Opposite sides of the same coin: syndrome evidence, child abuse and the wrongful conviction of Peter Hugh McGregor Ellis

Abstract

Introduction of evidence relating to the now-discredited behavioural-science syndrome known as ‘child sexual abuse accommodation syndrome’ in R v Ellis demonstrates the danger of syndrome reasoning in judicial fact finding. Comparable syndrome evidence is still used in the Family Court in the form of ‘parental alienation syndrome’. Ellis should sound the death knell for all forensic applications of unreliable syndrome reasoning in the courts.

Key words:

• child abuse
• expert evidence
• family court
• syndromes

Introduction

The case of The King v Peter Ellis is notorious. Ellis was convicted of the large-scale ritual sexual abuse of children at a crèche where he worked in Christchurch. His conviction long worried defense lawyers and experts on wrongful conviction. The criticisms of his trial, along with Ellis’ arguments on appeal, tend to focus on the evidence of the young complainants and its potential contamination.¹ This article focuses on a less-canvassed aspect of the case: expert evidence relating to an unreliable forensic application of a now-discredited behavioural-science syndrome known as ‘child sexual abuse accommodation syndrome’ (CSAAS). It argues that the use of CSAAS in Ellis demonstrates the danger of syndrome reasoning in judicial fact finding. It explains how comparable syndrome evidence is still used in the family court in the form of ‘parental alienation syndrome’ (PAS) and argues that Ellis should sound the death knell for all forensic applications of unreliable syndrome reasoning in the courts.

Background

In the 1980s, an ideological rift developed within clinical psychology. At one end of the ideological continuum, psychologists who specialised in treating abused children became concerned that decision makers in child-protection institutions (police, child-welfare agencies, counsellors and family courts) were underestimating the prevalence of child abuse and neglect (CAN) (particularly intra-familial abuse), disbelieving children’s truthful accounts, minimising the damaging impact of CAN, and failing to protect children from re-experiencing it (‘child-abuse specialists’).² At the other end of the ideological spectrum, a (smaller) subset of psychologists remained convinced that CAN were rare, ‘low level’ CAN were not harmful, and the mantra of ‘believe the children’ harmed both children making allegations of CAN and parents who were falsely accused (‘family-reunification practitioners’).³

The child-abuse specialists focused on criminal courts. They were concerned that child complainants in abuse trials were being brutalised on cross-examination and juries were wrongfully acquitting CAN perpetrators based on outdated societal myths around the untrustworthiness of children’s claims.⁴

The family-reunification practitioners focused on family courts. They were concerned that claims of CAN were being used tactically by mothers in custody proceedings, and fathers were being falsely accused, protective responses were dividing families and children were losing important paternal relationships.⁵ Saunders et al. (2011) document: ‘Among [family-court] evaluators, the belief that allegations of domestic violence are usually false was part of a constellation of beliefs, including beliefs that false allegations of child abuse and parental alienation by DV survivors are common’.⁶

These philosophical positions are not equivalently supported by research. Validated studies support the correlations between CAN and behavioural consequences.⁷ CAN are horrifyingly prevalent and under-detected, cause long-term damage to victims, and false claims are rare.⁸ The social science establishes that we have a problem with the under-reporting of family violence (FV), not with false claims or over-reaction. Nonetheless, the small subset of family-reunification practitioners continues to cling to beliefs about the relative infrequency of ‘real’ CAN and benign nature of most child mistreatment.⁹ Miller-Perrin and Perrin (2012) explain: ‘[T]he literature would suggest that each of these backlash claims is false. That is, sexual abuse is not uncommon. Children do not tend to exaggerate their victimization’.¹⁰

The more strongly the child-abuse specialists made their claims, the more of a backlash their claims evoked from the reactionary wing of the field.¹¹ The ‘backlash’ against new social-science insights and advocacy around CAN in the 1980s is well documented in the social-science literature. Jenny and Roesler (1993) and Perrin and Perrin (2012) documented the relationship between child-abuse advocates making claims that went beyond what research data could support and the backlash movements that seized on their distortions to argue that false allegations of child abuse are a more serious problem than child abuse itself.¹² Conte (1994) describes the ‘backlash’ as consisting of ‘extreme positions or points of view that challenge concerns about sexually abused children, and that nearly always have the purpose of defending adults accused of sexual abuse, or that otherwise minimize, rationalize, or deny the realities of childhood sexual abuse; lack of supporting research data for extreme positions or points of view; and near total rejection of the knowledge, experience, or realities of childhood sexual abuse’.¹³ Myers (1994) asserts that the backlash was fueled by exaggerated insistence on always believing children’s reports of abuse and employing poor interviewing techniques like those exhibited in the Ellis investigation.¹⁴ Finkelhor (1994) documents the reactive and oppositional nature of the child-abuse backlash, noting that the backlash was driven primarily by parents accused of CAN and lawyers.¹⁵

