SC 115/2015 International Cases [2017A] Part 5 Case 8 [NZSC]



Lyonel M.T.P. Taniwha

- vs -


Elias CJ

William Young J

Glazebrook J

Arnold J

OʼRegan J




The appeal is dismissed.


Justice Arnold


  1. In New Zealand, as in other jurisdictions from the common law tradition, criminal trials are conducted orally, in open court. The principle of orality, enshrined in s 83 of the Evidence Act 2006, recognises the fundamental importance of transparency in the administration of justice through the courts. The principle also rests upon the assumption that a fact-finder, whether a judge sitting alone or a jury, is likely to benefit from seeing and hearing witnesses give their evidence. There is, however, research which indicates that a person’s demeanour when giving evidence in court generally provides little or no assistance to a fact-finder charged with determining whether or not the witness is telling the truth.[1] A witness who presents as confident, articulate and honest may be mistaken or dishonest; a witness who presents as diffident, hesitant or awkward may be telling the truth and their evidence may be accurate. Not only can appearances be deceptive, but fact-finders may overestimate their ability to recognise those who are truthful from those who are not, by, for example, relying on unreliable behaviours such as fidgeting or looking away.[2] In this appeal, the Court is asked to consider how a judge instructing a jury in a criminal trial should deal with the question of demeanour.

  2. Background

  3. The appellant, Mr Taniwha, was convicted following a jury trial before Judge O’Driscoll of six of 12 counts involving physical and sexual violence against the complainant, with whom he was in a relationship for a few months. The Crown alleged that soon after the relationship began, Mr Taniwha began to act in a possessive and controlling way towards the complainant and later became abusive and violent towards her. Ultimately the police became involved, which led to the charges. Mr Taniwha denied the charges. While acknowledging that there were arguments and fights during their relationship, he said that the complainant had fabricated the allegations.

  4. It is not necessary that we set out the factual background in full. It is sufficient that we identify the two particular incidents that are relevant to the issues in the appeal. We will refer to them as the towel incident and the Police safety order incident.

  5. Before we outline these incidents, however, we should mention that prior to the trial, the Crown applied under s 103 of the Evidence Act for a direction that the complainant’s evidence be given by alternative means. Specifically, the Crown sought an order that the complainant’s evidential video be played (being the bulk of her evidence-in-chief) and that her oral evidence be given by way of closed circuit television (CCTV). After a contested hearing, these orders were made,[3] and the complainant’s evidence was given in this way at trial.

  6. The towel incident

  7. Mr Taniwha was convicted of two counts of rape arising out of events on Christmas Day 2012. Mr Taniwha and the complainant had gone to a family gathering at the home of one of Mr Taniwha’s relatives for Christmas lunch. During the afternoon, the complainant indicated that she had to go back to her house to have Christmas dinner with her flatmate. Mr Taniwha became angry. He arranged for a friend to drive the complainant and him to his house, where he demanded that the complainant have sex with him. When the complainant refused, Mr Taniwha raped her, putting a dirty towel into her mouth to stop her protestations. He raped her again shortly afterwards.

  8. When she gave evidence at trial, the complainant was shown a photograph of the bathroom at Mr Taniwha’s house. The photograph showed some towels. The prosecutor asked the complainant whether she recognised any of the towels. She said she did recognise one – it was the one which Mr Taniwha had forced into her mouth. When her attention was drawn to the towels in the photograph, the complainant apparently reacted, showing visible distress.

  9. Mr Taniwha also gave evidence. He said that the sexual intercourse that evening was consensual and denied that he had shoved a towel into the complainant’s mouth. The prosecutor asked Mr Taniwha about the complainant’s reaction when she was shown the photograph of the towels:

  10. Q.

    Did you see Mr Taniwha when [the complainant] was giving her evidence, when she looked at the photo of those pink towels .... in the bathroom? Did you see her reaction to that?




    Do you think she just put that all on for the jury?


    Yes. She smokes methamphetamine.


    So your evidence is that there was no towel and certainly not one in her mouth?


    She smokes more methamphetamine than I do.


    Yes you’ve said that, so, your evidence is that there was no towel and certainly not one in her mouth?


    There was no towel, no towel in her mouth.

    Later in the cross-examination, the prosecutor referred Mr Taniwha to some texts between the complainant and him the following day which referred to the towel incident.[4] The prosecutor asked Mr Taniwha what the complainant was talking about if there was no towel incident. Mr Taniwha replied:


    Random texts, making up stories. She does it a lot.

  11. In his closing address to the jury, the prosecutor referred to the complainant’s reaction:

  12. You might have noticed, members of the jury, when she saw those pink towels in the photograph booklet underneath the [bathroom] basin her instinctive reaction and upset. In my submission it’s a matter for you that that wasn’t made up, that wasn’t faked.

  13. In addition, the prosecutor identified five reasons why the jury should believe the complainant. Having emphasised that it was for the jury to decide whether or not they agreed with his submissions, he put the first of the five reasons in this way:

  14. First and importantly, think about the way in which [the complainant] gave her evidence, her demeanour. In my submission she was very compelling. Ask yourselves was that consistent with a person who had been through what she had been through with Mr Taniwha or with a person who for some reason is making all this up; acting as [he] suggested she was.

    Now obviously none of you are human lie detectors and you can’t just judge a case solely on how a person appears. But in my submission the way she gave her evidence was certainly very compelling. I don’t need to say anything more to you because you are best placed to assess that and factor that in to the other evidence that you have heard in deciding whether to accept her account.

    Consider the fact, members of the jury, that she was a long time in that witness box, I think around 10 hours in total, about half of which was cross-examination. In my submission she wasn’t shaken about one single material detail. Not one. Sure there were peripheral details and things such as what week it occurred, she might not have been sure about but on the material details she was unshaken. In my submission that would be very difficult to achieve if all this were a lie.

