The appeal is dismissed.
The appellant was convicted after jury trial in the District Court of one count of sexual violation by rape. At the same time he was found not guilty of a charge of sexual violation by oral connection arising out of the same incident and concerning the same complainant. The appellant’s appeal to the Court of Appeal was dismissed. He appeals further to this Court on two grounds. First, he says the trial Judge wrongly excluded evidence relevant to his defence of consent. Secondly, he says that the two verdicts were inconsistent, making his conviction on the rape charge unsafe.
The charges arose out of events after the appellant visited the complainant at her home at night. The complainant and the appellant and their respective partners were known to each other. Earlier in the evening both had been drinking at a pub in the small settlement where they both lived.
The complainant left the pub at about 9.30 pm and went home where she changed into her nightwear. She discovered a dead mouse in a trap in the kitchen. The complainant dislikes handling mice and usually asks someone else to do so. She telephoned the pub to try to persuade her daughter, who was there, to come to the house to remove the mouse. The daughter refused. The complainant said that the appellant had then turned up at the house, saying he had come to dispose of the mouse. The appellant said the phone had been passed to him at the pub (the daughter said it was possible, although she could not recall giving the phone to the appellant) and that the complainant had asked him to come to the house to deal with the mouse.
The evidence was that after the appellant had disposed of the mouse, he and the complainant sat outside smoking and then went back into the house and sat on a couch. The complainant said that the appellant pushed her down, took off her underpants and licked her vagina before raping her. The appellant, who had been drinking for some hours, when interviewed by the police said he could not recall the sexual activity. At trial however he acknowledged the sexual connection but said that it had been consensual. (He explained his statement to the police that he could not remember as having been made because he had been unable to contact his lawyer and did not want to make any statement until he had done so but did not want to appear uncooperative.)
The appellant said that the complainant, in asking him on the telephone to come to deal with the mouse, had said it would be worth his while to do so and that after the mouse had been disposed of she asked him to stay on and made it clear she wanted to have sexual intercourse with him. The complainant, who acknowledged being affected by alcohol but said it did not affect her recollection of what happened, denied consensual sexual activity. She said that “it all happened so fast” that she could not remember aspects of what happened but that she was pushing the appellant and telling him to stop.
As soon as the appellant left, the complainant sent a text message asking her daughter to come over to the house. When her daughter arrived, the complainant, who was very upset, complained of the sexual violation.
The excluded evidence
At the trial, counsel for the appellant, Mr Davison QC, advised the Judge that he proposed to call a man who would give evidence that some months previously he had been at the same pub in the middle of the day when the complainant had telephoned him and asked him to come to her house to block a hole through which mice were entering the house. When he arrived at the house the witness was uncomfortable to smell alcohol on the complainant’s breath and to find her in her nightwear. No sexual contact eventuated.
Counsel advised the Judge that the evidence was being called as a basis for the inference that the mouse issue was a “pretext” that the complainant had previously employed to “attract or encourage a male to be with her .... in the privacy of her home and in circumstances where she was dressed in a manner that might have indicated a willingness to engage in sexual activity to some degree or other”. He said that the evidence was probative because of its similarity to the events in issue and that it was necessary to enable the appellant to offer an effective defence.
The explanation given by counsel for the appellant indicated that the evidence to be called related to the complainant’s propensity to behave in a certain way. Under s 40(1)(a) of the Evidence Act 2006 “propensity evidence” is defined as:
.... evidence that tends to show a person’s propensity to act in a particular way or to have a particular state of mind, being evidence of acts, omissions, events, or circumstances with which a person is alleged to have been involved ....
Under s 40(3)(b), propensity evidence “about .... a complainant in a sexual case in relation to the complainant’s sexual experience may be offered only in accordance with section 44”. Section 44 provides:
The Judge ruled that the proposed evidence was inadmissible. He doubted its relevance but, in any event, considered that it was evidence that either went to the complainant’s sexual reputation or was evidence relating to the sexual experience of the complainant with someone other than the appellant. In the end, the Judge seems to have treated it as evidence of the sexual experience of the complainant with someone other than the appellant. He decided, under s 44(3) of the Evidence Act, that it was not contrary to the interests of justice for it to be excluded.
On appeal, the Court of Appeal considered that the evidence went to the reputation in sexual matters of the complainant, and so was excluded by s 44(2). It had accordingly been unnecessary for the trial Judge to consider under s 44(3) whether the evidence should be admitted (as would have been necessary if it had been evidence of sexual experience of the complainant with a man other than the appellant).
In this Court, Mr Lithgow QC for the appellant argues that the evidence did not engage s 44 on either basis because its relevance was simply to indicate that on another occasion the complainant had asked a man into her house to deal with mice. The evidence was said to be relevant to the complainant’s denial that she had invited the appellant to her house for the same purpose.
On the basis of the explanation given by trial counsel, the proposed evidence was relevant to a pretext used on another occasion by the complainant to get another man to her house in circumstances indicating preparedness to engage in sexual activity. That seems the only purpose of seeking to call evidence from the witness as to the way the complainant was dressed and his discomfort and impression that she was signalling her willingness to engage in sexual activity.
I consider that this basis for admission was to raise the propensity of the complainant to seek sexual opportunity in her home through the excuse of dealing with mice. As such, I consider it was evidence that was inadmissible under s 44(2) because it was evidence “that relates directly or indirectly to the reputation of the complainant in sexual matters”. Its purpose was to raise a foundation on which the jury would be invited to reason from the reputation sought to be established by the propensity evidence. Such reputational evidence is excluded by the legislative judgment that it is impermissible reasoning which is also deeply offensive to complainants and destructive of the protection of law. Because the sexual experience of the complainant with a person other than the defendant may be relevant for reasons other than to permit reputational propensity reasoning, the bar on such evidence is not absolute and it may be permitted for reasons of trial fairness by the judge. But the scheme of the legislation is that propensity reasoning from reputation is never permissible in sexual cases.
As the heading and contents of s 44 suggest, “sexual experience” covers “sexual experience of the complainant with any person other than the defendant” (evidence excluded under s 44(1), subject to judicial consent to its admission if it would be contrary to the interests of justice to exclude it) and “evidence .... that relates directly or indirectly to the reputation of the complainant in sexual matters” (the subject of absolute exclusion under s 44(2)). I consider that the linkage in s 40(3)(b) indicates that any evidence of sexual propensity of the complainant is covered by the two aspects of s 44. That is consistent with the sense of the legislation and avoids what would otherwise be an unaccountable gap, such as is countenanced by the interpretation preferred by William Young J. Evidence relating, directly or indirectly, to the reputation of the complainant in sexual matters is not, in this context, confined to evidence about general perceptions of the complainant. Reputational evidence is evidence as to the complainant’s character and attitudes from which propensity in sexual matters is sought to be taken.
I do not think that the evidence proposed can be said to have been directed to the sexual experience of the complainant with another person, as the judgment of McGrath, Glazebrook, and Arnold JJ treats it. That stretch in interpretation might deal with the particular case but leaves uncertain the application of s 44 to other cases of disposition, unless the suggestion that s 40(3) excludes such evidence is adopted in a case where it arises. The interpretation I prefer removes any such gap in s 44 and uncertainty in relation to the application of s 40(3). I consider that the evidence in the present case was, rather, evidence relating “directly or indirectly to the reputation of the complainant in sexual matters”. It was an attempt to set up such reputation by evidence and then to invite propensity reasoning from it. That is something, on the interpretation I prefer, the Evidence Act does not permit.
Such interpretation is I think consistent with the legislative history. The former s 23A of the Evidence Act 1908 required the leave of the judge before evidence could be led or questions put as to either the sexual experience of the complainant with someone other than the accused or “the reputation of the complainant in sexual matters” (the same division maintained in the current legislation in s 44). Leave could only be given if the evidence was of such direct relevance to the facts in issue (or the appropriate sentence) as to make its exclusion contrary to the interests of justice. Even so, the judge’s determination was subject to a proviso, the breadth of which suggested that general disposition or propensity in sexual matters was considered to be within the two exclusions provided for by s 23A(2):
Provided that any such evidence or question shall not be regarded as being of such direct relevance by reason only of any inference it may raise as to the general disposition or propensity of the complainant in sexual matters.
The Law Commission proposal for reform of s 23A would have strengthened the proviso by removing the ability of the judge to grant leave where evidence of the complainant’s reputation in sexual matters was for the purposes of supporting or challenging the complainant’s truthfulness or for the purpose of establishing the complainant’s consent. (Leave could be obtained if the evidence was introduced for some other purpose.) The Bill as introduced reverted to a provision in similar terms to the proviso in s 23A(3). It was changed on the recommendation of the Select Committee to its present form. The Select Committee’s explanation for the change was:
We recommend that clause 40 [now s 44] be amended to provide that no evidence can be given and no question be put relating to the sexual reputation of the complainant in sexual matters. We consider that any reference to a person’s sexual reputation is irrelevant and should not be admitted.
The evident legislative purpose was to reject as irrelevant the propensity in sexual matters of the complainant. That purpose is fulfilled if “reputation” is interpreted, as I think is the sense of the provisions in any event, to include evidence relating to disposition or character in sexual matters. The interpretations of the other members of this Court leave gaps, as they acknowledge. I therefore agree with the characterisation of the nature of the evidence adopted by the Court of Appeal rather than the view finally taken by the trial Judge and the majority in this Court. I consider that it was propensity evidence barred by s 44(2) because treated by the legislation as irrelevant.
Even if properly dealt with under s 44(3) as evidence of the complainant’s sexual experience with another man, however, I consider that the evidence was not sufficiently probative to justify its admission. In my view, the trial Judge was right to refuse to admit it.
If the complainant had denied that she relied on others to dispose of mice, the evidence that she had invited someone else in on another occasion to deal with mice might have been relevant. She did not however deny relying on others to dispose of mice for her, saying in her evidence that she always asked someone else to come over and deal with them.
There was uncontested evidence that the complainant had telephoned the pub to seek help from her daughter in disposing of the mouse in the trap. While the complainant denied that she had asked the appellant by telephone to come around to the house to deal with the mouse, she herself gave evidence that she had invited him into the house for that purpose when he turned up to say he had come to help. The evidence that the complainant had invited the appellant into the house to deal with the mouse, and that he had done so, was common ground. Had the evidence been shorn of the inference of sexual purpose that trial counsel had sought to call it for (which I consider was rightly excluded as inadmissible under s 44) it would have added nothing to the defence case that was not already before the jury and, if proffered for that purpose (as it had not been), would have been evidence the Judge might well have excluded under s 8(1)(b) of the Act. Even if the objectionable detail which invited reasoning from sexual propensity had been excluded, however, there is I consider no basis on which it could be said that the exclusion of the evidence of seeking help on another occasion to deal with mice in the house could have occasioned a miscarriage of justice.
Inconsistency in the verdicts
Where the verdicts returned by a jury are shown to be irreconcilable, so that there is “necessary inconsistency” and the verdicts “cannot stand together”, a conviction may be unreasonable, requiring it to be set aside under s 385(1)(a) of the Crimes Act 1961 or its replacement, s 232(2)(a) of the Criminal Procedure Act 2011. That position is reached in cases where the appellate court concludes that no reasonable jury could, on the evidence properly used, have arrived at different verdicts on the different counts, and that the conviction is unsafe.