This philosophical divide among clinical psychologists became entangled in broader societal ideological battles.¹⁶ Child-protection specialists became increasingly aligned with the feminist movement. Domestic-violence (DV) specialists almost universally recognise the prevalence of FV as an offshoot of the patriarchal entitlement of men to use violence against wives and children.¹⁷ It is a canon of international human-rights law that gender-based FV is a systemic violation of women’s and children’s rights.¹⁸

Family-reunification practitioners became embedded in larger ideological beliefs about the harms of the demise of the ‘traditional family’, including women’s equality in the workplace and changing gender norms around parenting.¹⁹ Family-reunification practitioners not only assert that false allegations of CAN are rampant and harming children’s relationships with fathers, but also that men and women are equally likely to perpetrate DV, and that separation, equal division of relationship assets and child-support obligations are unfairly detrimental to men.²⁰ These claims have been seized upon by ‘men’s rights’ groups (MRGs), who argue that false allegations of CSA are the ‘weapon of choice’ for vindictive women post-separation.²¹

There is nothing inherently wrong with mental-health professionals having ideological leanings, if they are candid about their assumptions, fair in their analysis and balanced in their presentation of competing interpretations. Unfortunately, however, candid and balanced analysis is not the hallmark of adversarial systems of justice. Because of the ideological nature of much of this work, lawyers began to seize on the two positions to support their clients’ respective (and often competing) positions. By the late 1980s, child-abuse specialists and family-reunification practitioners were regularly serving as partisan expert witnesses.

As discussed in greater detail below, in criminal cases, prosecutors called child-abuse specialists to corroborate and bolster testimony of child complainants. Their testimony was often based on the (validated) relationship between CAN and behavioural signs and symptoms that correlate with abuse. Sometimes, this was explicitly called ‘Child Abuse Syndrome’ or CSAAS. Other times, language of diagnosis or syndromes was scrupulously avoided, but syndrome reasoning was nonetheless evident. Child-abuse specialists would testify that a particular child was exhibiting behaviours ‘associated’ or ‘consistent with’ abuse.

At the same time, in child-custody cases, lawyers representing parents (overwhelmingly fathers) accused of FV started calling family-reunification practitioners to discredit children’s claims of abuse.²² As Walker (2020) explains, the concept of PAS ‘quickly became admissible in family courts and used by custody evaluators to blame mothers for children not wishing to have contact with their often angry and controlling fathers’.²³ She notes: ‘Mental health professionals were more inclined to believe mothers were “gatekeeping” by overprotecting children or somehow poisoning them against a relationship with their fathers’.²⁴

The testimony of these evaluators was based on the (not validated) correlation between ‘alienating behaviours’ by mothers and behavioural signs and symptoms in children purported to correlate with ‘parental alienation’ (PA).²⁵ This evidence was latched onto with furor by accused perpetrators and the MRGs as evidence of the vindictiveness and untrustworthiness of women’s DV complaints.²⁶ Sometimes, the evidence was explicitly called PAS. Other times, the language of diagnosis or syndromes was scrupulously avoided, but syndrome reasoning was nonetheless evident.²⁷ Family-reunification practitioners would testify that a particular child was exhibiting behaviours ‘associated’ or ‘consistent with’ PA.²⁸

In both sets of cases (CSAAS evidence in criminal cases and PAS evidence in family cases), the evidence suffered from at least two irremediable flaws. First, the ‘behavioural signs’ of CAN and PA were amorphous, subjective and common in all children. In fact, the ‘symptoms’ of CAN and PA are basically the same – fluctuating anger and conciliation with the parent accused of abuse, social withdrawal, developmental regression, acting out, anxiety, depression and self-harm. This is not surprising. These ‘symptoms’ are symptoms of any traumatic childhood experience, not to mention ‘symptoms’ of being a child, adolescent or teenager. The second, related problem with both sets of evidence was that it was fundamentally unreliable as used in forensic contexts. While it is important for clinicians to know that certain behavioural changes in a child could indicate abuse, it cannot be reliably said that any given child demonstrating those behavioural changes has experienced CAN (or PA).