  15. In concluding the closing address, the prosecutor said:

  16. One of the great things about a jury is that between you, you have years of experience of the world, a good pool of collective common sense. Having seen [the complainant] give evidence, you are the 12 best placed to assess whether she was telling the truth, bearing in mind all of the evidence, all of the other evidence you have heard. In my submission to you, she was telling the truth and the proper verdicts are guilty. But that, of course, is your province.

  17. The complainant’s reaction when she saw the towel in the photograph and the prosecutor’s observations about that and the compelling nature of the complainant’s evidence raise the demeanour issue.

  18. Police safety order incident

  19. Towards the end of February 2013 when the complainant was at Mr Taniwha’s house, Mr Taniwha woke her and said he wanted to have sex. Having already had sexual intercourse with him that evening and gone to sleep, the complainant refused his advances. Despite this, Mr Taniwha forced her to have sex with him (the jury convicted him of rape in relation to this incident). The complainant returned to her house the following morning. Later that evening, there were numerous texts between Mr Taniwha and the complainant, in which Mr Taniwha said that he wanted to stay the night at the complainant’s place. The complainant refused, ultimately saying that she was going to bed. Mr Taniwha said that he would go over to the complainant’s house anyway and demanded that she answer the door when he arrived. He turned up at her house early the following morning (6.30 am). He was drunk and belligerent. The complainant’s flatmate called the police, who detained Mr Taniwha “for detox” and served him with a Police safety order (PSO).[5]

  20. The effect of the PSO (which was for a period of five days) was that Mr Taniwha was not to contact or harass the complainant in any way. When the police released Mr Taniwha later in the day, he immediately breached the PSO by going to the complainant’s workplace, where he asked her for money and told her that she had better go round to his place that evening.

  21. In opening the case for the Crown, the prosecutor referred to the circumstances giving rise to the PSO and then said:

    Within about 12 minutes of being released by the police from their custody he had breached [the PSO] and was at the complainant’s [workplace] confronting her. The Crown says that his actions, in going to [the complainant’s workplace] within minutes of being released and being issued that [PSO], show his strong, controlling tendencies and also his lack of appreciation of appropriate boundaries.
  22. The police officer who issued the PSO gave evidence of the circumstances by reading his brief of evidence by consent. That evidence indicated that the police had gone to the complainant’s house as the result of a complaint; that they found Mr Taniwha outside banging on the windows; that he had been drinking and was uncooperative and belligerent; and that, after having been served with the PSO and given time to sober up, he had breached the PSO immediately upon release. In addition, when referring to this incident in her evidence, the complainant said that Mr Taniwha had threatened to “sort out” her flatmate. Mr Taniwha addressed the PSO incident briefly in his evidence-in-chief and the prosecutor questioned him about it and his breach of the PSO at the outset of his cross-examination. Finally, in his closing address, the prosecutor said:

  23. You’ve also got Mr Taniwha’s response to the [PSO], breached within about 15 minutes of release and then repeatedly in the days following by texting her. That kind of action is wholly aligned with what [the complainant] has told you about Mr Taniwha’s attitude to the involvement of the police in their relationship. And that’s also borne out by the text messages that you’ve seen, one of which I just read out, “Don’t bother calling the pigs because I don’t give a fuck about them.” Together these strands of corroboration lend substantial independent support to [the complainant’s] story.

  24. Two points arise from this sequence – first, whether the probative value of the PSO evidence was outweighed by its prejudicial effect, so that it was inadmissible and second, if the evidence was admissible, should a “proper use” direction[6] have been given.

  25. The issues

  26. This Court granted Mr Taniwha leave to appeal in respect of two issues, namely whether the Court of Appeal was correct that:[7]

    1. no miscarriage of justice arose as a result of the absence of a tailored demeanour direction in the Judge’s summing up to the jury; and

    2. evidence of the appellant’s breach of a Police safety order two days after the date covered by the final count alleged in the indictment was admissible and no “proper use” direction was required.

  27. Demeanour

  28. In brief, Mr Chisnall argued for Mr Taniwha that two aspects of the complainant’s demeanour had been highlighted by the prosecutor, namely the complainant’s reaction when shown the photograph of the towels and her general demeanour when giving evidence. He submitted that this was a case which depended on whether the jury believed the complainant, Mr Taniwha having given a different account of what had occurred. Given the use that the prosecutor attempted to make of the complainant’s demeanour in closing, the jury should have been given a tailored demeanour direction, as required by E (CA799/2012) v R.[8] In the absence of such a direction, there was a risk that the jury would give unjustified weight to the complainant’s demeanour in determining her credibility.

  29. The Court of Appeal rejected this argument, holding that the reference to the towel incident and the complainant’s demeanour in that context “was a very small part of the extensive evidence the jury heard”.[9] In relation to the prosecutor’s point in closing about the complainant’s general demeanour, the Court of Appeal said that while mentioned, her demeanour was “not given any great prominence”.[10] The Court also noted that the trial Judge had:

    1. emphasised several times in the summing up “the need for the jury to have regard to all of the evidence they had heard and to do so clinically and carefully”;[11] and

    2. given a demeanour warning at the start of the trial.[12]

    The Court was satisfied that there was no real risk that the jury would have placed excessive reliance on demeanour in reaching their verdicts.[13]

  30. Before we evaluate Mr Chisnall’s submissions, we should give more detail of relevant aspects of the Judge’s instructions to the jury.

