T (CA36/2022) v THE KING [2023] NZCA 299

T (CA36/2022) v THE KING [2023] NZCA 299 [17 July 2023]NOTE: PUBLICATION OF NAME, ADDRESS, OCCUPATION ORIDENTIFYING PARTICULARS OF COMPLAINANT PROHIBITED BYSS 203 AND 204 OF THE CRIMINAL PROCEDURE ACT 2011.IN THE COURT OF APPEAL OF NEW ZEALANDI TE KŌTI PĪRA O AOTEAROACA36/2022[2023] NZCA 299BETWEEN T (CA36/2022)AppellantAND THE KINGRespondentHearing: 10 May 2023Court: Miller, Woolford and Cull JJCounsel: H J Croucher and C A Hardy for AppellantZ R Hamill for RespondentJudgment: 17 July 2023 at 11.00 amJUDGMENT OF THE COURTA The application for leave to adduce further evidence is declined.B The conviction appeal is dismissed.C The application for leave to appeal sentence out of time is granted.D The sentence appeal is allowed. The sentences on the charges of breach of aprotection order in relation to NF are set aside and sentences of six weeks'home detention substituted, to run concurrently with the home detentionsentences for doing an indecent act and breach of a protection order inrelation to N. The appellant must resume his home detention sentence on24 July 2023. His bail is revoked from that date.____________________________________________________________________REASONS OF THE COURT(Given by Miller J)[1] T was found guilty at trial on one charge of doing an indecent act on a child,1his 8-year-old daughter N, and one of breaching a protection order by doing that act.2He had pleaded guilty to three charges of breaching the protection order in relation tothe child's mother, NF.3 These charges related to communications in which he soughtto resume his relationship with NF. Judge Moala sentenced him to concurrentsentences of eight months' home detention on all charges.4[2] T appeals his convictions relating to N on the ground that there is freshevidence tending to show he suffers from sexsomnia. He says that he has a viabledefence of sane or insane automatism.5[3] He appeals his sentence on the protection order charges relating to NF, sayingthat if his convictions relating to N are quashed then the sentence was excessive.The Crown agrees that a sentence of six weeks' home detention would be theappropriate sentence in that event, meaning he would be released because he servedpart of his sentence before being bailed pending this appeal.The trial[4] In 2019 T had shared care of N, who resided mostly with her mother but wouldvisit T overnight at his sister's, where he was living at the time. On 14 July 2019,N told NF that he had touched her "privates" while she was in bed. In an evidentialinterview, N said that she was sleeping at her aunt's house and T came up onto her bedand touched her private parts underneath her pyjamas. (She had been in a bed and hehad been on a mattress in the same room.) She then woke up and the morning alarmwent off and she got dressed and went to school.1 Crimes Act 1961, s 132(3).2 Family Violence Act 2018, ss 9(2)(b), 90(a) and 112(1)(a).3 Domestic Violence Act 1995, ss 19(2)(b), 19(2)(e) and 49(1)(b); and Family Violence Act,ss 90(b) and 112(1)(a).4 Police v [T] [2021] NZDC 24463 [Sentencing notes] at [14].5 C (CA223/2020) v R [2021] NZCA 80, [2021] 3 NZLR 152 at [38]–[50].[5] Trial counsel for T investigated a defence of sexsomnia, briefing Dr AntonioFernando, a consultant psychiatrist and sleep specialist. Dr Fernando interviewed Tand his sister (for any family history of parasomnias) and referred T for an overnightsleep study in a sleep clinic.[6] Dr Fernando provided a first report on 20 May 2021. At that time the sleepstudy had not been completed but he had interviewed members of T's family of origin.T had no childhood history of sleepwalking or sleep talking, but relatives had adultparasomnias. T had reported two episodes in which he had, according to NF,performed sexual activities while he was asleep. T has a history of bipolar disorderbut he said he was not experiencing symptoms at the time. T denied any memory ofthe incident. Dr Fernando could not exclude malingering.[7] Dr Fernando could not offer a confident diagnosis. There were multiple factorssuggesting sexsomnia was remotely possible but not very likely, and only one factorwhich strongly suggested it. That factor was a reported statement by NF, whoacknowledged to Dr Fernando in a phone conversation that T had "done things" to herin his sleep. She was otherwise uncooperative and declined to expand on thisstatement. Dr Fernando was unable to speak to a previous bed partner of T's.