The early parallels between the two unreliable forensic constructs – CSAAS and PAS – are striking. Their trajectories through the criminal and family courts, however, has been dramatically different. As described in greater detail below, decades ago, appellate courts and legislatures reined in the unreliable and unfair use of child-abuse syndrome evidence in criminal courts. This important regulatory correction has no parallel in family court. To this day, family-reunification practitioners not only push their unreliable forensic assessments based on PAS theory, but judges base their custody decisions on them.²⁹ It has been documented for decades that, in custody cases involving FV allegations, the accused perpetrator is more likely to be granted custody than the alleged victim.³⁰ PAS-based ideology is the source of this disparity. This is especially ironic given that the correlations that underlie CSAAS are scientifically validated. Children who have experienced abuse, collectively, are more likely to show CSAAS symptoms than children who have not. No such validation has been demonstrated for PAS.

One interesting question around the treatment of syndrome evidence is why there was a bifurcation between child-abuse specialists testifying in criminal courts and family-reunification practitioners testifying in family courts. Why have defendants accused of criminal CAN not offered evidence from family-reunification practitioners, and why have protective parents not offered evidence from child-abuse specialists in custody proceedings? The short answer is that they have tried. Criminal courts largely reject all syndrome evidence, so they would no more allow evidence of PAS today than evidence of CSAAS. On a handful of occasions, defendants have tried to offer evidence of PAS to support a claim that a child is fabricating claims of CAN because they have been ‘alienated’ by their other parent, but the criminal courts disallow PAS evidence on the ground that it lacks sufficient validity or reliability. For example, in People v Fortin,³¹ Fortin was charged with sexually abusing his 13-year-old niece. Fortin’s defence was that the complainant was lying about the abuse. Fortin sought to introduce expert evidence about PAS from Richard Gardner, the psychiatrist who coined the phrase ‘PA’, to support his claim that the complainant fabricated her allegations because of intra-familial discord. The court noted that Gardner had testified in ‘hundreds’ of family cases but only been permitted to testify in a single criminal case.³² The court held a pretrial hearing to give Fortin the opportunity to demonstrate the admissibility of Gardner’s evidence. Despite the regular admission of Gardner’s theories in family court, the criminal court concluded that PAS was not generally accepted within the field of psychology and therefore inadmissible.

In family court, on the other hand, the ideological project of PAS has largely succeeded. It is well-documented that family court judges and other personnel lack specialised FV expertise and exhibit a ‘pro-contact ideology’.³³ This is the result of the entrenchment of the science fiction of PAS. Because appellate courts and legislatures did not rein in unreliable PAS evidence in the 1990s when they were reining in unreliable syndrome evidence in criminal courts, PAS took over the family court.

Ellis and CSAAS

Ellis was employed at the Christchurch Civic Childcare Centre from 1986 to 1991.³⁴ In November 1991, a mother reported that her child made concerning statements about Ellis, and the centre suspended Ellis and made a police complaint. On 2 December 1991, centre staff hosted a meeting between parents, police and the Department of Social Welfare to advise parents that there had been ‘concerns’ with inappropriate conduct.³⁵ Parents were asked to ‘look for any noticeable changes in their children’s behaviour and any events which might explain them’.³⁶

In total, 118 children were interviewed. Most disclosed no abuse, but more than 20 disclosed to parents and investigators that they had been sexually abused by crèche staff. The ones who disclosed abuse were interviewed repeatedly, and their claims became increasingly bizarre over successive interviews, including being hurt with needles and burning paper, hung from the crèche roof in cages, taken through trapdoors, even infanticide and cannibalism – but medical examinations revealed no physical evidence to corroborate the claims.