The mere fact that the different verdicts were in cases where the case against the accused depended on the evidence of a single witness, does not set up a necessary inconsistency. And logical inconsistency in verdicts may not in itself cause the appellate court to regard a conviction as unsafe judged on the whole record unless the only possible explanation for the difference inevitably taints the conviction.
There are statements in some cases which suggest that inconsistency in verdicts may represent a “merciful” view of the facts which is open to juries. So, in the High Court of Australia, three members of the Court in 1996 expressed the view (referring to the Australian history of transportation) that exercising mercy is “a function which has always been open to, and often exercised by, juries”. Subsequently, in a further case in 2002, members of the High Court, while explaining why apparently inconsistent verdicts may reflect proper attention by the jury to the seriousness of the task and the onus of proof on the Crown, repeated the reference to the consideration that juries may act out of a sense of justice. Citing the first of these Australian High Court cases, although acknowledging New Zealand’s different history, the Court of Appeal in R v H suggested that it is a “valid reason” for inconsistency that the jury may have applied its “innate sense of fairness and justice”.
In the present case, the Court of Appeal, while indicating that there was a basis on which the verdicts could be reconciled, expressed the view that the jury might “also have thought that this single sexual incident should not result in two convictions” and that such reasoning was available to it:
In R v H this Court made it clear that a guilty verdict which is apparently inconsistent with an acquittal might be held to be not “unreasonable” if the innate sense of fairness and justice of the jury might properly have been applied in reaching the verdict of acquittal, for instance to avoid an unnecessary double conviction.
The other members of the Court endorse the approach that inconsistency may be valid if attributable to jury “leniency” and would allow “some scope” for such leniency in New Zealand law. I do not wish to be taken to accept that such approach is available to a jury. It is not consistent with the oath members of the jury take or with the instructions the judge is obliged to give to them as to their function. Juries are of course able legitimately to apply their own assessments in the context of the case when considering standards the law adopts: what is “reasonable”, or “indecent”, or “assistance”, for example, are all matters for jury assessment. But in the core function of determining guilt or innocence I do not think there is scope for other than conscientious discharge of the responsibility to decide on the evidence and according to law. Nor do I think that it is practicable for appellate courts to speculate about the processes of thought which might be said to be acceptable in order to give “some scope” to what is legitimate, as opposed to illegitimate, leniency.
Where verdicts are truly inconsistent (so that there is no rational explanation), the appellate court must still decide whether the circumstances are such that the conviction is unsafe. In some cases, of which the decision of the English Court of Appeal in R v Lewis is an illustration, the inconsistency may be immaterial to the safety of the conviction. In others, especially where the evidence depends on the credibility of a single witness (as is illustrated by R v Dhillon), the appellate court may conclude that the conviction is unsafe. It is not necessary to consider whether the conviction was otherwise unsafe, because I am satisfied that there is no inconsistency between the verdicts here.
The present case is not one of necessary inconsistency. To convict on both counts, the jury had to be satisfied beyond reasonable doubt of the sexual connection, the absence of consent, and the absence of reasonable belief on the part of the appellant that the complainant consented. The evidence given by the complainant was of a sudden assault by the appellant. She said that it all happened in a rush and had difficulty as a result in remembering aspects of the assaults. It was not impossible that the Crown did not exclude consent to the oral connection to the standard required but did exclude such consent in respect of the intercourse. It was perhaps more likely that the Crown had not excluded the appellant’s belief that the complainant had consented at the time of the oral connection but had excluded any such belief at the time of the penile penetration, because of her protests and pushing.
As with other members of the Court, I am of the view that the reference to “give and take” in the standard direction given to a jury that is unable to agree, may be misunderstood and is better avoided. But in the absence of necessary inconsistency in the verdicts, there is no basis for any speculation that the verdicts here represented a compromise.
For the reasons given, I would dismiss the appeal.
McGrath, Glazebrook & Arnold JJ
The appellant faced trial in the District Court on two counts of sexual violation, one by oral connection and one by rape. The counts arose from a single incident that occurred late one evening at the complainant’s home. The appellant admitted the conduct, but said that it was consensual. At his jury trial before Judge Duncan Harvey he wished to call certain evidence about the complainant. The Judge ruled that the evidence engaged s 44 of the Evidence Act 2006 and refused leave to call it.
The jury was unanimous in acquitting the appellant on the oral connection count but, by a majority verdict, convicted him on the rape count. Judge Harvey sentenced the appellant to imprisonment for a term of six years, three months. His appeal to the Court of Appeal was unsuccessful. This Court granted leave to appeal on two questions, namely whether:
in light of ss 7 and 44 of the Evidence Act 2006, the Judge should have permitted the applicant to lead all (or some) of the proposed evidence; and
the apparent inconsistency of the jury’s verdicts warranted the allowing of the appeal.
The appellant co-owned a small bar in a small coastal settlement. The complainant was a regular customer, drinking there most evenings. She was friendly with the appellant and his wife, she and her former partner having attended their wedding some years previously. The complainant lived on her own in a cottage in the settlement. Unfortunately, the cottage had something of a mouse problem and the complainant had a deeply-held aversion to mice.
The sexual violations were alleged to have occurred on a Sunday evening at the complainant’s cottage. The complainant had been away for a few days with her former partner and returned to the settlement around mid-afternoon. She contacted her adult daughter, who lived in the same settlement, and they agreed to go to the appellant’s bar together. The pair arrived at the bar sometime between 4 and 5 pm. The complainant remained there for about four hours. She accepted that she had drunk up to eight glasses of white wine over that period and said that, although she was “tiddly”, she was able to recall the evening’s events.
The appellant had spent the day playing in a golf tournament at a local course. He had had several drinks with the other golfers at the prize-giving, before going to the bar with some of his golfing companions, where he settled in for the evening.
The Crown case was that around 9.30 pm or shortly after, the complainant left the bar and got a lift to her cottage. When she arrived, she had a snack and changed into her nightie and dressing gown. At that point she noticed that there was a dead mouse in a mousetrap on her kitchen bench. (Her daughter gave evidence that while her mother was away, she had gone to the cottage and set several mouse traps, as she did occasionally.) The complainant telephoned the bar. The appellant’s father, who was working at the bar that evening, answered. The complainant asked to speak to her daughter. The appellant’s father misunderstood and handed the phone to another, similarly named, person. The complainant explained to him that she wanted to speak to her daughter. When her daughter came to the telephone, the complainant told her that there was a dead mouse in one of the traps and asked her to come and get rid of it. The daughter refused. The complainant hung up and sat down to watch television.
Soon after, the appellant arrived at the cottage, saying that he had come to deal with the mouse. The Crown’s case was that this resulted from the appellant having overheard part of the telephone conversation. The complainant invited him in and pointed out where the dead mouse was. Having disposed of the mouse, the appellant suggested that he and the complainant go outside to the deck for a cigarette, which they did. When they went back into the cottage, they sat on a couch in the living room. The complainant said that the appellant suddenly pushed her down against the back of the couch with her legs spread. She said he removed her underwear and licked her vagina. The complainant said she told the appellant to think about his pregnant wife and child and tried to push him away but could not because of his size. The appellant then raped her, although she again attempted to push him away and said: “Stop, stop, stop”. The appellant left shortly after.
The complainant immediately tried to telephone her daughter but, as her telephone was engaged, sent her a text asking her to come over right away. When her daughter arrived at the cottage, the complainant, who was in a distressed condition, told her that the appellant had raped her. A complaint was made to the police two days later. A subsequent medical examination revealed bruising the complainant’s thighs, in particular on the inside of her left thigh.
When interviewed by the police, the appellant said that he had gone to the complainant’s cottage to get rid of the mouse but said he could not recall any sexual activity. At trial, he accepted that the sexual activity had occurred. He gave evidence that after the complainant had spoken to her daughter about getting rid of the mouse, the daughter had handed the telephone to him. He spoke to the complainant, who asked him to come to the cottage to get rid of the mouse and said that she would make it “worth [his] while” if he did. The appellant said that after the two had had their cigarettes on the deck, the complainant had asked him not to leave and invited him to sit on the couch beside her. He said she made it clear that she wanted to engage in sexual activity with him. He said that the oral sex and intercourse which followed were consensual and described how they occurred. When asked in cross-examination why he had told the police that he could not recall any sexual activity, he said that he had attempted to contact his lawyer but had been able to contact only his lawyer’s secretary, who had told him not to answer any questions. He said he did not want to be unhelpful to the police so he decided to say that he could not remember what happened.
In her evidence, the complainant’s daughter said that when she took the telephone call from her mother at the bar, she found it difficult to hear because of the noise. There was a small serving bar, so she went round behind that to take the call. She thought the appellant and his father were also behind the bar at the time. She said that when her mother asked her to go over and get rid of the mouse, she told her not to be silly and to do it herself. Once the call was over she thought she made a comment along the lines of “Mum and that dead mouse”. She accepted that she did not recall what she had done with the telephone after speaking to her mother and acknowledged that it was possible that she might have given it to the appellant.
The jury retired to consider their verdicts at 3.17 pm. At 6.30 pm the jury sent a note to the Judge saying that they could not agree. After consulting counsel, Judge Harvey gave a Papadopoulos direction, in the form approved by the Court of Appeal in R v Accused (CA87/88). The jury retired again at 6.41 pm. They returned at 9.20 pm to give their verdicts. They were unanimous in acquitting the appellant on the oral connection charge but convicted him on the rape charge, by a majority.
The appellant appealed to the Court of Appeal against both conviction and sentence. His appeal was dismissed. Before us, he argued that Judge Harvey was wrong not to allow him to lead certain evidence about the complainant, with the result that there was a miscarriage of justice (under s 385(1)(c) of the Crimes Act 1961), and that the jury’s guilty verdict was unreasonable as it was inconsistent with the acquittal and so should be set aside (under s 385(1)(a) of the Crimes Act).
The proposed evidence
The appellant was represented at trial by Mr Davison QC. While the complainant was giving evidence, Mr Davison advised the Judge that he intended to call a man to give evidence about an incident with the complainant some months previously, about which he wished to cross-examine her. He was concerned that the proposed evidence might engage s 44 of the Evidence Act. Mr Davison advised the Judge that the proposed witness would say that some months earlier (the exact date was uncertain) the complainant contacted him while he was at the same bar around midday and asked him to call in at her cottage because she was having a problem with mice. When the man arrived at her cottage, the complainant was still in her nightie and dressing gown. The man said that he could smell alcohol, although he could not say that the complainant had been drinking. The complainant invited him in. The man dealt with the mouse problem (which involved blocking a small hole through which the mice had been entering the cottage) and then left. He would say that he felt distinctly uncomfortable in the circumstances.
Mr Davison described the effect of the evidence in this way:
The essence of the evidence that would be led and which is to be put in cross-examination pertains to a method that was employed in this instance and employed the defence says on an earlier occasion to attract or encourage a male to be with her .... in the privacy of her home and in circumstances where she was dressed in a manner that might have indicated a willingness to engage in sexual activity to some degree or other. And so – and there was in relation to this earlier matter that the defence wishes to put, no sexual activity of any kind. It was simply .... an inference that could have been drawn as to her motivation from her attire ....