  31. The Judge’s instructions to the jury

  32. In his opening remarks to the jury, Judge O’Driscoll drew the jury’s attention to the fact that they would have to assess the credibility and reliability of witnesses. He noted that a witness may be honest but unreliable, or reliable on some points and not on others. A witness’s inability to recall a particular matter did not mean the witness’s evidence was unreliable as a whole. The Judge gave a demeanour warning as follows:

  33. How you assess the respective witnesses is entirely a matter for you. Experience has shown that simply by looking at a witness and watching their demeanour as they give their evidence is not necessarily a good way to assess their evidence. For example, a witness may not appear confident in the witness box when giving evidence or the witness may make a number of pauses or hesitate when giving evidence. That doesn’t necessarily mean that their evidence is untruthful. The witness may be understandably nervous when giving evidence, particularly in front of a number of people.

    The Judge asked the jury to keep an open mind until they had heard all the evidence.

  34. The Judge returned to these matters in his summing up. In explaining the burden of proof, the Judge made it clear to the jury that they had to give “careful and impartial consideration to all of the evidence”. He emphasised that they were to consider each count separately and should isolate and consider the evidence relevant to each count, points reiterated in the written material provided to the jury. The Judge noted that this was a case where there were “huge gulfs” between the evidence of the complainant and that of Mr Taniwha, which could not really be explained on the basis of mistake or misinterpretation.

  35. When addressing the issues of the credibility and reliability of witnesses, the Judge posed a series of questions for the jury to consider:

  36. [87]

    .... Were the witnesses giving, for example, their evidence to the best of their ability or were they attempting, do you think, to deceive you when they gave their evidence? How reliable was the witness? Were they an actual active participant in an incident or were they a mere bystander? What was the relationship of the witnesses? Were the witnesses independent or did they have a close association with each other that you thought coloured or affected their evidence? Did you observe any bias or prejudice exhibited by the witnesses? Do you think any witness embellished, minimised or exaggerated their evidence on either their role or on the role of someone else? Was the evidence consistent or inconsistent with other witnesses?


    Importantly in this case you have a raft of text messages. How do the text messages fit in with what the witnesses have said occurred or didn’t occur? Are the witnesses consistent with what they have said on a previous occasion to what they have said in Court? How do you think the witnesses performed under cross-examination? Were they worn, tired down, confused or did they stand up to cross-examination and resilient when challenged? Were they prepared to make realistic and appropriate concessions and of course in this case you have got the impact of alcohol and drugs, which you need to consider where witnesses have taken alcohol and drugs ....

  37. Later, the Judge reminded the jury of the defence argument that the complainant was not a credible witness because some of her text messages to Mr Taniwha were inconsistent with her allegations and were consistent with arguments and verbal disagreements rather than assaults and sexual abuse.

  38. Finally, towards the end of his summing up, the Judge said:

  39. [112]

    Mr Foreperson, you have seen and heard the witnesses give evidence, you have seen them being subject to cross-examination. This case fundamentally involves your assessment of the witnesses and what you make of the witnesses. Look at the evidence clinically and carefully, don’t speculate or guess. Use your experiences of life, of people, of common sense in considering the weight that you attach to the evidence from the various witnesses. Look at all the evidence, including the text messages ....


  40. In their discussion paper entitled Evidence Law: Character and Credibility,[14] the Law Commission said:[15]

  41. The demeanour displayed by witnesses and the manner in which they present their testimony are traditionally regarded as relevant to assessing truthfulness. This has contributed to the reluctance of appellate courts to interfere with first instance findings of fact based on a determination of truthfulness. However, a determination of truthfulness by reference to demeanour has a subjective basis which will inevitably reflect the values, experience and cultural norms of the fact-finder. The danger of assessing truthfulness by reference to demeanour alone was well expressed by the Privy Council in AG of Hong Kong v Wong Muk Ping [1987] AC 501, 510:

    It is a commonplace of judicial experience that a witness who makes a poor impression in the witness box may be found at the end of the day, when his evidence is considered in the light of all the other evidence bearing upon the issue, to have been both truthful and accurate. Conversely, the evidence of a witness who at first seemed impressive and reliable may at the end of the day have to be rejected. Such experience suggests that it is dangerous to assess the credibility of the evidence given by any witness in isolation from other evidence in the case which is capable of throwing light on its reliability ....

  42. The Commission went on to point out that psychological research indicated that popular beliefs about what types of behaviour indicated that a speaker was being deceptive were erroneous, and that such misconceptions were particularly dangerous in the context of cross-cultural assessments.[16] An example is gaze. There is a widespread belief that liars look away from those they are trying to deceive; yet this has been shown to be an unreliable indicator of lying.[17]

  43. It is perhaps useful to clarify what demeanour means in this context. Lord Bingham described demeanour as follows:[18]

  44. [It is the witness’s] conduct, manner, bearing, behaviour, delivery, inflexion; in short, anything which characterizes [the witness’s] mode of giving evidence but does not appear in the transcript of what he [or she] actually said.

    In a recent article on which Mr Chisnall placed particular reliance, Hon Robert Fisher QC gave the following description:[19]

    The words used (the verbal content) can be contrasted with the way in which they were uttered (the non-verbal content). A speaker simultaneously manifests a package of non-verbal information which is normally subconscious. It consists of facial expressions, bodily movements and vocal characteristics. The vocal characteristics (known as “paralinguistic cues”) include pitch, pace, volume, timbre, expression and tremors. It is this package of face, body and voice that lawyers describe as “demeanour”.

  45. Demeanour is, then, a broad term. As a result, generalisations may be misleading. Sometimes, it will not be possible to understand what a witness meant simply from the words recorded in a transcript – non-verbal conduct falling within the description of “demeanour” may be vital in conveying meaning. Assume, for example, that a proposition is put to a witness under cross-examination that some years earlier he had acted in a particular way. The witness’s reply is recorded as “No, I don’t think so”. Depending on the way in which that response is delivered, the witness may be expressing uncertainty – “No, I don’t think I did that, but I’m not sure” – or may be rejecting the proposition out of hand – “No, I most certainly did not do that”.