Ellis was charged with the sexual abuse of 13 children at the crèche. He consistently denied wrongdoing. The complainants’ testimony was bolstered by evidence from Dr Karen Zelas, whom the Court of Appeal described as ‘a specialist child psychiatrist with international experience in the field of child abuse’.³⁷ Parents had reported behavioural changes by the children around the time of the alleged abuse, including regression in toileting habits, insomnia, nightmares, bed wetting, fear of spiders and tantrums. Dr Zelas testified that ‘the consistency of each complainant’s behaviour’ was ‘consistent with’ that of ‘sexually abused children of the same age group; the intellectual attainment, mental capability, and emotional maturity of the complainant; and the general development level of children of the same age group’.³⁸ Dr Zelas identified 20 behaviours that she opined were consistent with CSA, including common childhood behaviours like tantrums, toileting difficulties, sleep disturbances, sadness, headaches and stomach aches, anxiety, fear of animals and masturbation. She conceded that some behaviours (anxiety, tantrums) could have a ‘multitude’ of causes, including stress and anxiety caused by something other than CSA, but nonetheless opined that the children demonstrated ‘clusters of behaviours’ that indicated that they had been abused.³⁹ She concluded: ‘What tends to be apparent from studies of children who have been sexually abused however is that some [symptoms] are more likely to be indicative of sexual abuse than others and that clustering of a number of symptoms is more likely to indicate abuse than the existence of a solitary symptom’.⁴⁰

The Crown Solicitor repeatedly emphasised Dr Zelas’ evidence in his closing address, arguing that it corroborated the children’s accounts. The prosecution provided the jury a chart of the behavioural changes of each complainant for deliberations.

The High Court directed the jury regarding the CSAAS evidence:

The relevance of the behaviour is this. If a child says that she says [sic] that he or she has been sexually abused then you are entitled to weigh in support of their statements the fact that they have also exhibited behaviours which have been observed by other people in which a lot of children of the same age group who have been sexually abused have also shown.⁴¹

The phrase ‘CSAAS’ was coined by Roland Summit in 1983.⁴² He claimed that clinical studies of large numbers of children in proven cases of CSA demonstrated ‘a typical behavior pattern or syndrome’.⁴³ Summit identified five categories of responses that children who experienced CSA exhibited: secrecy, helplessness and lack of resistance, acceptance and accommodation (imaginary friends, substance use, self-harm, lashing out against the non-abusive parent, academic over-achievement), delayed and unconvincing disclosure (trivial discrepancies, lack of criminal prosecution) and retraction. Summit was not an academic researcher. He was a ‘community psychiatrist’, and his observations relating to CSAAS were the result of ‘personal experience’, ‘clinical anecdotes’ and ‘personal discussions’ with CSA experts.⁴⁴ He claimed that CSAAS was ‘derived from the collective experience of dozens of sexual abuse treatment centers in dealing with thousands of reports or complaints of adult victimization of young children’.⁴⁵

Summit was concerned that traditional ‘mythology and protective denial surrounding sexual abuse’ prevented children’s disclosures from being believed.⁴⁶ He worried that ‘mental health specialists may be more skeptical of reports of sexual abuse and more hesitant to involve themselves as advocates for children than many professionals with less specific training’.⁴⁷ Summit intended his syndrome to inform clinicians, ‘to provide a vehicle for a more sensitive, more therapeutic response to legitimate victims of child sexual abuse and to invite more active, more effective clinical advocacy for the child within the family and within the systems of child protection and criminal justice’.⁴⁸ Summit expressly cautioned:

A syndrome should not be viewed as a procrustean bed which defines and dictates a narrow perception of something as complex as child sexual abuse. Just as the choice to sexualize the relationship with a child includes a broad spectrum of adults acting under widely diverse motivations and rationalizations, the options for the child are also variable. A child who seeks help immediately or who gains effective intervention should not be discarded as contradictory, any more than the syndrome should be disregarded if it fails to include every possible variant. The syndrome represents a common denominator of the most frequently observed victim behaviors.⁴⁹

Summit’s prognostic caution was largely ignored by justice systems. Less than 10 years later, in ‘Abuse of the Child Sexual Abuse Accommodation Syndrome’, Summit complained that ‘court misuse’ of his syndrome had caused ‘distortions’ of it.⁵⁰ He expressed dismay at how his ‘clinical concept’ had been ‘elevated as gospel’, exploited ‘as ammunition for battles in court’ and polarised by ‘false claims advanced by prosecutors’.⁵¹ He admonished:

The CSAAS originated, then, not as a laboratory hypothesis or as a designated study of a defined population. It emerged as a summary of diverse clinical consulting experience, defined at the interface with paradoxical forensic reaction. It should be understood without apology that the CSAAS is a clinical opinion, not a scientific instrument.⁵²