When asked precisely what inference he wished the jury to draw, Mr Davison said that what was important was “the use of [a] pretext”. In his ruling, Judge Harvey summarised the position as follows:
Mr Davison said that what has occurred [in this case], coupled with what has occurred on the earlier occasion, might well be seen as a method [the complainant] has employed to encourage men to be with her in her home in dress which may seem to encourage sexual activity. Mr Davison wants the jury to simply hear this evidence and then draw from that evidence what inferences they see fit. Mr Davison stresses that this is not a situation where he is attempting to blacken the complainant’s character, and he argues in fact that the evidence does not engage s 44. If, however, the Court considers that s 44 is engaged, then Mr Davison argues that this cross-examination is essential to enable the accused to proffer an effective defence. The real relevance is that the evidence is probative due to its similarity and its relevance is to the issue of consent.
The trial Judge’s ruling
As we have said, Mr Davison raised the issue because he considered it possible that s 44 of the Evidence Act might be engaged, although he argued that it was not. Relevantly, s 44 provides:
In considering whether the evidence could be called, Judge Harvey said that he struggled to see its relevance. The Judge also said that the way in which the defence wanted to use the evidence invited the jury to speculate about whether the complainant had invited the proposed witness over in order to engage in sexual activity. It is a little unclear under which subsection of s 44 the Judge considered that the evidence fell. The Judge described the evidence as, indirectly, going to reputation. On that basis, it would have fallen within the absolute prohibition contained in s 44(2). However, the Judge went on to consider the test contained in s 44(3), which refers back to s 44(1) dealing with evidence of sexual experience with someone other than the accused. The Judge concluded that it was not contrary to the interests of justice to exclude the evidence.
The Court of Appeal considered that the evidence went indirectly to the complainant’s sexual reputation and so fell within the absolute prohibition in s 44(2). The trial Judge’s consideration of the test in s 44(3) was accordingly unnecessary.
Mr Lithgow QC argued that the appellant should have been allowed to adduce, and cross-examine on the basis of, the proposed evidence because it did not involve evidence of sexual experience or reputation, so that s 44 was not engaged. Rather, the evidence went to the “simple proposition that the complainant had previously used the available pretext of a mouse problem to be solved in order to invite a man to her house”. Mr Lithgow argued that the evidence could properly be used to challenge the complainant’s denial that she asked the appellant to come to the house.
We should emphasise that Mr Lithgow accepted that nothing could properly be taken from the fact that the complainant was still wearing her nightie and dressing gown when the man went to her cottage in the middle of the day. He submitted that the sole purpose of the proposed evidence was to show that on a previous occasion the complainant had asked a man to come to her house to deal with a mouse problem and that this was relevant to her denial that she had done so on this occasion.
Section 44 – general
Section 44 largely replicates s 23A of the Evidence Act 1908, New Zealand’s original “rape shield” provision enacted in 1977. Rape shield provisions control the extent to which complainants in sexual cases may be questioned about their previous sexual history. Such provisions are intended to reduce the humiliation and embarrassment faced by complainants and to prevent the use of reasoning based on erroneous assumptions arising from a complainant’s previous sexual history. In Bull v R, the majority of the High Court of Australia identified two erroneous lines of reasoning that might arise in this context: because a complainant has a particular sexual reputation, disposition or experience, either (1) he or she is the kind of person who would be more likely to consent to the activity which is the subject of charges or (2) he or she is less worthy of belief than a complainant who does not have those characteristics. Against these concerns, however, must be balanced the defendant’s right to a fair trial and the right to present an effective defence in particular.
One important difference between ss 44 and 23A is that there is an absolute prohibition on leading evidence of sexual reputation under s 44(2) whereas under s 23A(3)(a), evidence of sexual reputation could be led with the leave of the judge if it was of such direct relevance to a fact in issue that it would be contrary to the interests of justice to exclude it (reputation evidence could go to an accused’s belief in consent, for example). As McLachlin J concluded in her judgment for the majority in R v Seaboyer, prohibitions on leading sexual conduct evidence which do not permit the exercise of judicial discretion create the risk that some defendants will not be able to present legitimate defences.
As Mr Lithgow pointed out, s 44 is not without its difficulties. It refers to sexual experience and sexual reputation but does not refer to sexual disposition, as the proviso to s 23A(3) did and the equivalent provision in Western Australia does. Analytically, sexual disposition is a distinct concept from sexual experience or sexual reputation, although a particular sexual disposition may lead a person to have particular sexual experiences with others or to develop a particular sexual reputation. But sexual disposition may be revealed in a way that does not involve experience with others or lead to a relevant reputation, as, for example, through sexual fantasies recorded in a personal diary. This raises the question of how s 44 applies to such evidence.
The answer may lie in s 40 of the Evidence Act. Disposition evidence is, of course, propensity evidence. Under s 40(3)(b), propensity evidence about a complainant in a sexual case in relation to that person’s sexual experience may only be offered in accordance with s 44 of the Act. On the face of it, this language is apt to include evidence of, say, sexual fantasies in a personal diary (even though not manifested in sexual experience with others or in sexual reputation). Accordingly, it is arguable that the effect of s 40(3)(b) is that, because sexual disposition is not referred to explicitly in s 44, evidence of sexual disposition involving sexual fantasies and such like cannot be led at all. Such an outcome seems consistent with the policy underlying s 44 and other rape shield provisions.
A further difficulty arising out of s 44 is that, although analytically distinct, the concepts of experience, reputation and disposition may overlap in practice, which may result in problems of application. So, for example, evidence of a complainant’s sexual experience with others may also relate “directly or indirectly” to the complainant’s reputation in sexual matters. The existence of such difficulties suggests that s 44 needs further legislative clarification.
Section 44 – this case
If the only purpose of the evidence was, as Mr Lithgow submitted, to show that on a previous occasion, the complainant had asked a man to come to her cottage to deal with a mouse problem, we accept that it would, in principle, have been relevant to one of the issues in the present case, namely whether the complainant had asked the appellant to come over to dispose of the dead mouse. How the complainant was dressed at the time would have been irrelevant, and s 44 would not have been engaged. At trial, prosecuting counsel accepted this analysis. But the evidence would have been, at best, peripherally relevant given the way the trial developed, because the complainant accepted in her evidence that she did not deal with dead mice herself but asked someone else to come over and deal with them, generally her daughter or former partner. She said this was because she hated mice and did not like to touch or deal with dead ones.
It is clear, however, that Mr Davison wanted to take the evidence further than that. He wanted the jury to draw an inference, based principally on the complainant’s attire, that she had invited the man over on the previous occasion on the pretext that she wanted him to deal with the mouse problem but in reality because she was interested in having sex with him. Mr Davison wanted the jury to take this into account when considering the complainant’s denial that this was what happened on the present occasion.
Although Mr Lithgow disavowed this use of the proposed evidence in his oral submissions, his position was somewhat equivocal given the extract from his written submissions quoted at  above. There he referred to the mouse problem as being a “pretext”. A pretext is “an ostensible or alleged reason or intenon”.54 Describing the mouse problem as a “pretext” invites the question, “For what?” From the defence perspective, the answer had to be that the complainant wanted to create the opportunity for a sexual liaison. On that basis, the proposed evidence did engage s 44: it fell within s 44(1) being evidence of the complainant’s sexual experience with another person. The fact that no sexual conduct occurred does not take the incident outside the scope of “sexual experience” with another. If the complainant had actively propositioned the proposed witness to have sexual intercourse with her and he had refused, evidence of that would undoubtedly be evidence of sexual experience with another. In effect, this is what the “pretext” argument involved.
We do not agree with the Court of Appeal that the proposed evidence was evidence of sexual reputation, falling within the absolute prohibition in s 44(2). Although the proposed witness said he felt uncomfortable in the circumstances, his evidence did not concern the beliefs or opinions that other people held about the complainant (reputation); rather, it concerned the complainant’s interactions (allegedly sexual) with a person other than the appellant on a previous occasion (experience). It is difficult to see how an incident with one person could amount to evidence of sexual reputation (unless, of course, it became the foundation for a more widespread belief about the complainant).
As evidence of the complainant’s sexual experience with a person other than the appellant, the proposed evidence would only have been admissible if the trial Judge had granted leave under s 44(3). He refused leave, and in our view was right to do so. Under s 44(3), to admit the evidence the Judge had to be satisfied that it was “of such direct relevance to facts in issue in the proceedings .... that it would be contrary to the interests of justice to exclude it”, the so-called heightened relevance test. The only factor that was arguably suggestive of any sexual purpose was that the complainant was still wearing her nightie and dressing gown in the middle of the day when the man arrived. There was no evidence that she had said or done anything else that might be regarded as suggestive. As previously noted, Mr Lithgow accepted that the fact that the complainant was still in her nightie and dressing gown could not, on its own, legitimately be used to support an inference of sexual purpose. As the Judge said, the jury was really being asked to speculate as to the complainant’s purpose in inviting the proposed witness to her home, and was being encouraged to draw the inference that the complainant invited men to her home on a pretext, in order to engage in sexual activity with them.
In some circumstances, the evidence might have been admissible even though it could not, on its own, support an inference of a sexual purpose. For example, if the complainant had claimed that she would never have invited the appellant to the cottage while dressed in her nightwear, the proposed evidence could have been called to contradict that assertion. But the complainant made no such claim.
We conclude, then, that the Judge was right to rule that the proposed evidence was inadmissible for the purpose for which Mr Davison wished to use it. We agree with Mr Lithgow, however, that the fact that the complainant had on a previous occasion telephoned one of the patrons at the bar to ask him to come and deal with a mouse problem (albeit of a different sort) was potentially relevant in light of the appellant’s claim that the complainant had telephoned and asked him to come over to deal with the dead mouse on this occasion. But we consider that the evidence was of little or no relevance in fact, given the complainant’s acknowledgement in evidence that she did not deal with dead mice herself but asked others to come to her cottage to dispose of them. In the result, then, we do not consider that there was any risk of a miscarriage of justice in this respect.
The Court of Appeal rejected the appellant’s contention that the jury’s verdicts were inconsistent, on two grounds: first, that there was a logical basis on which the difference in the verdicts could be explained and second, because the jury may have applied its innate sense of fairness and justice in convicting of only one offence arising out of what was in reality a single incident. Mr Lithgow challenged both lines of analysis.
The purpose of an inconsistent verdict argument is to show that a jury’s guilty verdict is unreasonable and should be quashed. It invokes s 385(1)(a) of the Crimes Act (see now, ss 232(2)(a) and 240 of the Criminal Procedure Act 2011). As this Court held in R v Owen, a jury’s verdict “will be unreasonable if, having regard to all the evidence, the jury could not reasonably have been satisfied to the required standard that the accused was guilty”. Before examining the arguments raised in this case we will set out the principles relevant to inconsistent verdict arguments.
As the majority judgment of the High Court of Australia in MacKenzie v R noted, courts faced with submissions of jury verdict inconsistency must take account of conflicting objectives. On the one hand, courts seek to uphold the integrity of the jury system. Our criminal justice process places a high value on the role of juries as fact-finders in criminal cases. The law acknowledges the fundamental importance of juries in a variety of ways, for example, by conferring a right to trial by jury in respect of offences punishable by imprisonment for two or more years, by criminalising attempts to influence jurors by threats, bribes or other corrupt means, and by protecting jury deliberations from disclosure. Courts will always be reluctant to conclude that juries have not acted consistently with their oaths. On the other hand, the courts’ concern with doing justice may be engaged in respect of particular verdicts. Where they deliver multiple verdicts which are not capable of logical reconciliation, juries give some insight into their thought processes. Logically irreconcilable verdicts may indicate that the jury’s thinking has gone awry in some fundamental way: in particular, the jury may have acted on a misunderstanding of the law or reached an illegitimate compromise. In such circumstances, a court may feel it necessary to intervene in order to ensure that justice is done, despite its respect for the jury’s function in the criminal justice process. An obvious difficulty, however, is how a court can determine the basis for a jury’s apparently inconsistent verdicts given that jury deliberations are protected from outside scrutiny.