  46. The Court of Appeal made this point in R v Munro.[20] Delivering the judgment of herself, Chambers, Arnold and Wilson JJ, Glazebrook J said:[21]

  47. Tone of voice is important in conveying meaning and this cannot be captured in a written transcript. For example, the concession “it’s possible” could, depending on the tone, have a meaning ranging from “that’s quite likely” to “yes, but pigs might fly”. Pauses can be significant and are rarely shown in a transcript. Gestures and facial expressions can also add to meaning and are often not recorded in a transcript unless vital to a witness’ answer and even then, when digital recording systems are in operation, only when specifically read into the record by counsel or the judge.

  48. Glazebrook J then went on to discuss some of the authorities and research which question the utility of demeanour in assessing witness credibility, noting in particular that:[22]

    1. Giving evidence in a courtroom setting is an artificial process so that the ability of witnesses to tell their stories in their own way is impeded.

    2. Commonly accepted behavioural clues as to when a person is lying (such as hesitation or eye or body movements thought to indicate shiftiness) have been shown to be inaccurate.

    3. Appearance influences credibility assessments. People perceived as confident, open and attractive are more likely to be believed whereas people perceived as unattractive or unsavoury are less likely to be believed. This may lead to an undiscriminating assessment of a witness’s evidence, in the sense that evidence is believed, or disbelieved, in its entirety – the so-called “halo effect”.

    Glazebrook J also noted that everyone is prone to “heuristic illusions”, which can lead to systematic errors and biases in decision-making.[23] An example is a tendency, with hindsight, to overestimate what could have been anticipated to occur.

  49. In his article, Robert Fisher summarised some of the psychological and other studies dealing with demeanour and concluded:[24]

  50. The overwhelming conclusion is that demeanour is not a useful guide to veracity. There are no observational advantages when assessing the honesty of a witness’s evidence. Those confined to reading the transcript will do just as well. Of particular concern is the fact that so many sincerely believe that they are capable of assessing veracity through demeanour. The widespread belief that they can is a popular fallacy.

    He did accept, however, that non-verbal cues may give meaning to the words used by a witness and are important for that reason, although he doubted that this is a significant issue in many instances.[25]

  51. Mr Chisnall’s submissions were built around two key propositions. The first was that credibility findings based on demeanour are idiosyncratic and unpredictable and people do no better than chance (for example, tossing a coin) in detecting deception or dishonesty through demeanour. The second was that the law must keep pace with current knowledge and develop some means to address the risk of demeanour findings by juries who sincerely believe that that they are capable of assessing veracity through demeanour. Mr Chisnall argued that juries should be told that they must not make credibility assessments on the basis of demeanour, and should be advised of the reasons why it is unsafe to do so.

  52. As we noted at the outset, criminal trials in New Zealand are conducted by way of oral evidence given before a fact-finder (whether a judge sitting alone or a jury). This is reflected in the Evidence Act, s 83(1)(a) of which provides:

  53. 83.

    Ordinary way of giving evidence


    The ordinary way for a witness to give evidence is,―


    in a criminal or civil proceeding, orally in a courtroom in the presence of―


    the Judge or, if there is a jury, the Judge and jury; and


    the parties to the proceeding and their counsel; and


    any member of the public who wishes to be present, unless excluded by order of the Judge;

    The section goes to on address the circumstances in which evidence may be given by way of affidavit.

  54. As s 83(1)(a) indicates, the oral tradition is linked to the requirement that the administration of justice through the courts be conducted by way of an open, public process. The requirement for an open, public court process is seen as vital to the preservation of public confidence in the courts, to the extent that the common law recognised only very narrow exceptions to it.[26] More relevantly, however, the principle of orality also reflects an assumption that a fact-finder is likely to benefit from seeing and hearing a witness give evidence in person. In their 2013 review of the Evidence Act, the Law Commission expressed the view that:[27]

  55. .... the principle of orality in criminal proceedings is an important one; it allows a defendant to fully test a witness’s evidence by cross-examining them on it and provides the fact-finder with the opportunity to assess firsthand the credibility and veracity of the witness giving evidence.

    It is against this background that Mr Chisnall’s arguments must be considered.

  56. In E (CA799/2012) v R the Court of Appeal concluded that in cases where demeanour might be relevant, a warning about the risks of relying on demeanour when assessing credibility was not inevitably required, but, when a warning was required, it should be tailored to the circumstances of the particular case. To adopt a universal rule would, the Court said:[28]

  57. .... tend to deprive juries of the accepted benefits of seeing and hearing the witnesses; it would risk juries interpreting the Judge’s direction as an invitation to disbelieve the witness or to place little weight on his or her evidence; and it would restrict the ability of the trial judge to tailor a direction appropriate to the circumstances of the case.

  58. The Court went on to say, however, that the jury should be advised of two points.[29] The first was that the assessment of the credibility and reliability of a witness should be broadly based and take into account the evidence as a whole and other relevant factors. The second was that, while demeanour could properly be taken into account, it was best not considered in isolation but should be considered as one factor in a broader assessment.

  59. In relation to the factors potentially relevant to that broader assessment, the Court identified:[30]

  60. (a)

    The consistency of the witness’s evidence with what is agreed, or clearly shown by other evidence, to have occurred.


    The internal consistency of the evidence of the witness.


    Consistency with what the witness has said or deposed on other occasions.


    The credit of the witness in relation to matters not germane to the litigation.