He complained that, nonetheless, ‘lawyers and a few clinical expert witnesses have tended to seize on the CSAAS as a major weapon’.⁵³ He noted that ‘it should have been obvious to a careful reader that the CSAAS was not addressing an illness or disorder’.⁵⁴ He explained: ‘The words identification, detection, diagnosis, symptom, disorder, illness and pathology, which might infer a diagnostic focus, do not appear in the paper [in which he first coined the phrase CSAAS], nor is there a promise of verifying the alleged abuse with such words as test, validate, evaluate, confirm, or prove’. He conceded: ‘The CSAAS acknowledges that there is no clinical method available to distinguish ‘valid’ claims from ‘those that should be treated as fantasy or deception’, and it gives no guidelines for discrimination’.⁵⁵ He noted:

Some of the adversarial alarm and distortion stems from misunderstanding of the word syndrome. In medical tradition it means a list, or pattern of otherwise unrelated factors which can alert the physician to the possibility of disorder. Such a pattern is not diagnostic, and the cause-and-effect relationship among the factors themselves and with the possible problem is generally obscure. In court circles, syndrome seems to mean a diagnosis which an expert witness contrives to prove an injury.⁵⁶

He explained: ‘Had I known the legal consequences of the word at the time, I might better have chosen a name like the Child Sexual Abuse Accommodation Pattern to avoid any pathological or diagnostic implications’.⁵⁷ He explained:

There has been some tendency to use the CSAAS as an offer of proof that a child has been abused. A child may be said to be suffering from or displaying the CSAAS, as if it is a malady that proves the alleged abuse. Or a child’s conspicuous helplessness or silence might be said to be consistent with the CSAAS, as if not complaining proves the complaint.⁵⁸

Ellis offered expert evidence in response to the prosecution’s CSAAS evidence to attempt to address some of this distortion. He called an expert psychiatrist, Dr Keith Le Page, who opined that, while the children’s ‘symptoms’ could be consistent with CSA, they were also ordinary childhood behaviours or could be explained by other stressful events in their lives.⁵⁹

In 1993, Ellis was convicted of 16 counts of indecency with young children and sentenced to 10 years in prison.⁶⁰ In 1994, the Court of Appeal set aside three of Ellis’s convictions, after one child recanted, but otherwise dismissed his appeal. The Court did not address Ellis’ submission that Dr Zelas’ evidence caused a miscarriage of justice. The Court nonetheless noted:

Parents reported behavioural changes around the time the children were at the creche, the most common being regression in toileting habits in the case of children who were supposed to have been fully toilet-trained by the time they moved into the preschool section. Those changes were accompanied in the majority of cases with sleeping problems, nightmares, and night terrors, and there were other difficulties. While Dr Zelas acknowledged that they could be consistent with domestic and other upsets disclosed in the evidence, she considered that what she saw as ‘clusters’ were consistent with sexual abuse.⁶¹

In 1999, the Court of Appeal rejected Ellis’ claim that he suffered a miscarriage of justice after his case was referred for reconsideration by the Governor-General under section 406 (a) of the Crimes Act 1961. The Crown adduced new evidence from experts regarding CSAAS to rebut Ellis’ claim that the recantation by some children was not given sufficient weight during his trial.⁶² The experts claimed that CSAAS ‘suggested that retraction of allegations by sexually abused children is a normal part of the process of having been sexually abused, and that therefore these retractions are not genuine’.⁶³

Ellis offered evidence from Dr Barry Parsonson. Parsonson explained that CSAAS was identified by Summit and ‘suggested characteristics of children who had been sexually abused’.⁶⁴ He explained that more recent research suggested that there was ‘no distinctive behavioural indicator of child sexual abuse’ and ‘[m]any behaviours that were once considered consistent with reports of child sexual abuse are now shown to be characteristic of non-abused children as well’.⁶⁵

Rejecting Ellis’ claim that he had been the victim of a miscarriage of justice, the Court of Appeal did not specifically address the question of whether the trial court erred in admitting Dr Zelas’ CSAAS evidence.⁶⁶

CSAAS and syndrome reasoning in CSA trials

Legislative history

The New Zealand Oxford Dictionary defines a ‘syndrome’ as ‘a group of concurrent symptoms of a disease’ or ‘a characteristic combination of opinions, emotions, behaviour, etc’.⁶⁷ It is the latter definition that is typically at issue in forensic contexts, when one party seeks to introduce evidence that a relevant individual demonstrates behaviours or emotions that are ‘characteristic’ or ‘consistent with’ an underlying cause.