The majority judgment in MacKenzie went on to identify six general propositions drawn from the authorities. These propositions, which are largely echoed in the New Zealand authorities, are as follows:
There is a distinction between cases involving legal inconsistency and those involving factual inconsistency. Legal inconsistency occurs when two verdicts cannot, as a matter of law, stand together. Examples are where a jury convicts a person of both an attempt to commit an offence and the completed offence or as the thief and the receiver of the same property on the same occasion. Factual inconsistency occurs where, given the evidence, two verdicts cannot stand together.
Factual inconsistency can arise either between verdicts involving the same accused or between verdicts involving different persons charged in connection with related events. In R v Pittiman the Supreme Court of Canada said that it will often be more difficult for an appellant in a multiple accused case to establish inconsistency as there is likely to be greater scope for differing verdicts in such cases.
In relation to factual inconsistency arising from “guilty” and “not guilty” verdicts on a multiple count indictment against one defendant, the test is one of “logic and reasonableness”. As the Court of Appeal said in R v Irvine:
|The question which we must ask ourselves is whether the acquittal on count one, in all the circumstances of this particular case, renders the verdict of guilty in respect of count two unsafe, in the sense that no reasonable jury could have arrived at different verdicts on the two different counts.|
Courts are reluctant to conclude that that jury verdicts are inconsistent, both because the jury’s function must be respected and because there is general satisfaction with the way juries perform their role. If there is some evidence to support the verdict said to be inconsistent, an appellate court will not usurp the jury’s function by substituting its view of the facts for that of the jury. We note in this connection that the Court of Appeal emphasised in R v O (No 2) that any reasonable explanation for the difference between the two verdicts “must be found in the evidence properly used”. The Court said:
It will not be a reasonable explanation if it depends on a use of evidence or a process of reasoning which the law does not permit.
In addition, the majority judgment in MacKenzie acknowledged that an appellate court “may conclude that the jury took a ‘merciful’ view of the facts upon one count: a function which has always been open to, and often exercised by, juries”. The Court of Appeal accepted this view in R v H, a case to which we will return.
There will be cases where the different verdicts returned by a jury represent “an affront to logic and commonsense which is unacceptable and strongly suggests a compromise of the performance of the jury’s duty”. In such cases an appellate court will intervene. Hard and fast rules are not possible; rather, the assessment must be made on a case by case basis. In R v Pittiman, the Supreme Court of Canada said that inconsistent verdicts may be held to be unreasonable “when the evidence on one count is so wound up with the evidence on the other that it is not logically separable”.
The obligation to establish inconsistency rests with the person challenging the conviction. Where inconsistency is established, the court must make such consequential orders as the justice of the case requires.
The High Court of Australia returned to the topic of inconsistent verdicts in MFA v R and confirmed the principles articulated in MacKenzie.
In R v Pittiman, the Supreme Court of Canada made the further point that “[w]hile an appellate court inevitably compares the basis for acquittals as well as convictions in assessing inconsistent verdicts, the decisive question is not whether the acquittals are reasonable, but whether the conviction was not”. This focus on the reasons for conviction rather than the reasons for acquittal explains in part why the United States Supreme Court has refused to countenance criminal appeals based on inconsistent verdicts in relation to federal offences, at least in cases of factual inconsistency. The two leading cases are Dunn v United States and United States v Powell. In Powell, the Court accepted that Dunn established “the unreviewable power of a jury to return a verdict of not guilty for impermissible reasons”. The Court went on to say that the rule that a defendant may not upset an inconsistent verdict embodied a “prudent acknowledgment of a number of factors”.
The first factor was that inconsistent verdicts are not necessarily a windfall to the prosecution at the defendant’s expense but may be unfair to the prosecution, which has no ability to appeal an acquittal. As the Court rather colourfully put it:
Inconsistent verdicts therefore present a situation where “error,” in the sense that the jury has not followed the court’s instructions, most certainly has occurred, but it is unclear whose ox has been gored.
In light of these two factors – uncertainty and the prosecution’s inability to appeal an acquittal – the Court considered that it was “hardly satisfactory to allow the defendant to receive a new trial on the conviction as a matter of course”. The Court acknowledged that inconsistent verdicts may be the product of what it described as “jury lenity” and saw this as a reflection of the jury’s historic function in criminal trials of providing a check against arbitrary or oppressive exercises of power by the executive branch.
The second factor was that the Court rejected as “imprudent and unworkable” a rule that would allow defendants to challenge verdicts as inconsistent on the ground that they were the result of adverse jury error rather than leniency. The Court said:
Such an individualized assessment of the reason for the inconsistency would be based either on pure speculation, or would require inquiries into the jury’s deliberations that courts generally will not undertake. Jurors, of course, take an oath to follow the law as charged, and they are expected to follow it.
Finally, the Court noted that criminal defendants have protection against juror irrationality or error through the independent review of the sufficiency of evidence that is available through trial and appellate courts. The Court said:
Sufficiency-of-the evidence review involves assessment by the courts of whether the evidence adduced at trial could support any rational determination of guilty beyond a reasonable doubt .... This review should be independent of the jury’s determination that evidence on another count was insufficient. The Government must convince the jury with its proof, and must also satisfy the courts that given this proof the jury could rationally have reached a verdict of guilty beyond a reasonable doubt. We do not believe that further safeguards against jury irrationality are necessary.
In the New Zealand context, such review is most obviously available through the “no case to answer” procedure and the “unreasonable verdict” ground of appeal.
Some commentators have been critical of the Supreme Court of the United States’ approach and a small number of state courts have recognised some scope for inconsistent verdict appeals in relation to state offences. It is perhaps significant that the Supreme Court appears to have seen itself as having to choose between two inflexible rules: either that inconsistent verdicts would result in a new trial on the relevant conviction “as a matter of course” or that inconsistent verdicts would never be a basis for allowing a conviction appeal. Presumably this was because the Court considered an individualised approach to be unworkable. This scepticism about an individualised approach is generally reflected in the decisions of state courts which permit inconsistent verdict appeals, in the sense that most permit such appeals only in respect of legally inconsistent verdicts. The reason for this appears to be that the relevant courts consider that they have no ability to determine whether factually inconsistent verdicts result from mistake, compromise or leniency (that is, a merciful verdict) on the part of the jury. The result is, then, that apparently inconsistent jury verdicts that result from mistake, compromise or leniency are tolerated as an incident of the jury system.
The law of New Zealand, like that of Australia, Canada and the United Kingdom, does allow some scope for inconsistent verdict appeals, both factual and legal, although that scope is limited. Accordingly, we turn to consider the two bases upon which the Court of Appeal concluded that there was no inconsistency and upheld the appellant’s conviction for sexual connection by rape.
We note at the outset that, for the purposes of analysis, we will deal with the “innate sense of justice” point separately from the “logical basis for differentiation” point. There is a degree of artificiality in this as there may be some interaction between the considerations that arise under each point in the minds of jurors, as Gleeson CJ, Hayne and Callinan JJ noted in their joint judgment in MFA. It may be, for example, that jurors in a particular case will have some relatively minor doubts about one count which, when combined with a sense that a conviction on another count fairly captures the accused’s overall culpability, lead them to acquit on that count while convicting on the other.
Logical basis for differentiation
Before the Court of Appeal, the appellant’s then counsel argued that the jury’s verdicts were inconsistent because all the factors relied upon by the Crown to indicate the absence of consent, or of any basis for a reasonable belief in consent, were present throughout the entire incident and so related to both charges (that is, factual inconsistency). After considering the evidence, the Court of Appeal rejected that contention. The Court said:
On the basis of the evidence we have summarised ...., the jury could quite possibly have given [the appellant] the benefit of a reasonable doubt as to whether the complainant was consenting to the oral sex, or whether on reasonable grounds [the appellant] thought she was. But that doubt evaporated in the face of the complainant’s emphatic evidence that she tried to push [the appellant’s] body away with her hands and repeatedly told [the appellant] to stop when he set about having sexual intercourse with her.
Mr Lithgow challenged this conclusion, essentially on the same basis as counsel before the Court of Appeal.
As Gleeson CJ, Hayne and Gummow JJ noted in their joint judgment in MFA, juries are typically given two directions which bear on the issue of inconsistency. The first is that they should consider each count, and the evidence relevant to each count, separately. What Judge Harvey said to the jury in the present case followed the standard pattern:
As I have said, you may reach different verdicts on different counts. You see, essentially we have had two separate trials going on over the past five days. For convenience both counts are heard together. However, when you retire to consider your verdicts you should isolate the evidence that relates to each individual charge and be very careful that you do not use evidence in relation to one charge to support or bolster evidence relating to another charge. For example, it would not be proper for you to say that, well because we find the accused is guilty of count 1 he must be guilty on count 2. Each charge falls to be considered solely on the evidence that has been given in relation to that charge. But, of course, recognising that in this case much of the evidence does in fact relate to both.
The second instruction is that the jury may accept the evidence of a witness in whole or in part. Again what Judge Harvey said to the jury in the present case is typical:
.... the sole responsibility for deciding all questions of fact rests with you. It is for you to decide what evidence you accept, and what evidence you reject. It is for you to decide what weight you will give to any part of the evidence. It is open to you to accept some parts of what a witness has said and reject other parts.
This point is important in a case such as the present, where the Crown case rested almost entirely on a complainant’s evidence, which the accused contested – a so-called “she said, he said” case. In R v Shipton, the Court of Appeal said:
Time after time in appeals to this Court it is argued, as counsel argued here, that because the jury must have “disbelieved” a witness to acquit on one count, it was inconsistent to rely on her to convict on another count. The argument is utterly fallacious; there may be all sorts of valid reasons why the jury may be convinced by a witness on one count but not on another. To put this another way, there is no reason why credibility must be static. As was said in R v G  Crim LR 483, “A person’s credibility is not a seamless robe, any more than is their reliability”. It is not necessarily illogical for a jury to be convinced as to the credibility of some aspects of one person’s story, but not as to others, a fortiori where it is convinced, but not beyond a reasonable doubt.
In their joint judgment in MFA, Gleeson CJ, Hayne and Callinan JJ gave a rather fuller explanation of the point, as follows:
.... [E]mphasis will invariably be placed upon the onus of proof borne by the prosecution. In jurisdictions where unanimity is required, such as New South Wales, every juror must be satisfied beyond reasonable doubt of every element in the offence. In the case of sexual offences, of which there may be no objective evidence, some, or all, of the members of a jury may require some supporting evidence before they are satisfied beyond reasonable doubt on the word of a complainant. This may not be unreasonable. It does not necessarily involve a rejection of the complainant’s evidence. A juror might consider it more probable than not that a complainant is telling the truth but require something additional before reaching a conclusion beyond reasonable doubt. The criminal trial procedure is designed to reinforce, in jurors, a sense of the seriousness of their task, and of the heavy burden of proof undertaken by the prosecution. A verdict of not guilty does not necessarily imply that a complainant has been disbelieved, or a want of confidence in the complainant. It may simply reflect a cautious approach to the discharge of a heavy responsibility. In addition to want of supporting evidence, other factors that might cause a jury to draw back from reaching a conclusion beyond reasonable doubt in relation to some aspects of a complainant’s evidence might be that the complainant has shown some uncertainty as to matters of detail, or has been shown to have a faulty recollection of some matters, or has been shown otherwise to be more reliable about some parts of his or her evidence than about others.