[8] A further report was provided on 9 July 2021. Through the intervention ofthe police and defence counsel, NF had agreed to speak to Dr Fernando in the interim.She mentioned a single incident, in 2009, in which T had displayed abnormal sexualbehaviour. When they were in bed he rubbed his hands on his genitals then put hishands on her face and mouth. He was quiet when she asked him what he was doing.She was unsure if he was asleep and said he was having a bipolar episode at the time.She reported that when manic he wanted sex all the time. She was not aware of anyother unusual sleep-related behaviours.[9] The sleep study had been completed. T was diagnosed with moderateobstructive sleep apnoea, and he also displayed periodic limb movements. These areknown causes of sleep disturbance, which can trigger parasomnias. No other evidenceof abnormal sleep behaviours was noted. The study did not support a diagnosis ofsexsomnia, but neither could a one-night study exclude it. The sleep history providedby T remained confusing.[10] Dr Fernando concluded that the likelihood of sexsomnia explaining the allegedoffending was low.[11] The trial commenced on 2 August 2021. N and NF gave evidence, the latter todepose that N had reported the incident to her after the sleepover. Consistent withDr Fernando's reports, the defence did not run sexsomnia. NF was not asked aboutany history of T exhibiting sexual behaviour during sleep.[12] The defence was that N was mistaken about being touched. It was suggestedthat the complaint may have stemmed from conflict between T and NF over N's care.[13] The jury found T guilty on both charges. As noted, he pleaded guilty to otherbreaches of the protection order involving NF.The new evidence[14] T appealed against conviction on 26 January 2022.[15] NF and T have been engaged in Family Court proceedings over the care of N.His conviction has naturally raised questions about the extent to which he should havecontact with N. He has maintained that he did not consciously touch her sexually.Both parents have sworn affidavits. On 14 March 2022 T deposed that NF had toldhim of one episode of sexsomnia during their relationship. NF responded in anaffidavit dated 27 March 2022 that "[i]t is correct that [T] touched me while he wasasleep."[16] This statement led appellate counsel to seek a third report from Dr Fernando.It is dated 2 September 2022. He stated that because there were no other bed partnerswho could confirm or deny sleep-related sexual behaviour, NF's statement was"crucial" in making the diagnosis. He formed the opinion that the possibility of Thaving sexsomnia "should be strongly considered".6 He drew attention to supportinginformation, in the form of the family history of parasomnias and results of thesleep study.[17] An application was filed to adduce fresh evidence on appeal. Affidavits weresworn by Dr Fernando (on 3 November 2022) and Dr Peter Dean (on23 February 2023). Dr Dean is a consultant forensic psychiatrist called by the Crown.Both gave oral evidence before us. NF was not called, but we were provided withanother affidavit which she swore in the Family Court proceeding, on13 December 2022. There she sought to correct her statement that T had touched herwhen he was asleep. She said she had been too embarrassed to describe the incidentfully. She now did so:3. For a start, I never said it was an episode of sexsomnia. On theoccasion in question [T] started touching me. I initially thought hewas asleep. But then I believe he was awake because he was touchinghis genitals then rubbing my mouth and sticking his fingers in mymouth. Then he would smell his hand, he did this a few times,touching his genitals then rubbing them on my face. When I realisedwhat he was doing I lay there shocked and disgusted until I grabbedhis hand and moved it away and moved myself further away from himin bed. He then stopped. I believe he was awake as he was able tocontrol his actions. There would have been no need for him to smellhis hand if he was asleep. I got up and then went to the toilet to getaway. When I came back, he had rolled over and I didn't say anythingand faced the opposite direction, feeling sick to my stomach.[18] It will be seen that NF was describing a single incident during theirrelationship. Her reaction suggests it was not normal sexual behaviour for them.She did not say that either spoke during the incident. (In her previous account she saidthat she had asked what he was doing but he did not speak.) She formed the opinionthat he was awake because he was able to control his actions.