A mixture of common-law and statutory rules has long authorised the admission of counter-intuitive social-science evidence.⁶⁸ CSAAS evidence was expressly authorised by the Evidence Act 1908 (EA 1908).⁶⁹ The EA 1908 contained a special section, s 23 G, authorising expert psychologists to give evidence that ‘the complainant’s behaviour is, from the expert witness’s professional experience or from his or her knowledge of the professional literature, consistent or inconsistent with the behaviour of sexually abused children of the same age group as the complainant’.⁷⁰

In 1999, the New Zealand Law Commission (NZLC) reviewed the Evidence Code.⁷¹ The review included section 23 G, which the NZLC recommended should remain in the EA. The NZLC explained that section 23 G ‘enables an expert to express an opinion on whether the complainant’s observed behaviour is or is not consistent with the behaviour of sexually abused children of the same age group. The purpose of such evidence is not diagnostic. Rather, the purpose of the evidence is educative: to impart specialised knowledge the jury may not otherwise have, in order to help the jury understand the evidence of and about the complainant, and therefore be better able to evaluate it. Part of that purpose is to correct erroneous beliefs that juries may otherwise hold intuitively. That is why such evidence is sometimes called ‘counter-intuitive evidence’: it is offered to show that behaviour a jury might think is inconsistent with claims of sexual abuse is not or may not be so; that children who have been sexually abused have behaved in ways similar to that described of the complainant; and that therefore the complainant’s behaviour neither proves nor disproves that he or she has been sexually abused. The purpose of such evidence is to restore a complainant’s credibility from a debit balance because of jury misapprehension, back to a zero or neutral balance’.⁷²

Parliament rejected the NZLC’s recommendation, repealing section 23 G when it enacted the EA 2006.⁷³ The Associate Minister of Justice recommended to Cabinet that section 23 G be repealed and ‘the admissibility of expert evidence in child sexual abuse cases be dealt with in the same way as expert evidence in any case’.⁷⁴

New Zealand caselaw

The purpose of section 23 G was to allow counter-intuitive evidence to debunk myths about the behaviour of CSA victims that was inconsistent with societal expectations around ‘normal’ reactions.⁷⁵ In practice, however, courts permitted prosecutors to introduce CSAAS evidence not just as counter-intuitive evidence or to rebut defence suggestions that victims’ behaviour was inconsistent with CSA, but also as substantive evidence that complainants were abused because their behaviour was consistent with CSA.

JR Billington QC conceded on behalf of the Crown in his oral submissions in Ellis that ‘there is something of a disconnect in what the Law Commission is saying [in its 1999 report] and what was actually happening in practice … and the difficulty is that apart from the witness being able to comment on the development, the emotional and physical development of the child they had the option of commenting on matters that were consistent with or inconsistent with the abuse and that then led, as we see in the cases … to running the risk of ultimately being ultimate issue questions’.⁷⁶ He noted that, under section 23 G: ‘You can’t speak to the individual victim so all you can give is some generalities’.⁷⁷ He conceded that ‘child behavioural symptoms cannot be considered as indicative or diagnostic of sexual abuse’, and ‘symptoms’ of CSAAS could not indicate CSA because they could occur in children who were not abused.⁷⁸ The Crown acknowledged the fundamentally probabilistic nature of the evidence underlying CSAAS, noting that ‘in a study in relation to where the general population where [sic] identified those who were known to be sexually abused showed a higher proportion of those symptoms than those who did not’ and ‘in surveys of the general population children who have been sexually abused displayed a higher proportion of those symptoms’.⁷⁹ The Crown conceded that the appropriate use of the correlation underlying CSAAS was clinical but not forensic, noting: ‘The clinician knows to explore it. It simply doesn’t prove it. Now one can’t argue with that’.⁸⁰