    The inherent plausibility of the evidence of the witness (does it make sense?) and


    Where appropriate, consistency with any contemporaneous documentary evidence.

  61. In describing when, and what type of, tailored direction might be required, the Court of Appeal said:

  62. [46]

    We stress the importance of tailoring the jury direction to the circumstances of the case. That will depend to an extent on the points made by counsel in their closing addresses, the relative importance of demeanour in the particular case and the existence or otherwise of other factors which might in the circumstances assume greater importance in the overall assessment.


    For example, the demeanour of a witness may be far less important in a case where there are substantial internal inconsistencies in the evidence of the witness or with his or her prior statements.


    On the other hand, in a “she said/he said” type of case, demeanour may assume greater importance in the absence of other factors such as inconsistency or any inherent implausibility. Counsel may also have given prominence to demeanour in their addresses. In such a case, the judge may consider it appropriate to draw on the suggestions made by the Canadian Judicial Council:


    What was the witness’s manner when he or she testified? Do not jump to conclusions, however, based entirely on how a witness has testified. Looks can be deceiving. Giving evidence in a trial is not a common experience for many witnesses. People react and appear differently. Witnesses come from different backgrounds. They have different abilities, values and life experiences. There are simply too many variables to make the manner in which a witness testifies the only or most important factor in your decision.


    If the judge considers a direction along these lines is appropriate in the circumstances, then it is important to tailor this to the specifics of the case rather than applying it in a formulaic manner. One suggestion is to deal with this issue in the context of counsel’s closing addresses on this topic. This should ensure that any such direction is tied to the specifics of the case. If in doubt on this issue, judges may wish to raise with counsel before summing up whether a demeanour warning is required and, if so, what form it should take.

  63. Against the background of his submissions that the settled scientific position is that demeanour is an unreliable predictor of truthfulness and that the impact of this had yet to be completely embraced by either Parliament or the judiciary, Mr Chisnall argued that the Court of Appeal in E (CA799/2012) was wrong to accept that demeanour may afford some assistance provided it is not determinative or the dominant consideration. A fact-finder, Mr Chisnall submitted, can derive no benefit from a witness’s demeanour and this must be made clear to jurors by a direction given in all cases where credibility is at issue. He supported the direction suggested by Robert Fisher in his article.[31]

  64. In our view, Mr Chisnall’s submission that judges must sum up to juries on the basis that they can derive no benefit from demeanour in assessing truthfulness is inconsistent with important features of the legislative context. These include the oral tradition recognised in s 83(1) and the assumptions which underlie it, which this Court has previously acknowledged;[32] the fact that the Evidence Act does not require a warning to be given (as it does in other contexts); and the fact that the Law Commission has not recommended that the Act be amended in this respect.[33] Moreover, we agree with the Court of Appeal that an invariable requirement for a demeanour warning may operate unfairly in some cases, for example, where it will create a risk that what could otherwise be significant evidence will be devalued or undermined unfairly. As we will explain below, we consider that such a risk arises in the present case.

  65. We have already said that what could be described as demeanour – gestures or tone of voice, for example – may affect meaning and so cannot be ignored. In addition, where a witness gives evidence over a lengthy period, a fact-finder may be able to form a view about whether a witness is intelligent or unintelligent, well or poorly educated, clear-thinking or muddled and so on. In some circumstances, such assessments may be relevant to determining credibility and reliability. Demeanour may be relevant in other ways. For example, where a defendant facing fraud charges is generally articulate, succinct and focussed when answering questions but becomes evasive, obtuse and long-winded when asked difficult questions going to the heart of the case, the fact-finder may well be entitled to draw an adverse inference as to the reason for that. On the other hand, a marked change in response or demeanour by a complainant in a sexual case when asked about intimate details of the alleged offending may simply reflect embarrassment or the difficulty in re-living a traumatic event. As noted earlier, generalities are dangerous.

  66. Accordingly, we consider that there is no invariable requirement for judges to give demeanour warnings when summing up to juries in cases where credibility is at issue. Rather, we consider that the need for a warning should be assessed in each case. Whether a warning is required will depend upon the nature of the evidence in the case and the way the trial has unfolded. The key consideration for the Judge will be whether there is a real risk that witness demeanour will feature illegitimately in the jury’s assessment of witness veracity or reliability. We must express a note of caution, however, given the risk that a jury will interpret a “tailored” direction[34] as an expression of doubt by the judge as to the veracity of a particular witness or witnesses. Obviously, any direction should be formulated in a way that avoids this.

  67. Given the difficulty in formulating tailored directions in a summing up that avoids the risk just mentioned, we consider that trial judges could usefully, as a matter of course where credibility is likely to be a major issue at trial, include in their opening remarks to the jury a brief statement about the approach the jury should take to assessing competing accounts from witnesses, as the Judge in the present case did. Such advice at the outset of a trial should reduce the risk of the jury construing the warning as being directed at specific evidence, which could unfairly devalue that evidence. Moreover, this advice may well be more helpful to jurors in advance rather than after they have heard all the evidence and are attempting to evaluate it as by that stage impressions based on demeanour may already have been formed. In addition, in a case where, because of the way the trial has developed, a judge thinks it necessary to say something about demeanour in summing up to the jury, a reminder of what he or she said in the opening statement may be sufficient (and relatively risk-free).

  68. We have quoted the Judge’s opening remarks in the present case at [21] above. We suggest a rather fuller statement, beginning with an identification of factors which will help jurors to determine whether a witness is telling the truth. This should reflect the particular circumstances of the case as far as they can be assessed in advance, but could include reference to considerations such as:

    1. Whether the witness’s evidence is consistent with the evidence of other witnesses which the jury has accepted.

    2. Whether the witness’s evidence is consistent with objective evidence such as documents or text messages, and if it is not, what explanation is offered for any inconsistencies.