Mr Lithgow accepted that a jury may be entitled in some situations to convict on one count and acquit on another on the basis that they accept the evidence of a critical witness in relation to one count but not the other. He argued, however, that this was not appropriate in the present case because the evidence relating to each count was essentially the same: the complainant said that she resisted and told the appellant to stop during both the oral sex and the intercourse, which occurred within a short time period as part of a single interaction. At its heart, Mr Lithgow’s proposition was that, given the nature of the evidence, there had to be convictions on both counts or on neither – an all or nothing case. This was, of course, directly contrary to the Judge’s instructions to the jury.
A similar argument was accepted by the Criminal Division of the English Court of Appeal in R v Dhillon. There the Court observed that it was “notoriously difficult successfully to challenge a jury’s verdict on the ground that inconsistent verdicts have been returned”. To succeed there had to be a logical inconsistency in the verdicts which could not be explained by a line of reasoning reasonably available to the jury. In sex cases where sexual incidents are alleged to have occurred on separate occasions, inconsistency will not arise simply because the jury accepted part of a complainant’s evidence but was not sure about other parts. It may be different, however, where the various offences are “simply different facets or acts in the course of a single sexual encounter”. In those circumstances, the Court said:
.... if the jury is unsure of the complainant’s evidence with respect to one count on the grounds that it may be unreliable or lacking credibility, it is likely to be more difficult than it would be with respect to chronologically separate encounters for a jury to be sure that the evidence on the other counts is reliable and credible.
The Court said that, when assessing the jury’s reasoning, it was “important to have regard to how a fair minded jury would approach the evidence that was properly before them”. Adopting that approach, the Court concluded that the verdicts in that case were logically inconsistent as there was no satisfactory explanation which could render them consistent given the evidence.
In Dhillon, as in the present case, several (in that case, five) sexual offences were alleged to have been committed in the course of a single sexual encounter. The appellant was convicted on two counts and acquitted on three. While no issue was raised by his acquittal on two of the counts, his acquittal on the third raised the question whether the guilty verdicts were logically consistent. The two counts on which he was convicted involved digital penetration of the complainant’s vagina and sexual assault by touching the complainant’s breasts. The count on which he was acquitted involved oral sex. The appellant admitted that the three incidents had occurred but said that they were consensual.
The Court said that the jury’s verdicts could only be explained on the basis either that:
the jury was unsure whether the oral sex had occurred but was sure about the other two acts; or
the jury considered that the complainant had consented to the oral sex but not to the other less serious acts; or
the jury considered that the appellant had a reasonable belief that the complainant consented to the oral sex but not to the touching of the vagina and breasts.
The Court rejected the first possible explanation, on the basis that no reasonable jury could have found that two of the incidents occurred but not the third. Counsel accepted that the second explanation was untenable. The Court rejected the final explanation. The evidence showed that the acts had occurred at the same time – indeed, it was not clear from the evidence whether the touching of the breasts had preceded the oral sex or vice versa. Moreover, neither the Judge nor counsel had suggested that the appellant might have had a reasonable belief in consent in respect of one act and not the other.
Accordingly, the Court concluded that the verdicts were inconsistent. That did not necessarily mean that they were unsafe, however, and that the appeal should be allowed. The Court considered the possibility that, on the evidence, the appellant was unjustifiably acquitted of the oral sex charge rather than that he was unjustifiably convicted on the other charges. While accepting that there was some force in that proposition, the Court did not accept it. Noting that the verdicts were by a majority and had come after six hours of deliberation, the Court considered that there was a real risk that the jury had reached an illegitimate compromise and quashed the convictions.
In the present case, because the appellant gave evidence, the Judge gave a third direction which bears upon the issue of inconsistency, namely the tripartite direction. Judge Harvey directed the jury that:
If they considered that the defence evidence was credible, reliable and a convincing answer to the Crown case, they should acquit.
If they were not convinced by the defence evidence but it left them unsure as to the true position, the accused would have raised a reasonable doubt and should be acquitted.
Finally, the Judge said:
However, you may think the defence evidence or parts of it is entirely unconvincing, and you simply reject it as being unworthy of belief. Now, if that is your view, you should be careful not to jump from that conclusion to an automatic conclusion of guilt or even regard it as somehow adding to the case against the accused, because it doesn’t. You should simply set that evidence aside, go back to the rest of the evidence and ask yourselves whether on the basis of that evidence you are satisfied about guilt.
In her evidence in chief, the complainant described an evolving situation in which the appellant used his body (he was much bigger than she) to pin her left leg against the back of the couch, removed her underpants, licked her vagina and then penetrated her. While this was occurring, she attempted to push the appellant away, asked him to think about his pregnant wife and child and told him to stop. Mr Davison cross-examined the complainant closely about the details of what happened, for example as to how exactly the appellant had removed her underpants and how the appellant’s pants had come to be removed. At numerous points in her answers to these questions, the complainant said she could not remember, explaining several times that it had all “happened so fast”. When she was asked whether she was pushing the appellant away when he was performing oral sex, the following exchange occurred:
The jury clearly rejected the appellant’s claim that the complainant consented to the sexual intercourse and, in all probability, also rejected his claim that she consented to the oral sex. Having reached that point, the jury had to consider whether it was a reasonable possibility that the appellant thought, on reasonable grounds, that the complainant was consenting. Following the Judge’s instructions, they would have considered the complainant’s evidence to satisfy themselves that the Crown had established all the ingredients of the offences, including lack of reasonable belief in consent. In his closing address, Mr Davison read the jury some of the extracts from the notes of evidence in which the complainant had said that she could not remember some of the details, in particular about how her underpants came to be removed. If the jury considered the complainant’s account by reference to what she said under cross-examination rather than what she said in her evidence in chief and took account of the fact that she had been drinking, it is possible that they would have been left with a reasonable doubt about the point at which the appellant could no longer have had a reasonable belief in consent. The jury may well have given the appellant the benefit of the doubt in relation to the oral sex but concluded that by the time he penetrated the complainant, her lack of consent had been manifested unequivocally, both by words and conduct, so that he could not possibly have had a reasonable belief that she was consenting. This is an analysis which, in our view, was open on the evidence.
Accordingly, we agree with the Court of Appeal that, in light of the evidence, the jury’s verdicts were not inconsistent.
Innate sense of justice
In the judgment under appeal, the Court of Appeal gave a further ground for dismissing the appeal, as follows:
As we have said, the Court of Appeal in R v H accepted that apparently inconsistent verdicts could be justified on the basis that they were the result of the jury’s innate sense of justice. Mr Lithgow argued that R v H was wrongly decided and should be overruled or limited, for reasons which we summarise below.
In R v H, the appellant was charged with seven representative counts of sexual offending against the daughter of his partner, spanning a seven year period. The appellant denied the complainant’s allegations but was convicted on four counts and acquitted on three. He appealed on the ground of inconsistent verdicts. The argument focussed on the appellant’s acquittal on one count but convictions on four other closely related counts, in respect of which the complainant’s evidence was identical.
The Court reviewed a number of Australian, Canadian, New Zealand and United Kingdom authorities, before accepting that a conviction which is apparently inconsistent with an acquittal might not be held to be unreasonable if “the innate sense of fairness and justice of the jury might properly have been applied in reaching the verdict of acquittal, for instance to avoid an unnecessary double conviction”. The Court noted that the appellant had initially faced one representative charge of rape covering a four and a half year period but this had been amended to include two representative charges of rape, one covering a little over 20 months of the four and a half year period and the other covering the remainder. This amendment had been made to reflect an amendment to s 128 of the Crimes Act, whereby the word “vagina” had been replaced by the word “genitalia”. The Court said that the jury may well have refused, because of a sense of justice and fairness, to convict on two counts of rape “simply because of a legislative quirk”.
Mr Lithgow argued that the approach accepted in R v H permitted a court to speculate inappropriately as to the jury’s reasons for acquitting in an inconsistent verdict situation. Juries are told that the judge will instruct them on the law, which they must accept, but they are to determine the facts. They are told that they are to reach their verdicts only on the basis of the evidence presented in court and must not be influenced by feelings of prejudice against, or sympathy for, the accused. This is reinforced by the jury oath, in which jurors undertake to give a verdict according to the evidence. Where there is no rational explanation for verdicts of guilty and not guilty, the court should be reluctant to speculate that the members of the jury have ignored their instructions and exercised leniency.
In the alternative, Mr Lithgow sought to confine the effect of R v H by noting that it was an unusual case given the legislative change that led to the amendment of the indictment to produce two rape counts where previously there had been one. He argued that where inconsistent verdicts indicate that jurors have not followed their oaths, the verdicts should be treated as presumptively unreasonable. If there was an obvious situation of unfairness or injustice, sufficient to explain the breaching of the oath to achieve some other objective of the justice system, the court was entitled to conclude that the verdicts were reasonable. Mr Lithgow argued:
[O]nly a special and obvious explanation for an inconsistency not based on the evidence can properly be described as a valid reason. If speculation is required then the verdicts remain unreasonable.
Mr Lithgow also argued that there was a distinct possibility that the jury had reached a compromise verdict in this case. He noted that the jury had been given the standard Papadopoulos direction, which includes the following:
One of the strengths of the jury system is that each member takes into the jury room his or her individual experience and wisdom and is expected to judge the evidence fairly and impartially in that light. You are expected to pool your views of the evidence and you have a duty to listen carefully to one another. Remember that a view honestly held can equally be honestly changed. So within the oath there is scope for discussion, argument and give and take. This is often the way in which in the end unanimous agreement is reached.
Mr Lithgow submitted that the use of the words “give and take” might have encouraged the jury to reach a compromise.
We agree with Mr Lithgow’s contention that R v H is an unusual case. We also accept that when considering inconsistent verdicts lacking any apparent logical explanation, an appellate court should be slow to conclude that the jury were being merciful in acquitting. But that does not mean that a court should never reach that conclusion. In this context, we note that in its work on juries in criminal trials, the Law Commission identified juries as fulfilling four functions, two of which were acting as the community conscience in criminal cases and providing a check on the oppressive exercise of state power. In an interesting account of jury leniency in the United States (often referred to as “jury nullification”), Alan Scheflin writes:
Proper understanding of the concept of jury nullification requires it to be viewed as an exercise of discretion in the administration of law and justice. Jury discretion in this context may be a useful check on prosecutorial indiscretion. No system of law can withstand the full application of its principles untempered by considerations of justice, fairness and mercy. Every technical violation of law cannot be punished by a court structure that attempts to be just. As prosecutorial discretion weeds out many of these marginal cases, jury discretion hopefully weeds out the rest.