In ‘Endangered by Junk Science’, I previously wrote about the ‘G2i problem’ that arises when probabilistic syndrome evidence is deployed in individual forensic applications.⁸¹ This characterisation of CSAAS evidence demonstrates the problem. Even if those who have experienced CSA are more likely to have ‘symptoms’ of CSAAS than those who have not, there are still individuals who experience CSA who do not exhibit the symptoms and, more concerningly in forensic application, individuals who have not experienced CSA who exhibit behaviours that are indistinguishable from the ‘symptoms’ of CSAAS. As Justice Glazebrook noted during the oral submissions in Ellis: ‘[T]hese behaviours can be consistent so that there are a range of behaviours that can be consistent with trauma, that trauma can be sexual abuse, it can be other factors. It also, these behaviours can be exhibited by perfectly normal children who were just exhibiting behaviours’.⁸² Or as Justice Williams put it more succinctly: ‘Look, [the CSAAS behaviours are] consistent but, to be fair, everything is consistent’ with CSA.⁸³

R v B I

Since before the Ellis trial, New Zealand appellate courts have wrestled with the line between appropriate and inappropriate uses of CSAAS evidence. The earliest Court of Appeal case to deal with the relevant principles governing admissibility of CSAAS evidence was R v B (R v B I).⁸⁴ B was charged with indecent assault on his daughter. The prosecution sought to introduce expert evidence from a child psychologist that the daughter exhibited ‘typical’ behaviours of sexually abused children, including ‘typical ambivalence’ of ‘desperately trying to make things alright for the family, but wanting to gain protection for herself’, ‘typical’ anger, ‘typical’ withdrawal and secrecy, ‘typical’ inability ‘to tell anyone of her untoward experiences’ out of fear of not being believed, self-blame, shame and ‘typical’ fluctuation between ‘anger and hatred’ and ‘love and affection’ for the perpetrator.⁸⁵ The prosecution appealed the District Court’s refusal to admit the evidence.

The Court of Appeal unanimously upheld exclusion of the evidence. Justice McMullin noted that the psychologist had ‘acceptable academic qualifications in the field of child psychology’, ‘practiced as a consulting psychologist’ and ‘had considerable experience in dealing with children who have been sexually abused’.⁸⁶ He correctly noted, however, that the ‘competence of psychologist as an expert in her field’ was not the issue, but rather it was ‘whether she may give evidence on the matters’ in her opinion evidence.⁸⁷ He concluded:

As child psychology grows as a science it may be possible for experts in that field to demonstrate as matters of expert observation that persons subjected to sexual abuse demonstrate certain characteristics or act in peculiar ways which are so clear and unmistakable that they can be said to be the concomitants of sexual abuse … . However, what is said to be given in evidence here, as disclosed in the statement of evidence, falls short of demonstrating any such state.⁸⁸

Evidence Act 1908 23 G was enacted following the Court’s decision in R v B I.⁸⁹

R v Accused

The Court of Appeal revisited CSAAS evidence in R v Accused.⁹⁰ Accused was charged with indecent assault on his daughter. His defence was that she fabricated her complaint to avoid punishment for school problems. The District Court admitted the evidence of a psychologist that the daughter was exhibiting characteristics ‘associated with sexually abused children’.⁹¹ The judge found that she was ‘well experienced’ and able to opine about the ‘likely cause of’ the daughter’s behaviour.⁹² He found that she could opine whether the daughter’s behaviour was ‘consistent with behaviour of the type that is alleged here’.⁹³ The psychologist testified that ‘sexually abused children exhibited changes in their behavioural patterns which can be identified’ and she observed ‘behavioural characteristics’ that were ‘very differently consistent with’ the ‘characteristics of sexually abused children’, including self-harm, secrecy, shame, embarrassment, exhaustion, depression, difficulty discussing personal issues, avoidance of eye contact and physical contact and a dramatic decline in school performance.⁹⁴

Accused appealed against his conviction on the ground that the judge erred in admitting the evidence. The Court of Appeal allowed the appeal, finding that the evidence should not have been admitted. The Court explained that ‘as a precondition of admissibility the subject-matter upon which the expert expression of opinion is given must be a sufficiently recognised branch of science at the time the evidence is given’.⁹⁵ It admonished that ‘before a psychologist or other similarly qualified person can be allowed to give evidence that a particular child has exhibited traits displayed by sexually abused children generally, it must be demonstrated in an unmistakable and compelling way and by reference to scientific material that the relevant characteristics are signs of child abuse’.⁹⁶ The Court explained:

While the characteristics mentioned by the psychologist were said to be consistent with those the witness had come to know as the characteristics of sexually abused children, some at least of those characteristics, eg pen tattoos on hands and arms, cigarette burns and cuts and lack of eye contact, may very well occur in children who have problems other than sexual abuse.⁹⁷