    3. Whether the witness’s account is inherently plausible – does it make sense? Is it likely that people would have acted in the way suggested?

    4. Whether the witness has been consistent in their account over time and, if not, why not?

    It should be emphasised that the jury must consider a witness’s evidence in the context of all the evidence in the case.

  69. Following that, a direction along the following lines could be given:

  70. I must warn you, though, that simply observing witnesses and watching their demeanour as they give evidence is not a good way to assess the truth or falsity of their evidence. For example, a witness may not appear confident or may hesitate, fidget or look away when giving evidence. That doesn’t necessarily mean that their evidence is untruthful. The witness may be understandably nervous giving evidence in an unfamiliar environment in front of unknown people. Or there may be cultural reasons for the way a witness presents. On the other hand, a witness may appear confident, open and persuasive but nevertheless be untruthful. And remember that even an honest witness can be mistaken.

    Things like gestures or tone of voice may sometimes help you to understand what the witness actually means. But you should be cautious about thinking that they will help you much in determining whether or not the witness is telling the truth.

  71. Two things follow from what we have said. First, the references that judges sometimes make to the help to be obtained from observing demeanour or body language when determining credibility are likely to be misleading and are better avoided, for the reasons explained above. Further, simply inviting the jury to use “robust common sense” in assessing conflicting evidence is unlikely to provide any real assistance and may positively mislead them. Second, there are various ways in which the risk of inappropriate reliance on demeanour by juries can be addressed in summing up. While there may be occasions where tailored demeanour directions are the only option, judges should always consider whether there are other options given the risks associated with such directions.[35] If observations along the lines earlier discussed are made in the judge’s opening remarks, it may well be possible to address any potential prejudice from inappropriate reliance on demeanour in a way that avoids risk by simply referring back to those observations.

  72. We now turn to consider the two aspects of the case that Mr Chisnall submits raise demeanour concerns that should have been addressed in the Judge’s summing up.

  73. First, there is the complainant’s reaction when shown the photograph containing images of the towels. People may react in a discernible way when reminded of occasions which have been deeply traumatic for them. Such a reaction can be a natural response. In this case, an issue for the jury was whether the complainant’s reaction was a genuine reaction to a traumatic event or not. That issue was squarely raised by the prosecutor in his cross-examination of Mr Taniwha, and the effect of Mr Taniwha’s response was that the complainant’s reaction was not genuine but was put on for the jury.

  74. There was strong contemporaneous evidence corroborating the complainant’s account. On the day following the towel incident, Mr Taniwha sent a text to the complainant asking if she was coming to his place that evening. She replied “No I don’t want a repeat of yesterday”.[36] Mr Taniwha then sent a text asking whether they were going to catch up. The complainant’s text in response said: “Not tonight. [X] said you’re drinking again. Not in the mood for a fight and a towel being shoved in my mouth. That’s fucked up shit.” Mr Taniwha’s text in reply said: “I won’t do that honey.” Obviously, this exchange was open to the interpretation that Mr Taniwha acknowledged that the towel incident had occurred, but said it would not happen again. Indeed, it is difficult to explain the exchange on any other basis.

  75. The Judge instructed the jury that they were to decide the case uninfluenced by prejudice or sympathy. As the Court of Appeal noted in Rua v R, the risk of unfair prejudice to a defendant arising from a complainant’s demeanour (such as evident distress when giving evidence) may be met by a strong sympathy and prejudice direction.[37] The Judge also instructed the jury to assess the evidence of the witnesses “clinically and carefully” and in light of all the evidence presented. Further, the Judge emphasised to the jury that they had access to many, many text messages between the complainant and Mr Taniwha during their relationship, which they should consider in evaluating the witnesses’ evidence (the defence had argued strongly that text messages sent by the complainant to Mr Taniwha during their relationship were inconsistent with her account of events). All of this would have brought home to the jury the need for caution.

  76. If the Judge had given some form of demeanour warning in his summing up in the particular context of the towel incident, we think there would have been a substantial risk that the difficulty identified by the Court of Appeal in E (CA799/2012) would have arisen, namely that the jury would have interpreted it as an indication that they should disbelieve or place little weight on the complainant’s evidence in respect of this incident. Because the issue of whether the complainant’s reaction was a genuine reaction to the memory of a traumatic event or was fabricated for the benefit of the jury was squarely raised, a warning from the Judge about the dangers of relying on demeanour in assessing credibility may well have resulted in the jury disregarding or devaluing the significance of the reaction rather than determining whether or not it was genuine.

  77. In these circumstances, we do not consider that the towel incident required some form of demeanour warning from the Judge in his summing up.

  78. Second, there is the prosecutor’s reference to the complainant’s general demeanour while giving evidence as the first of the five reasons that he advanced for the jury to accept the Crown case,[38] which is of greater concern. As we have noted, in this case the complainant’s evidence was given by way of playing her evidential video and, viva voce, by means of CCTV. There is a question as to whether this has any relevance for the purpose of the demeanour analysis. Defence counsel often attempt to resist Crown applications for evidence to be given by alternative means by arguing that CCTV and the like impede the ability of juries to assess whether or not the witness is telling the truth, and this undermines the defendant’s right to a fair trial. In general, the courts have not been sympathetic to such claims,[39] and such research as there is indicates that the mode of giving evidence makes no meaningful difference to the ability to assess veracity.[40] Accordingly, we regard this feature of the case as irrelevant for present purposes.

  79. Submissions based on general demeanour such as that made by prosecuting counsel in this case are frequently made to juries by prosecuting and defence counsel alike. To the extent that they are an outright appeal to a demeanour-based assessment, such submissions are better not made, certainly in an unqualified way,[41] although the fact that such a submission is made will not necessarily result in a miscarriage of justice. In this case, we do not consider that the Judge’s failure to give a tailored demeanour direction in summing up to the jury gave rise to a miscarriage of justice, for the following reasons.