Potentially, however, jury leniency undermines the rule of law, creates uncertainty and operates unequally as between comparable defendants. Importantly for present purposes, jury leniency is in direct conflict with the premises on which jury trials are conducted, as reflected in trial judges’ instructions to juries. But while jury leniency may not be encouraged – indeed, it is actively discouraged in jury instructions – that does not necessarily mean that its existence must always be ignored. There is widespread acknowledgement that juries do sometimes apply their innate sense of justice by convicting a defendant on one count and acquitting on another, even though the evidence would support convictions on both, and a general (albeit not universal) acceptance that the fact that this sometimes happens is, on balance, a beneficial feature of the jury system (the jury acting as the conscience of the community). Where an appellate court considers that a jury’s “not guilty” verdict is explicable on this basis, it seems perverse that the court should be required to quash the conviction because it is not logically consistent with the acquittal. While logic in the law is important, it is not everything.
As we have already seen, both the High Court of Australia in MacKenzie, and later in MFA, and the Supreme Court of the United States in Powell, acknowledged that juries may exercise leniency in reaching verdicts and accepted that this is within the jury’s proper function. In MacKenzie, Gaudron, Gummow and Kirby JJ quoted the following extract from the judgment of King CJ in R v Kirkman, describing the Chief Justice’s remarks as “practical and sensible”:
[J]uries cannot always be expected to act in accordance with strictly logical considerations and in accordance with the strict principles of the law which are explained to them, and courts, I think, must be very cautious about setting aside verdicts which are adequately supported by the evidence simply because a judge might find it difficult to reconcile them with the verdicts which had been reached by the jury with respect to other charges. Sometimes juries apply in favour of an accused what might be described as their innate sense of fairness and justice in place of the strict principles of law. Sometimes it appears to a jury that although a number of counts have been alleged against an accused person, and have been technically proved, justice is sufficiently met by convicting him of less than the full number. This may not be logically justifiable in the eyes of a judge, but I think it would be idle to close our eyes to the fact that it is part and parcel of the system of administration of justice by juries. Appellate courts therefore should not be too ready to jump to the conclusion that because a verdict of guilty cannot be reconciled as a matter of strict logic with a verdict of not guilty with respect to another count, the jury acted unreasonably in arriving at the verdict of guilty.
The Court in R v H referred to several United Kingdom authorities which did not accept that there was any rule that simply because a jury returned verdicts which were on the face of it inconsistent, an appellate court was obliged to quash the conviction. A recent example is the decision of the English Court of Appeal in R v Lewis. The case is principally concerned with joint enterprise liability but there was a relevant inconsistent verdict issue in relation to one of the appellants, Mr Ward. The case arose from the appellants’ attack on four victims, one of whom died (V1) and two of whom (V2 and V3) suffered injury. Mr Ward had fought with V2, but did not physically assault V1or V3. He was charged with murder in relation to V1 but convicted of manslaughter. He was also convicted on one count of attempting to cause grievous bodily harm with intent to V2 and a further count of attempt to cause grievous bodily harm in relation to V3. The Court considered that Mr Ward’s conviction for manslaughter was legally inconsistent with his conviction on the other two counts but upheld the convictions. The Court said:
In the present case the judge made it very clear to the jury, both in relation to count 1 [the murder count] and again in relation to count 5 [(attempting to cause grievous bodily harm with intent to V3)], that they could not convict Ward of either offence unless they were satisfied that he was a party to the joint enterprise from the outset. We think that there can be no doubt, therefore, that in convicting him on both those counts they were satisfied of that fact. Moreover, in convicting him on count 4 (attempting to cause grievous bodily harm with intent to [V2]) it is clear that they were satisfied that when he took part in the violence he intended to cause really serious harm. In those circumstances we find it impossible to believe that the jury was not satisfied that he had foreseen that Lewis and Cook might act in a similar way. If that is so, Ward’s conviction on count 5 is safe. The anomaly is the conviction for manslaughter on count 1; on that basis the jury should have convicted him of murder. In our view the explanation offered by [counsel] is clearly the more plausible: the jury was unwilling to convict Ward of murder in circumstances where he had played no direct part in the death of [V1]. He may be fortunate in having been convicted of manslaughter rather than murder, but we are quite satisfied that the conviction is not unsafe.
The Supreme Court of Canada has not addressed jury leniency in an inconsistent verdict context, which may indicate that the Court does not see it as relevant in that context. It has discussed the concept in other contexts, however. In R v Morgentaler, one of the several cases involving the well-known pro-choice advocate Dr Morgentaler, Dickson CJC was critical of defence counsel for making a submission to the jury that although they were bound to accept the judge’s instructions on the law, they had to decide whether to apply the law to the facts and had a right to say that it should not be applied. The Chief Justice said that the submission should not have been made, although he accepted the reality that juries have a de facto discretion to apply the law. The Supreme Court has subsequently held that guarding against jury leniency is a “desirable and legitimate exercise for a trial judge”.
Apart from R v H, there appears to be only one other New Zealand case where the Court of Appeal has refused to allow an appeal in respect of a multiple count indictment where it could not identify any logical reason why the jury convicted on some counts and acquitted on others. That case is R v Keeley, where the Court of Appeal dismissed an inconsistent verdict appeal and said:
Juries do not act with complete harmony, or complete logic, in arriving at their verdicts, and in this case there was .... ample evidence on which the jury could have convicted on all the counts. They have chosen to convict on some only. That does not allow of the argument that their acquittal on some counts is inconsistent, within the meaning of the authorities in which that term is employed in the criminal law, with a conviction on other counts.
Judges instruct juries that their task is to decide the facts and the judge’s task is to state the law. The practice of providing juries with question trails or issues sheets reinforces that role allocation and may serve to reduce the instances of jury leniency. But juries will occasionally depart from their instructions and exercise leniency, and this may result in inconsistent verdicts. Where an appellate court concludes that this has occurred, we think it legitimate, in line with the Australian and United Kingdom authorities, that the court uphold the conviction.
We agree with the Supreme Court of Canada in Pittiman that the decisive question in a case such as the present is not whether the acquittal was reasonable but whether the conviction was unreasonable. This reflects the terms of the ground of appeal (in s 385(1)(a) of the Crimes Act and ss 232(2)(a) and 240 of the Criminal Procedure Act) and requires the appellate court to examine the evidence before the jury. If the court finds it difficult to understand on what basis the jury accepted a complainant’s evidence at one point and not another, it is entitled to consider whether the jury may have departed from its instructions in giving a not guilty verdict, out of an innate sense of justice.
However, this is an explanation that an appellate court would accept only rarely. It is likely to do so only in cases of factual inconsistency arising on a multiple count indictment involving both acquittal(s) and conviction(s) in respect of the same defendant. Even in such a case, it may not provide an adequate explanation of an apparently illogical inconsistency. Whether it does or not will depend on the particular factual circumstances. The court will, of course, need to be satisfied that the verdict was not simply the result of an improper compromise. We acknowledge that making these assessments will be difficult, given that the court will have no direct knowledge of the jury’s thought processes, but where the court is in doubt (as in Dhillon), it should allow the appeal. In the present case, we have said that we consider that there was a basis on the evidence for the jury’s differing verdicts. However, we also accept that this is one of those relatively rare cases where the jury might have thought that a conviction on the rape count sufficiently captured the appellant’s culpability for what was in substance a single sexual interaction of relatively brief duration.
The Papadopoulos direction
Although it did not feature significantly in argument, Mr Lithgow did submit that the reference to “give and take” in the standard Papadopoulos direction might mislead a jury into thinking that a compromise verdict was legitimate. We do not agree, given that the passage upon which he relied (quoted at  above) was immediately followed by the following truction:137
But, of course, no one should be false to his or her oath. No one should give in merely for the sake of agreement or to avoid inconvenience. If in the end you honestly cannot agree, after trying to look at the case calmly, objectively and weighing carefully the opinions of others, you must say so. If, regrettably, that is the final position, you will be discharged and in all probability there will have to be a new trial before another jury.
That instruction makes it clear that the ultimate responsibility of every juror is to give a verdict in accordance with his or her view of the evidence.
The “give and take” referred to relates to the exchange of views or opinions. We think it would assist juries if this were made clear, as is done in Canada. Accordingly, the penultimate sentence of the extract quoted at  above should be amended by deleting the words “give and take” and adding the following italicised words:
So within the oath there is scope for discussion, argument and listening with an open mind to the opinions of others.
The appeal is dismissed.
William Young J
A preliminary comment
I disagree with the majority as to the application of s 44(1) of the Evidence Act 2006 and the significance they tentatively place on s 40(3)(b). Otherwise, I agree with their reasons and am of the view that the appeal should be dismissed. In these brief reasons, I will address the excluded evidence and discuss the inconsistency of verdicts issue.
The excluded evidence
The statutory text
Section 44 provides:
If s 44(2) is engaged, the reputation evidence in question is inadmissible and any inquiry into relevance is therefore precluded. Under s 44(1) and (3), evidence which is relevant but does not meet the heightened relevance standard under s 44(3) is inadmissible. The policies primarily underlying s 44 are that those who allege sexual offending should not be subject to humiliating cross-examination and that trials for sexual offences should not be derailed by collateral inquiries of little or no actual relevance into the complainant’s sexual experiences. These policies are fundamental to the interpretation of s 44. But so too are rights of fair trial. Given the New Zealand Bill of Rights Act 1990 and particularly ss 25(e) (the right to present a defence) and 25(f) (the right to cross-examine prosecution witnesses) along with s 6 (interpretation consistent with Bill of Rights to be preferred), I think that caution should be exercised before resolving ambiguities in the language of s 44 against defendants.
In R v A (No 2), the House of Lords utilised the Human Rights Act 1998 – s 3 of which corresponds to s 6 of the New Zealand Bill of Rights Act, albeit that it is expressed in firmer language – to interpret s 41 of the Youth Justice and Criminal Evidence Act 1999 – corresponding in function although not language to s 44 of the Evidence Act – so as to allow the admissibility with leave of evidence which was inadmissible on the language used in s 41. The interpretative approach taken in R v A (No 2) was radical, to say the least, as its effect was to create an exception to the prohibitions in s 41 which was (a) not provided for in the text of the section and (b) therefore inconsistent with that text. The result was that s 41 was distinctly read down. My reliance on s 6 of the New Zealand Bill of Rights Act in this case is far more modest. I regard the policies underlying s 44 as reflected in its language and s 6 as relevant, at least for present purposes, to whether the policies underlying s 44 warrant the section being read up.
I see two relevant ambiguities: one as to the scope of the s 44(2) prohibition against the admission of sexual reputation evidence and the other as to what is encompassed by the phrase “sexual experience .... with any person other than the defendant” in s 44(1).
Under s 41(1) and (3), evidence relating to the sexual experience of a complainant is apparently admissible if: (a) it was with the defendant (and is relevant) and otherwise (b) with the leave of the judge. But if such evidence also “relates directly or indirectly to the reputation of the complainant in sexual matters” – as may be the case – is it thereby rendered inadmissible?
I think that the better interpretation is one which restricts the operation of s 44(2) to evidence which is: (a) only potentially relevant because it relates to a complainant’s reputation in sexual matters and (b) confined to the reputation rather than any relevant behaviour underlying, or giving rise to, such reputation. Admissibility of evidence about the underlying behaviour of the complainant can then be addressed by s 44(1) and (3). I would therefore construe the section as proceeding on the basis that evidence about behaviour may be admissible but evidence confined to reputation is not.