[19] In evidence Dr Fernando confirmed his opinion that sexsomnia is a possibleexplanation for the index offending and should be investigated. He could not go sofar as to diagnose the condition based on the information available to him.He explained that what changed his previous opinion was NF's statement in her6 He expressed his conclusion in this way in his affidavit sworn on 3 November 2022. It is slightlymore emphatic than the report itself, but we accept that the affidavit better reflects the consideredopinion he had formed at that time.27 March affidavit that T had touched her when he was asleep. He acknowledged thatin her subsequent affidavit she denied that, saying that she believed T was awake, buthe pointed out that she based this opinion on the mistaken assumption that a personcannot engage in complex actions when asleep.[20] Dr Dean found the likelihood that sexsomnia explains the index offending was"almost completely absent". He did not interview T. He did interview NF, whodisclaimed the statement in her 27 March 2022 affidavit, stating that it was due to amisunderstanding between her and her lawyer. She said that she had protested duringthe 2009 incident and did not believe that T was asleep. She appeared to betraumatised by the relationship; he indicated that she described a history of abusivebehaviour. Dr Dean highlighted the absence of any history of parasomnia in T, the factthat he was not initially sleeping in the same bed as N but later got into bed with herduring the alleged offending, and NF's claim that T had been suffering from a manicepisode in 2009.Sexsomnia and forensic diagnosis[21] The term "sexsomnia" was coined in 2003, but the phenomenon of "sleep sex"had been identified in earlier literature.7 It is a parasomnia, one of a large class ofsleeping disorders in which patients experience undesirable events and sleep-relatedbehaviours before, during or immediately after sleep.8 Parasomnias includesleepwalking, sleep talking, sleep terrors, nightmares, restless legs, sleep eating, teethgrinding and sleep sex.9 Sexsomnia is recognised in the most recent editions of theDiagnostic and Statistical Manual of Mental Disorders (DSM-5)10 and theInternational Classification of Sleep Disorders (ICSD-3).11 There is some evidencethat it may be more common than initially assumed, though there is debate about thestatistical quality of studies reporting it.7 Colin M Shapiro, Nikola N Trajanovic and J Paul Fedoroff "Sexsomnia — A New Parasomnia?"(2003) 48 Can J Psychiatry 311 at 314–315.8 Brian J Holoyda and others "Forensic Evaluation of Sexsomnia" (2021) 49 J Am Acad PsychiatryLaw 202 at 202.9 At 202.10 American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders (5th ed,American Psychiatric Association Publishing, Washington DC, 2013) at 399.11 American Academy of Sleep Medicine International Classification of Sleep Disorders (3rd ed,Darien, 2014) at 232.[22] The DSM-5 and ICSD-3 criteria for sexsomnia overlap extensively but are notidentical. They were summarised in a 2021 article in the Journal of the AmericanAcademy of Psychiatry and the Law.12 Dr Fernando was one of the authors.We summarise the criteria as:(a) Recurrent episodes of incomplete awakening from sleep. The DSM-5adds that these usually occur during the first third of the major sleepepisode.(b) Little or no dream imagery.(c) Amnesia for the episodes (complete or partial).(d) Substance use or a co-existing mental or physical condition do notexplain the episodes.(e) Clinically significant distress or impairment.(f) Inappropriate or absent responsiveness to the efforts of others tointervene or redirect the person during the episode.[23] Dr Fernando added that, while parasomnias are very rarely observed by aclinician, the person may exhibit behaviours during sleep that are consistent withparasomnia. The clinician may record an abnormally high number of awakeningsfrom the N3 stage of sleep, or the person may exhibit signs of other sleep pathologies,such as sitting up or restless leg syndrome. These phenomena can be observed in asleep study using video polysomnography. Sleep studies neither prove nor excludesexsomnia but they do aid diagnosis. Sleep studies may also identify phenomena suchas sleep apnoea that can disturb sleep and trigger parasomnias.