  80. First, although the prosecutor relied on the complainant’s demeanour in his closing address, he did not dwell on it or over-emphasise it. When making the submission, the prosecutor made it clear that a person could not be judged simply by how that person appears. He noted that the complainant was in the witness box for around 10 hours, both giving her evidence-in-chief and under cross-examination, and was consistent in the material details of her account and argued that it would have been difficult to achieve this if she had been lying.

  81. Second, what the Judge said about demeanour in his opening remarks to the jury, together with the points which he made in his summing up (as described above), were sufficient to bring home to the jury what their approach to the assessment of the competing accounts of the complainant and Mr Taniwha should be. Some of the questions which the Judge posed to assist the jury to determine credibility and reliability[42] were, with respect, not particularly helpful as they simply restated the issue. But others were of assistance, for example, whether the witness’s evidence fitted in with the text messages, whether it was consistent with the evidence of other witnesses and whether the witness was consistent in the account which he or she gave.

  82. Accordingly, we consider that this is not a case where there is a risk of miscarriage of justice as a result of the jury making an illegitimate, demeanour-based assessment of credibility.

  83. Police safety order

  84. As we have noted, the circumstances giving rise to the issue of the PSO were canvassed in evidence and the prosecutor made reference to it in both his opening and closing addresses. Defence counsel made no reference to it in closing, nor did the Judge in his summing up.

  85. Mr Chisnall submitted that it was a central plank of the Crown’s case that Mr Taniwha was controlling and intimidating towards the complainant. He argued that the Crown had deployed the PSO incident in support of its theory of the case, although it post-dated the final offence for which Mr Taniwha was charged.[43] While he accepted that the PSO incident had some relevance to the issues in dispute at trial (whether the complainant had consented to the sexual activity and, if not, whether Mr Taniwha reasonably believed she had), it contained other elements (specifically the threat to the complainant’s flatmate and Mr Taniwha’s attitude to police) that meant the risk of unfair prejudice if it was admitted outweighed its probative value. Accordingly, it was inadmissible. If it was admissible, a “proper use” direction should have been given in light of the prejudice associated with it.

  86. Similar arguments were made to the Court of Appeal, which rejected them.[44] The Court held that given the dispute between the complainant and Mr Taniwha as to the nature of their relationship, and in particular Mr Taniwha’s contention that the complainant was free to leave the relationship at any time, the circumstances in which the relationship did come to an end were relevant. The Court was satisfied that the probative value of the evidence was not outweighed by its prejudicial effect, particularly given the other material before the jury as to the nature of the parties’ relationship at that time. The Court held that the evidence was admissible and that no “proper use” direction was required.

  87. We will deal with this issue shortly. Section 40(1) of the Evidence Act defines propensity evidence and s 43 sets out the test which must be met where the prosecution wishes to offer propensity evidence about a defendant. These provisions were discussed by this Court in Mahomed v R.[45] All members of the Court agreed that the rationale for the admission of propensity evidence rests principally on the concepts of coincidence and probability.[46]

  88. In the much cited minority judgment of McGrath and William Young JJ, William Young J noted that propensity evidence relating to interactions between a defendant and a victim may have an important explanatory value in terms of the background or relationship between those involved. This includes situations where the events are so interconnected with the alleged offending that the jury will not be able to understand properly what happened without hearing evidence about them. In this type of case:[47]

  89. .... the Crown will not usually be much reliant on ideas about coincidence and probability and the wrongfulness of the defendant’s conduct will usually be so closely connected to the core elements of the case against the defendant as to leave little scope for unfair prejudicial effect.

  90. William Young J went on to consider when a propensity direction will be required. He identified two main situations, namely where:[48]

  91. (a)

    the Crown is relying on propensity reasoning and in doing so is invoking ideas about coincidence or probability; and


    the evidence involves aspersions on the character of the defendant in respects not directly associated with the alleged offending.

    William Young J also noted that a direction may be required where there is a danger that the jury will not realise the relevance of the evidence or some risk of unfair prejudice associated with the evidence. Where the evidence in question fell within the description of “propensity evidence” but was not led primarily in reliance on coincidence or probability reasoning, a specific direction might well not be required.[49]

  92. The majority in Mahomed did not find it necessary to engage with this analysis given the view it took of the case. We think the analysis is correct, and we endorse it.

  93. The events leading up to the issuing of the PSO, and those after it was issued, were essential elements of the factual narrative. They explained how the relationship between Mr Taniwha and the complainant came to police attention, how the relationship came to an end and how ultimately the complainant came to lay a complaint with police about his conduct towards her. Evidence of these events was admissible on this basis. Moreover, these events were so closely linked to the core elements of the case against Mr Taniwha that there was little risk of any unfair prejudicial effect. Other clearly admissible evidence indicated that Mr Taniwha was controlling and aggressive towards the complainant for much of their relationship and that he had little regard for police. The evidence at issue was simply more of the same. We do not accept Mr Chisnall’s submission that it contained additional or new prejudicial material sufficient to raise the risk that the jury might have used it improperly.

  94. In the result, we consider that the evidence in relation to the PSO incident was admissible and did not require a “proper use” warning.

  95. Decision

  96. The appeal is dismissed.

[1] See, for example, Aldert Vrij Detecting Lies and Deceit: Pitfalls and Opportunities (2nd ed, John Wiley & Sons, Chichester (Eng), 2008).

[2] See, for example, Eiten Elaad “Effects of Feedback on the Overestimated Capacity to Detect Lies and the Underestimated Ability to Tell Lies” (2003) 17 Appl Cognit Psychol 349.