Given the limited role which I ascribe to s 44(2), I think that the Courts below were wrong to rely on it and I agree with the approach of the majority in this respect. It likewise follows that I disagree with the approach favoured by the Chief Justice. On her approach, any evidence as to sexual character and attitudes, and thus sexual disposition, engages s 44(2) and is therefore inadmissible. There are a number of reasons why I think that this approach is wrong:
I see “reputation” in s 44(2) as denoting the way in which a complainant is regarded by others. On the approach of the Chief Justice it means that person’s actual disposition. This is not an orthodox meaning of “reputation”.
It contains an unnecessary step. The suggestion is that the appellant in this case wished to call evidence thought to indicate propensity and thereby attempted to set up reputation from which propensity could be inferred. But all the appellant sought to do was call evidence from which propensity could be inferred. Reputation, as such, had no role to play in the proposed exercise.
If the purpose of the legislature was to prohibit evidence of sexual disposition, it could have done so far more directly by using the word “disposition” as it did in s 23A of the Evidence Act 1908.
I find it difficult to see how s 44(1) and (3) would operate because evidence which would otherwise be admissible would, under those subsections would, on the approach to reputation taken by the Chief Justice, almost always be inadmissible as also going to reputation.
On the second ambiguity, s 44(1) can only apply in the present case if the courts construe the subsection on the basis that conduct with possible sexual overtones amounts to “sexual experience”. There is also the related problem, which does not arise in this case, but would on the example given by the majority at , namely, whether the “any person other than the defendant” must be someone other than the complainant. Given ss 6 and 25(e) and (f) of the New Zealand Bill of Rights Act, I favour an ordinary meaning of “sexual experience” rather than, what I think is the expanded meaning adopted by the majority. I therefore disagree with what is said in  of the reasons of the majority, albeit for reasons which line up with what is said in . This is because I do not consider that a woman who invites a man to her house when wearing a nightie and a dressing gown thereby has a sexual experience with that man; this irrespective of what he thinks she may have had in mind.
The majority discuss s 40(3)(b) at . This provides that propensity evidence about a complainant in a sexual case in relation to that person’s sexual experience may only be offered in accordance with s 44 of the Act. They consider it arguable that “sexual experience” in this subsection is not confined to “sexual experience .... with any person”. They also see as arguable a reading up of the expression “sexual experience” as encompassing non-experiences, such as for instance those recorded in diary fantasy entries. On this basis, they seem to favour the view that such diary entries would never be admissible – and that this would be so no matter how compellingly relevant they may be in a particular case. I disagree with this approach. I see “sexual experience” in s 40(3)(b) as simply a short-hand for sexual experience with another person – that is in the sense in which the expression is used in s 44. I also think that the word “experience” should be restricted to things that have happened, rather than encompass things that have not. Most importantly, I consider that we should not be reading up the language used by the legislature so as to preclude defendants calling evidence which may in fact be very relevant to their fair trial rights.
For these reasons I am of the view that s 44 was not engaged. I turn to consider whether the evidence was relevant under s 7(3) of the Evidence Act.
Was the excluded evidence relevant?
There are three bases on which it has been suggested that the evidence was relevant:
It lent support to the appellant’s contention that she had used the mouse problem as a pretext for inviting the appellant around for sex. This was the relevance suggested by defence counsel at trial.
Her aversion to mice was such that she was prepared to invite people around to deal with mice problems as and when they arose. This is the theory which was primarily advanced – at least ostensibly – on appeal. I say “ostensibly” because, as noted by the majority, counsel for the appellant used the word “pretext” in the course of argument before us. And:
Her aversion to mice was such that she was prepared to invite men around even though dressed only in a nightie and dressing gown. This theory is addressed in  of the reasons of the majority although it was not formally advanced by counsel. It is a development of the second basis.
For the reasons explained, the first basis does not directly engage s 44. It does, however, come very close. For instance, if it was alleged that on the other occasion the complainant had had sex with the other person, s 44(1) would obviously have applied. So there is something of a conundrum. As well, there can be no doubt that the policy behind s 44 was engaged. Away from the florid forensic environment of a hard-fought criminal trial, the mouse pretext theory seems ludicrous and perhaps risible. This, however, would not have detracted from the offensiveness of the proposed cross-examination from the point of view of the complainant. So although of the view that the heightened relevance standard provided by s 44(3) does not apply (given that s 44 was not engaged), I consider that the relevancy claim warranted heightened scrutiny – scrutiny which it cannot sustain:
The complainant’s expressed aversion to mice was not a pretext. It was common ground that she was averse to mice and sought assistance from others when confronted with mice problems. On the night of the offending, her initial response to the discovery of the dead mouse was to ask her daughter to dispose of it. When the appellant arrived, he was asked to dispose of the dead mouse, which he did.
The coincidences relied on by the appellant – a mouse problem and a call to the bar to resolve it – do not have any plausible connection to complainant’s alleged interest in having sex (in respect of the first occasion) or alleged consent to sex (in respect of the second). And:
Generally and most importantly, the complainant’s supposed interest in having sex on the other occasion cannot logically provide any support for the theory that she consented to have sex with the appellant on the night in question.
The second and third theories are closely related and can be discussed together. The argument for admissibility on the basis of these theories would have been far stronger if the complainant, when denying that she had invited the appellant to her house, had asserted that this was something that she would never have done. As it turned out, her denials were not embroidered by such detail. Indeed it was perfectly clear that she was averse to mice, and had on other occasions asked people to resolve mice problems. As well, there was no suggestion in her evidence of particular discomfort when the appellant arrived. On the other hand, she did tell the jury what happened after she got home (before the appellant arrived) which included getting into her night attire. It is at least possible that jurors may have thought that this adversely affected the appellant’s claim that he had been invited to the house. I would therefore be prepared to accept that the evidence – in pared down form – would have been relevant, albeit only marginally so. It was therefore admissible subject to the question whether its distinctly limited probative value was outweighed by the risk that its admission would “needlessly prolong the proceeding” under s 8(1)(b).
In his submissions to us, counsel for the appellant hypothesised two possibilities:
the complainant being questioned if she had on another occasion asked a man around to help with a mouse and her accepting the truth of the proposition, which, as he put it, “could be the end of it”; and
the complainant, in response, denying the earlier incident thus leaving it open to the defendant to call the evidence to contradict her, and opening up a line of argument based on her clinging “to denials even of the demonstrably untrue .... [as an] indicator of her truthfulness and presentation”.
Subject to a lingering concern about s 8(1)(b), I am inclined to think that the case should be looked at on the basis that the question referred to in (a) could have been properly asked. This is because the evidence, if given, would have been relevant on the second and third bases mentioned above. As well, if the question was put without tendentious detail, it could have been quickly asked and answered. But if the question had been answered affirmatively, the pool of evidence before the jury would have been added to only inconsequentially. As well there is the reality that counsel for the appellant at trial did not propose so limited an exchange. So I do not see any significant prejudice to the appellant associated with the fact that the postulated exchange was not permitted to occur.
Substantial forensic advantage for the appellant would only have accrued if: (a) there had been a denial by the complainant and (b) the appellant had been permitted to call rebutting evidence, particularly if that rebutting evidence was allowed to encompass suggestions of pretext and sexual inclination. This would have enabled the defence to open up another front as to what had happened on the other occasion and in this way would have transferred a “she says, he says” case into a “she says, they say” case.
I am by no means sure that a denial by the complainant would have warranted the calling of any rebuttal evidence; this given s 8(1)(b). Once there was a denial, the probative value of a collateral inquiry into what had happened on the other occasion would have been counterbalanced by the inconvenience of the exercise in needlessly prolonging the proceedings. And, in any event, if allowed, rebuttal evidence would have properly been confined to whether she had asked for assistance in the circumstances alleged on the other occasion. It would have been entirely inappropriate to allow a full-scale attack on the complainant’s veracity of the kind envisaged in the submissions, an attack which, on my appreciation, would have been inconsistent with the veracity rules provided for in ss 36–39 of the Evidence Act. In particular, such evidence would not have satisfied the substantial helpfulness test in s 37(1).
For these reasons, the exclusion of the evidence as to the earlier occasion in which the complainant sought assistance with a mouse problem and the refusal to permit the proposed cross-examination did not give rise to a miscarriage of justice.
The appellant’s argument on this ground has some apparent merit. The events giving rise to the counts alleging sexual violation and rape occurred in quick succession. On the complainant’s evidence and thus the Crown case, there was a single sequence of sexual activities to which she did not consent and the appellant could not reasonably have believed otherwise. There was no evidence external to that provided by the complainant which made the Crown case on the rape count stronger than on the sexual violation count. It is difficult to postulate a plausible narrative of events in which the appellant was not guilty on the sexual violation count (because either the complainant consented or he reasonably believed that she did) but was guilty on the rape count.
It is important, however, to consider the issue from the view point of the jury. The offences alleged against the appellant occurred in the course of a single but sequential course of events. My experience is that the later in the sequence a particular alleged offence, the easier it is for a jury to draw state of mind of mind inferences adverse to the defendant. This is illustrated by the present case as all the circumstantial detail which the complainant provided was available in relation to the rape count but, arguably anyway, this was not necessarily so in respect to the sexual violation count; this given the difficulty with the precise timing of the complainant’s specific consent-negativing actions in relation to the particular actions of the appellant. While I personally would be inclined to dismiss as make-weights arguments about these timing difficulties, that is not necessarily the way they would have been perceived by jurors who were required to make a collective decision and who had been told to treat the two counts separately.
We know that the jury had difficulty reaching their verdicts. They may have found it difficult to agree unanimously or by an appropriate majority on the first count, perhaps because of the timing difficulties referred to. If so, a sensible course of action would have been to move on to the second count, in respect of which these timing issues did not arise. With a verdict of guilty on the second count arrived at, jurors may have been prepared to give the appellant the benefit of the doubt on the first count given the timing difficulties just mentioned. I do not see a reasoning approach along those lines as irrational.
 B (CA862/2011) v R  NZCA 602 (Wild, Chisholm and Courtney JJ) [B v R (CA)].
 R v B DC Kaikohe CRI-2010-029-285, 20 September 2011 at  [R v B evidentiary ruling].
 At .
 At –.
 At .
 B v R (CA), above n 1, at .
 At .
 See below at  of William Young J’s judgment.
 See below at  of the majority judgment.
 Section 23A(3).
 Law Commission Evidence: Volume 2 – Evidence Code and Commentary (NZLC R55, 1999) at 124–127.
 Evidence Bill 2005 (256-2) (select committee report) at 7.
 R v Keeley  NZLR 565 (CA) at 567.
 R v Durante  1 WLR 1612 (CA) at 1617, citing with approval Devlin J in R v Stone (13 December 1954, unreported).
 R v O (No 2)  1 NZLR 326 (CA) at 333.
 R v Irvine  1 NZLR 96 (CA) at 99.
 R v Shipton  2 NZLR 218 (CA) at .
 R v Pittiman 2006 SCC 9,  1 SCR 381 at 387.
 MacKenzie v R  HCA 35, (1996) 190 CLR 348 at 367 per Gaudron, Gummow and Kirby JJ.
 MFA v R  HCA 53, (2002) 213 CLR 606 at  per Gleeson CJ, Hayne and Callinan JJ and at  per McHugh, Gummow and Kirby JJ.