[24] The 2021 article recommended best practice for diagnosing sexsomnia:13 The diagnosis of sexsomnia requires a thorough clinical history, sleephistory, and collateral history. In addition, an overnight sleep study with full12 Holoyda and others, above n 8.13 At 203–204 (footnotes omitted).electroencephalogram (EEG) and video monitoring should be obtained in aneffort to capture nocturnal sexual behaviors. The clinical history should assessan individual's stress and fatigue levels, psychiatric comorbidities,medications, neurologic history, family history of sleep disorders, alcohol andillicit drug use, and history of violence. A detailed sleep history should screenfor past and current sleep pathologies, including sleepwalking, sleep talking,sleep terrors, nightmares, other parasomnias, obstructive sleep apnoea,periodic limb movements, restless leg syndrome, and nocturnal enuresis.An evaluator should ask about shift work and the degree of sleep deprivationduring the episodes of sexsomnia. In addition, one should ask the evalueeabout environmental factors that disrupt sleep, including ambient noise,sleeping partner noise, and sleeping partner movements. As most people withsexsomnia have poor recall of the events, asking about recollection of theepisode or assessing the degree of amnesia of the event is important.Obtaining collateral history from bed partners, victims, or family memberswho are aware of a childhood history of sleep abnormalities or have witnessedepisodes of sexsomnia, parasomnic behaviors, or sleep-disordered breathingmay also help support or refute a diagnosis.Video polysomnography, or the "sleep study," may assist in diagnosingNREM parasomnias, including sexsomnia. In a recent descriptive study ofpatients complaining of NREM parasomnias, individuals reporting sexsomniaand those reporting other parasomnias both displayed an abnormally highnumber of awakenings from the N3 (or slow-wave) stage of sleep. Mostpatients with sexsomnia in this particular study did not show sexual behaviorsduring the study, however. In fact, there are very few published cases of actualsexsomnia observed during sleep studies. In general, NREM parasomnicbehaviors are rarely captured in sleep laboratories. Sleep studies ofparasomnias, including sexsomnia, are similar to the EEG study of seizuredisorders, in that the diagnostic test may or may not identify the pathology inquestion. Failure to capture behaviors consistent with sexsomnia on a sleepstudy does not exclude the possibility that sexsomnia occurred during thealleged event. Conversely, capturing sexsomnia during a sleep study may beuseful diagnostically but does not automatically allow the examiner to statethat the alleged crime occurred as a result of sexsomnia. Despite theseproblems, some researchers have recommended the routine use of videopolysomnography in cases of suspected NREM parasomnia due to theirpotential diagnostic yield and the identification of additional underlying sleeppathologies. Repeating sleep studies in the hope of capturing sexsomnia,however, may be impractical and of limited utility.[25] It will be seen that triggers for parasomnia may include environmental factorsthat disrupt sleep, such as movement or noise from sleeping partners. A sexual historyfrom bed partners is an important aid to diagnosis. Dr Fernando explained thatbed partners may not know whether the person was asleep, but they can identifybehaviours that differ in some way from normal sexual behaviour in the couple'srelationship.14 The behaviour itself is not always identical but it usually exhibits atheme, such as the manner of touching and lack of engagement with the partner.14 See C (CA223/2020) v R, above n 5, at [22].[26] Dr Dean did not dispute that sexsomnia is now a recognised condition ordisagree with the diagnostic criteria. However, he was plainly sceptical aboutsexsomnia as a likely explanation for alleged offending and wary of diagnosesultimately resting on self-report. He cited a 2010 book stating that the discipline offorensic sleep medicine is at an embryonic stage and noting methodological andethical difficulties in obtaining valid data.15[27] Sleep practitioners such as Dr Fernando engage in the clinical diagnosis andtreatment of sleep disorders, including sexsomnia. This is therapeutic rather thanforensic work and it seems the patient commonly involves their bed partner in thetherapy. There is no reason to doubt the accounts given by the person being diagnosedand their partner.[28] But inappropriate sexual behaviour can have legal consequences, whichintroduces the possibility of malingering. Dr Dean remarked that it is remarkablycommon for people accused of an offence to disclaim any memory of the incident butrarely does parasomnia sustain a defence of automatism in practice. He explained thathe has prepared more than 100 medico-legal reports per year since 2000 and in onlytwo of the cases he has been involved in has the defendant successfully pleadedautomatism.