[3] R v Taniwha DC Wellington CRI-2013-032-496, 17 December 2013 (Judge I G Mill).

[4] For their content, see below at [50].

[5] See Domestic Violence Act 1995, pt 6A.

[6] See Mahomed v R [2011] NZSC 52, [2011] 3 NZLR 145 at [7] per Tipping J (for Elias CJ, Blanchard J and himself).

[7] Taniwha v R [2015] NZSC 200.

[8] E (CA799/2012) v R [2013] NZCA 678.

[9] Taniwha v R [2015] NZCA 434 (Ellen France P, Courtney and Clifford JJ) [Taniwha (CA)] at [26].

[10] At [27].

[11] At [28].

[12] At [28]. For the contents of the Judge’s warning, see below at [21].

[13] Taniwha (CA), above n 9, at [29].

[14] Law Commission Evidence Law: Character and Credibility (NZLC PP27, 1997).

[15] At [115] (footnotes omitted).

[16] At [116] and [117].

[17] See, for example, Vrij, above n 1, at 60.

[18] Lord Bingham “Assessing Contentious Eyewitness Evidence: A Judicial View” in Antony HeatonArmstrong and others (eds) Witness Testimony, Psychological, Investigative and Evidential Perspectives (Oxford University Press, Oxford, 2006) 327 at [18.2].

[19] Robert Fisher “The Demeanour Fallacy” [2014] NZ L Rev 575 at 577 (footnote omitted).

[20] R v Munro [2007] NZCA 510, [2008] 2 NZLR 87.

[21] At [74].

[22] At [79]–[80].

[23] At [81].

[24] Fisher, above n 19, at 582 (footnotes omitted).

[25] At 590–591. Lord Bingham accepted the same point: see Bingham, above n 18, at [18.19].

[26] For a useful discussion of the authorities, see R v Tait (1979) 46 FLR 386 (FCA) at 401–405. In that case, two common law exceptions are identified – to prevent disruption of a trial by rioters or protestors and to preserve commercially secret processes: at 489–490. See also Attorney General v Leveller Magazine Ltd [1979] AC 440 (HL) at 449–450 per Lord Diplock and R (on the application of C) v Secretary of State for Justice [2016] UKSC 2, [2016] 1 WLR 444 at [1] and [16]–[17]. The courts are now, of course, given broader powers by statute to restrict public access to, or reporting of, criminal trials in certain situations: see, in particular, s 197 of the Criminal Procedure Act 2011.

[27] Law Commission The 2013 Review of the Evidence Act 2006 (NZLC R127, 2013) at [11.14] (footnote omitted).

[28] E (CA799/2012) v R, above n 8, at [41].

[29] At [43].

[30] At [44]–[45] (footnotes omitted).

[31] Fisher, above n 19, at 600–601.

[32] In Owen v R [2007] NZSC 102, [2008] 2 NZLR 37 (an unreasonable verdict case), the Court accepted that a jury may have advantages over an appellate court required to review the evidence at trial, particularly as to the assessment of the honesty and reliability of witnesses: see [13](b).

[33] We note that in 2013, an earlier Cabinet decision to introduce a new requirement that jurors be directed that they should not draw any inference from the demeanour of child witnesses when giving evidence by alternative means was rescinded: Cabinet Paper “Amendments to the Evidence Act 2006” (signed by the Minister on 12 November 2013) at [43]–[46].

[34] As suggested in E (CA799/2012), above n 8.

[35] See below at [51].

[36] The text messages are written in texting language. They have been put into standard English for the purpose of the judgment.

[37] Rua v R [2014] NZCA 599 at [49](b).

[38] See above at [9].

[39] See the discussion in Richard Mahoney and others The Evidence Act 2006: Act and Analysis (3rd ed, Thomson Reuters, Wellington, 2014) at [EV103.09(1)].

[40] See, for example, HK Orcutt and others “Detecting Deception in Children’s Testimony: Factfinders Abilities to Reach the Truth in Open Court and Close-Circuit Trials” (2001) 25 Law & Hum Behav 339; Sara Landström, Pär Anders Granhag and Maria Hartwig “Witnesses Appearing Live Versus on Video: Effects on Observers’ Perception, Veracity Assessments and Memory” (2005) 19 Appl Cognit Psychol 913; Law Commission The Evidence of Children and Other Vulnerable Witnesses (NZLC PP26, 1996) at [113]; and Elisabeth McDonald and Yvette Tinsley “Evidence Issues” in Elisabeth McDonald and Yvette Tinsley (eds) From Real Rape to Real Justice: Prosecuting Rape in New Zealand (Victoria University Press, Wellington, 2011) 279 at 283–284.

[41] In Rua v R, above n 37, at [46] the Court of Appeal described a submission to the jury similar to that made by the prosecutor in this case as “more than a little unwise given the developing judicial consensus as to the general limitations of demeanour as a reliable guide to assessing the credibility and reliability of witnesses”.

[42] See above at [23].

[43] Mr Chisnall did not argue that the evidence was inadmissible simply because it post-dated the date of the incident giving rise to the last charge, however.

[44] Taniwha (CA), above n 9, at [30]–[53].

[45] Mahomed, above n 6.

[46] At [3] per Tipping J (for Elias CJ, Blanchard J and himself) and at [51] and [81] and following per William Young J (for McGrath J and himself).

[47] At [90] (footnotes omitted).

[48] At [91].

[49] At [92].


N P Chisnall and M F Laracy (instructed by Public Defence Service, Wellington) for Appellant.

M D Downs and F G Biggs (instructed by Crown Law Office, Wellington) for Respondent.

all rights reserved