 R v H  2 NZLR 581 (CA) at .
 At .
 B v R (CA), above n 1, at .
 At .
 See below at  of the majority judgment.
 R v Lewis  EWCA Crim 496,  Crim LR 870.
 R v Dhillon  EWCA Crim 1577,  2 Cr App R 10.
 See below at  of the majority judgment.
 The direction in its current form is found in R v Accused (CA87/88)  2 NZLR 46 (CA) at 59. It was originally approved in R v Papadopoulos  1 NZLR 621 (CA).
 R v B DC Kaikohe CRI-2010-029-285, 20 September 2011 at  [B v R (evidentiary ruling)].
 R v B DC Whangarei CRI-2010-029-285, 16 December 2011.
 B(CA862/2011) v R  NZCA 602 [B v R (CA)].
 B (SC 12/2013) v R  NZSC 37.
 R v Papadopoulos  1 NZLR 621 (CA). In Hastie v R  NZSC 58,  1 NZLR 297 this Court said that it would be rare for a Papadopoulos direction to be given before a majority verdict direction is given: at . Here, the Judge advised the jury of the possibility of a majority verdict in his summing up, but said he would return to that topic if necessary. It is not clear from the material before us whether any further majority verdict direction preceded or followed the Papadopoulos direction.
 R v Accused (CA87/88)  2 NZLR 46 (CA) at 59. Cooke P delivered the unanimous judgment of a seven member Bench.
 B v R (CA), above n 32.
 By virtue of s 397 of the Criminal Procedure Act 2011, s 385 of the Crimes Act 1961 continues to apply to this appeal.
 B v R (evidentiary ruling), above n 30, at .
 At .
 At .
 At .
 At .
 B v R (CA), above n 32, at .
 At .
 Evidence Amendment Act 1977, s 2. For discussion of s 44, see Law Commission The 2013 Review of the Evidence Act 2006 (NZLC R17, 2013) ch 7; Richard Mahoney and others The Evidence Act 2006: Act and Analysis (2nd ed, Brookers Ltd, Wellington, 2010) at [EV44.01]–[EV44.02]; Elisabeth McDonald and Yvette Tinsley (eds) From “Real Rape” to Real Justice: Prosecuting Rape in New Zealand (Victoria University Press, Wellington, 2011) at 325–336.
 Bull v R  HCA 24, (2000) 201 CLR 443 at  per McHugh, Gummow and Hayne JJ (Gleeson CJ and Kirby delivered separate judgments).
 New Zealand Bill of Rights Act 1990, s 25(a) and (e).
 See R v Bourke CA207/06, 15 August 2006 at – (a case decided under s 23A of the Evidence Act 1908).
 R v Seaboyer  2 SCR 577 at 624–625 per McLachlin J. See also Neil Kibble “Judicial Discretion and the Admissibility of Prior Sexual History Evidence under s 41 of the Youth Justice and Criminal Evidence Act 1999: Sometimes sticking to your guns means shooting yourself in the foot”  Crim L R 263.
 Evidence Act 1908 (WA), ss 36B (sexual reputation evidence prohibited), 36BA (sexual disposition evidence prohibited), and 36BC (sexual experience evidence admissible with leave of judge); discussed in Bull v R, above n 46.
 As is recognised by s 40(3) of the Evidence Act 2006, which provides that, in a sexual case, propensity evidence about a complainant’s sexual experience may be offered only in accordance with s 44.
 See Australian Law Reform Commission and New South Wales Law Reform Commission Family Violence – Improving Legal Frameworks: Consultation Paper (ALRC CP 1 and NSWLRC CP 9, 2010) at [18.14]–[18.23].
 See – above.
 Tony Deverson and Graeme Kennedy (eds) The New Zealand Oxford Dictionary (Oxford University Press, Melbourne, 2005) at 892.
 B v R (CA), above n 32, –.
 See Richard Mahoney and others The Evidence Act 2006: Act and Analysis (2nd ed, Brookers, Wellington, 2010) at [EV44.01].
 B v R (CA), above n 32, at  and .
 At .
 R v Owen  NZSC 102,  2 NZLR 37 at .
 MacKenzie v R (1996) 190 CLR 348 at 365 per Gaudron, Gummow and Kirby JJ; Dawson and Toohey JJ delivered a separate judgment but agreed with the majority’s reasoning on the inconsistent verdict ground: see 351.
 New Zealand Bill of Rights Act, s 24(e).
 Crimes Act, s 117(b).
 It is a contempt of court to disclose details about jury deliberations: see Solicitor-General v Radio New Zealand  1 NZLR 48 (HC) at 53–55; and John Burrows and Ursula Cheer Media Law in New Zealand (6th ed, LexisNexis, Wellington, 2010) at [9.11].
 Jurors must swear, declare or affirm that they will try the case before them to the best of their ability and to give their verdict according to the evidence: Jury Amendment Rules 2000, sch 1, form 2.
 See generally Eric L Muller “The Hobgoblin of Little Minds? Our Foolish Law of Inconsistent Verdicts” (1998) 111 Harv L Rev 771.
 See, for example, R v Keeley  NZLR 565 (CA); R v Glassey  NZLR 1115 (CA); R v Irvine  1 NZLR 96 (CA), R v Irving CA234/87, 4 March 1988, R v K(CA49/96) CA49/96, 13 August 1996; R v O (No 2)  1 NZLR 326 (CA); R v H  2 NZLR 581 (CA); R v Shipton  2 NZLR 218 (CA); and Dempsey v R  NZCA 297.
 MacKenzie, above n 60, at 366–368.
 R v Pittiman 2006 SCC 9,  1 SCR 381 at . This was a multiple accused case.
 R v Irvine, above n 66, at 99.
 R v O (No 2), above n 66, at 333.
 At 333.
 MacKenzie, above n 60, at 367. The majority endorsed an extract of the judgment of King CJ (with Olsson and O’Loughlin JJ concurring) in R v Kirkman (1987) 44 SASR 591 (SACA) at 593, which we quote at  below.
 R v H, above n 66, at –.
 MacKenzie, above n 60, at 368.
 R v Pittiman, above n 68, at .
 See also R v Pittiman, above n 68, at .
 MFA v R  HCA 53, (2002) 213 CLR 606 [MFA] at – per Gleeson CJ, Hayne and Callinan JJ and at – per McHugh, Gummow and Kirby JJ.
 R v Pittiman, above n 68, at .
 Dunn v United States  USSC 15; 284 US 390 (1932).
 United States v Powell  USSC 237; 469 US 57 (1984).
 At 63, citing Harris v Rivera  USSC 230; 454 US 339 (1981) at 345–346. It should be noted, however, that the Court expressly reserved the question of the proper resolution of a case where a defendant is convicted of two crimes but the conviction on one logically excludes a finding of guilt on the other: at 69, fn 8. Presumably the Court was referring to situations of legal inconsistency.
 At 65.
 At 65.
 At 65.
 At 65–66.
 At 66.
 At 67.
 See Criminal Procedure Act, s 147. Section 147 permits the court to dismiss a charge if: (a) the prosecutor has not offered evidence at trial; (b) in a judge-alone trial, the court is satisfied that there is no case to answer, and (c) in a jury trial, the judge is satisfied that as a matter of law, a properly directed jury could not reasonably convict the defendant.
 See Criminal Procedure Act, ss 232(a) and 240.
 See, for example, Muller, above n 65.
 As the Supreme Court’s decision in Powell was not constitutionally based, state courts were free to take a different view in relation to state offences: see United States v Powell, above n 80, at 65. For discussion of the current position in various states, see the decision of the Maryland Court of Appeals in McNeal v State 44 A 3d 982 (Md 2012) at 989–990.
 See above at  above.
 This is subject to the possible exception identified at n 81 above.
 McNeal v State, above n 91, at 989–990.
 At 992–993.
 MFA, above n 77, at  per Gleeson CJ, Hayne and Callinan JJ.
 B v R (CA), above n 32, at .
 At .
 R v Shipton, above n 66, at .
 MFA, above n 77, at .
 R v Dhillon  EWCA Crim 1577,  2 Cr App R 10 (CA).
 At .
 At , citing R v Cross  EWCA Crim 1553 at .
 At .
 At .
 At .
 At .
 At .
 At .
 At .
 At .
 At .
 At .
 At .
 At .
 B v R (CA), above n 32.
 R v H, above n 66, at .
 At .
 R v Accused (CA87/88), above n 35, at 59.
 Law Commission Juries in Criminal Trials Part One: A Discussion Paper (NZLC PP32, 1998) at –; and Law Commission Juries in Criminal Trials (NZLC R69, 2001) at –. See also Law Reform Commission of Canada The Jury in Criminal Trials (Working Paper 27, Ontario, 1980) at 8–13.
 Alan W Scheflin “Jury Nullification: The Right to Say No” (1972) 45 S Cal L Rev 168 at 181.
 See, for example, R v Morgentaler  1 SCR 30 at  per Dickson CJC (with whom Lamer J joined); and Andrew D Leipold “Rethinking Jury Nullification” (1996) 82 Va L Rev 253.
 R v Kirkman, above n 72, at 436.
 MacKenzie, above n 60, at 367–368 per Gaudron, Gummow and Kirby JJ.
 R v H, above n 66, at , citing R v Drury (1971) 56 Cr App R 104 (CA); and R v Durante  1 WLR 1612 (CA).
 R v Lewis  EWCA Crim 496.
 At .
 At .
 At .
 At .
 At .
 R v Morgentaler, above n 122, at 76–78. Lamer J joined with Dickson CJC.
 At 78–79.
 R v Latimer 2001 SCC 1,  1 SCR 3 at .
 R v Keeley, above n 66, at 567.
 See  above.
 R v Accused (CA87/88), above n 35, at 59.
 See R v G (RM)  3 SCR 362 at  per Cory J for the majority.
 Emphasis added.
 R v A (No 2)  UKHL 25,  1 AC 45.
 For present purposes the primary difference is that s 41 permits evidence of sexual experience between the complainant and defendant as admissible in respect of consent only in extremely limited circumstances.
 Providing the judge was satisfied that admission of the evidence was necessary for the trial to be fair.
 See for instance the discussion in IH Dennis The Law of Evidence (5th ed, Sweet & Maxwell, London, 2013) at [15-015]–[15-018].
 In saying this, I am not taking a position as to the application of s 44(2) in circumstances of the kind which arose in R v Bourke CA207/06, 15 August 2006. As to this, see for instance Elisabeth McDonald “Complainant’s Reputation in Sexual Matters”  NZLJ 251.
 See  of the majority judgment.
 See – of the Chief Justice’s judgment.
 See  of the majority judgment.
 Normally no harm is done if a defendant is able to lead evidence of limited or no relevance. It will presumably be ignored by the jury if they conclude that it is of no value. For this reason, I would normally afford the defendant the benefit of the doubt on relevance. But where, as here, other interests are engaged, doubtful relevancy claims should be carefully scrutinised.
 And thus does not meet the s 7(3) test.
 That is, without reference to the alleged sexual connotations.
 Compare with  of the majority judgment.
 The meaning is clear despite the double negative.
R M Lithgow QC and N Levy for Appellant.
M D Downs and K A L Bicknell (instructed by Crown Law Office, Wellington) for Respondent.
all rights reserved