[29] Dr Fernando acknowledged that he is not an expert in malingering.He accepted that a forensic psychiatrist may be able to detect it, though there is nostandard test. Where malingering is in issue, a sleep specialist and a forensicpsychiatrist should work together. For his part, Dr Dean acknowledged that he is nota sleep expert, although he frequently deals with people who assert no memory of anevent and is familiar with the literature on parasomnias. A forensic psychiatrist canevaluate information taken, as in T's case, from a patient's history. He agreed thatsleep studies undertaken by a sleep expert can assist diagnosis, and he would defer tothe opinion of a sleep expert where the outcome depended on such a study.15 Irshaad O Ebrahim and Colin M Shapiro "Medico-legal consequences of parasomnias" inMichael J Thorpy and Giuseppe Plazzi (eds) The Parasomnias and Other Sleep-RelatedMovement Disorders (Cambridge University Press, Cambridge, 2010), as referred to by Dr Dean.[30] The authors of the 2021 article recommended that a forensic psychiatristshould obtain a detailed clinical history and collateral information.16 The latter shouldinclude, when possible, information from previous sexual and bed partners.The authors identified indicators that a person is feigning sexsomnia: efforts toconceal behaviour, repeated episodes of sexual abuse after becoming aware of thesleep-related sexual behaviour, recollection of the episode, and new-onset sexsomniapresenting as a sole parasomnic behaviour.17The appellate test[31] We assess the new evidence against the settled criteria of freshness, cogencyand materiality to verdict, bearing in mind that the ultimate criterion is the interests ofjustice.18Is the evidence fresh?[32] The new evidence on which the appeal rests is that of NF in her affidavit of27 March 2022. It is her statement that T touched her while he was asleep that ledDr Fernando to change his opinion. That statement still underpins his opinion thatsexsomnia should be considered in this case.[33] Ms Hamill, for the Crown, contended that the evidence is not fresh. We acceptthat it recounts conduct which predated the trial and might have been the subject ofevidence there. T had reported his claimed amnesia, and enquiries might haverevealed that NF could confirm he had touched her while he was asleep. However,defence counsel did make enquiries. NF's position at the time was that she was unsureif T was asleep during the single 2009 incident she reported. In her 27 March 2022affidavit, sworn after trial, she appeared to accept that T was asleep. In thecircumstances we are prepared to treat the evidence in that affidavit as fresh, as aresome of the details in NF's 13 December 2022 affidavit. We accept that had trialcounsel known of this information an attempt may have been made to lay a foundationfor a defence of insane or sane automatism.16 Holoyda and others, above n 8, at 205–206.17 At 207–208.18 Lundy v R [2013] UKPC 28, [2014] 2 NZLR 273 at [120]; R v Bain [2007] UKPC 33, (2007) 23CRNZ 71 at [34] and [103]; and Ieremia v R [2020] NZSC 143, [2021] 1 NZLR 168 at [36].Is the evidence cogent?[34] The evidence led NF to offer a revised account, which we have summarised at[17] above. We have also noted what she said to Dr Dean. She gave evidence at trial,as a Crown witness, to recount what N said to her when reporting the alleged offence,but as explained earlier, she was not asked about sexsomnia. She is hostile toward T,understandably so from her perspective, and wishes to sever contact between himand N.[35] We accordingly approach the question of cogency by assuming that whencalled at a retrial and asked about sexsomnia, NF would acknowledge that she said inher 27 March affidavit that T had touched her when he was asleep during the 2009incident, but she would explain it by saying, as recounted in her 13 December affidavit,that it is not her opinion; rather, it is what she assumed at the time. She wouldotherwise confirm that the incident was abnormal behaviour and he did not speakduring it. These features are consistent with sexsomnia. She would say he wasexperiencing a manic episode at the time, and expert evidence would confirm that is apossible explanation for T's behaviour. That can explain an abnormal consciousinterest in sex. The jury would hear expert evidence that NF's disbelief rested on themistaken assumption that a person who is asleep cannot perform complex behaviours.They would also hear that the sleep study identified possible triggers of sleepdisturbance.[36] We think the new evidence is not cogent, meaning it is not probative ofsexsomnia as an explanation for the index charge. This conclusion is no criticism ofDr Fernando's expert evidence, which we found balanced and constructive. It restsrather on the quality of the evidence of fact. We make several points.[37] First, there is very little collateral information to support a diagnosis ofsexsomnia, which is an uncommon condition.(a) T himself has no history of sleepwalking or sleep talking. That doesnot preclude a diagnosis of sexsomnia for T but does make it somewhatless likely; it appears that the majority of diagnosed cases have a historyof childhood parasomnias.(b) There is only a single incident in T's known sexual history that mightbe explained by sexsomnia. That does not preclude a diagnosis ofsexsomnia either but does make it somewhat less likely.(c) There is some collateral information pointing to an alternative mentalhealth explanation for the 2009 incident.(d) The sleep study neither confirms nor excludes sexsomnia; the most thatcan be said about sleep apnoea and leg movement is that they are knowncauses of sleep disturbance, which can trigger parasomnias.[38] Second, the index offending differs materially from the 2009 incident.The complainant is a child, not an adult partner, and they were not in the same bedwhen they went to sleep. The behaviour cannot be said to fit a pattern.[39] Third, we accept that a person may engage in complex behaviour, such aswalking about, cooking or even driving, when asleep, but Dr Fernando told us that,while sleep sex is as varied as conscious sex, the most common behaviours aretouching, fondling, masturbation and intercourse, and it seems that the other person isusually sleeping in the same bed. Partner movement or noise has been identified as atrigger for sleep disturbance. In this case, T moved from a mattress to N's bed duringthe episode and it would be speculative to suggest that the movement was triggeredby anything she did.[40] Fourth, the behaviour happened just before the morning alarm went off,suggesting that it was not in the early phase of T's sleep. As Dr Fernando pointed out,it is possible he had only recently fallen asleep, but again that seems less likely.[41] Finally, for these reasons Dr Fernando is presently unable to confirm adiagnosis of sexsomnia. And as he was careful to explain, it need not follow from adiagnosis of sexsomnia that sexsomnia explains the index offending.Might the evidence result in a different verdict?[42] In its present state, the evidence is in our opinion insufficient to lead a jury,acting reasonably, to entertain a reasonable doubt.19 A finding of automatism foundedon sexsomnia could not be reached without expert evidence. The expert evidence herepoints only to a possible diagnosis which cannot be confirmed without furtherinvestigation, such as an interview with T's previous bed partner. In the absence ofevidence an appellate court will not speculate on what such investigations mightdisclose.[43] Indeed, the trial judge might decline to leave the defence to the jury on theevidence as it stands. The Court explained in C (CA223/2020) v R that classificationof the defendant's condition as insane or sane automatism is a decision for thetrial judge, guided by expert evidence, and, depending on the classification, thedefendant faces a legal or evidential burden.20The conviction appeal: conclusion[44] We are not persuaded that the new evidence points to a miscarriage of justice.The application for leave to adduce further evidence is declined. The convictionappeal is dismissed.The sentence appeal[45] T brought his sentence appeal 84 days out of time, after being made aware ofthe fact that should his conviction appeal succeed, the sentences for the breaches ofthe protection order in relation to NF would remain at eight months' home detention.The sentence appeal was conditional on the conviction appeal succeeding because itmakes no practical difference to T if he must still complete his sentences for theindecent act and breach of a protection order in relation to N.[46] However, we agree with both counsel that the sentences for breach ofprotection order in relation to NF are manifestly excessive and in the circumstances19 R v Bain, above n 18, at [103].20 C (CA223/2020) v R, above n 5, at [47], [58]–[59] and [82].we think the correct course is to grant the extension of time, allow the appeals andreduce the sentences on those charges to six weeks' home detention, to be servedconcurrently with the home detention sentences for doing an indecent act andbreaching the protection order in relation to N.[47] T must now resume serving his sentence of home detention on 24 July 2023.That will give sufficient time for Corrections to make the necessary monitoringarrangements. His bail is revoked, with effect from that date.Solicitors:Public Defence Service, Auckland for AppellantCrown Law Office, Wellington for Respondent