R v Crime Appeal

In 1991, the Court of Appeal addressed CSAAS evidence again in R v Crime Appeal.⁹⁸ The defendant was charged with sexual offences against his stepdaughter. The prosecution called a psychologist who testified that ‘certain characteristics exhibited by the complainant were ‘strongly indicative’ of her having been sexually abused’.⁹⁹ She identified two categories of ‘relevant behavioural traits’: those ‘specific to sexual abuse and not to other forms of abuse’ and those ‘symptomatic of abuse in general’ that had been ‘frequently observed as occurring in sexual abuse victims’ but were not ‘exclusively’ consistent with CSA.¹⁰⁰ She ‘identified’ both types of traits in the complainant, including ‘preoccupation with sex’, greater than normal ‘sexual experience’ and ‘lack of sexual self-respect’.¹⁰¹ The defendant called a child psychiatrist to testify about ‘possible explanations for the behavioural traits observed in the complainant’.¹⁰² He testified that the behaviour could be the result of CSA but that there could also be psychiatric explanations, including depression or personality disorder.

The Court of Appeal found that the psychologist’s evidence was inadmissible and set aside the conviction. The Court noted that the testimony ‘reinforce[d] the view that this was a case of sexual abuse and considerably limited the scope for the intended defence rejoinder that there might well be another cause for the observed traits, such as a psychiatric disorder’.¹⁰³ The Court admonished:

This is a problem inherent in the kind of evidence that can be given by those practicing the professional treatment of alleged abuse victims. Their opinions properly carry great weight with the layman, who is not always able to differentiate between an opinion based on incontrovertible research material and that which is based on little more than experiment or hypothesis.¹⁰⁴

R v B II

In 2003, the Court of Appeal revisited CSAAS evidence in R v B (R v B II).¹⁰⁵ B was charged with indecent assaults on his 13-year-old brother. He claimed that the allegations were ‘malicious fantasies’.¹⁰⁶ Prior to disclosure, the complainant was involved in various forms of antisocial behaviour, including acts of dishonesty, aggressive behaviour and significant relationship problems.

The prosecution called Dr Gail Ratcliffe, a clinical psychologist with extensive experience assessing and treating children who experienced CSA, to give corroborative evidence under EA 1908 23 G. Dr Ratcliffe testified that the complainant’s behaviour was consistent with the behaviour of children of a similar age who experienced CSA. She opined that the complainant suffered from post-traumatic stress disorder (PTSD), which she described as ‘a recognised consequence of sexual abuse’ in as many as ‘36 per cent of adult survivors’, and that he suffered from ‘severe post-traumatic stress disorder, the source of his disorder being sexual abuse’.¹⁰⁷ The ‘indicators’ upon which she based her diagnosis included nightmares, flashbacks, fear, anger, sleep disturbances, loss of interest in activities and estrangements.¹⁰⁸ She testified that CSA was the only available cause for the onset of PTSD.

The Court of Appeal explained that section 23 G only authorised a psychologist to give ‘background assessments’ but not to express an opinion as to whether CSA occurred.¹⁰⁹ The Court found that Dr Ratcliffe’s expert evidence ‘went much too far’ to ‘whether the appellant had abused the complainant’.¹¹⁰ They explained: ‘The evidence of Dr Ratcliffe was that the onset of PTSD occurred after the event and was a direct result of the sexual abuse which the complainant suffered at the hands of the appellant’.¹¹¹ They noted that ‘the opinion evidence was that PTSD may be diagnostic of abuse and in this instance the expert witness went the further step of saying that on her evaluation of the evidence there was but one source of trauma capable of causing the onset of the disorder’.¹¹²

G v R

G v R¹¹³ is a particularly interesting example of the syndrome battles because it was decided under the modern regime for screening expert evidence, and the trial essentially boiled down to a battle of two syndromes: the Crown pushing a CSAAS theory, and the defence pushing a PAS theory.

G was charged with the indecent assault of his 7-year-old daughter. G denied wrongdoing and blamed his daughter’s mother and grandmother, suggesting that the disclosures resulted from their obsessive concerns.

The Crown led evidence from the mother and other witnesses about the daughter’s unusual and sexualised behaviour to support her testimony that G abused her. The mother gave evidence that the daughter had nightmares, wet the bed, pushed a pencil up a doll’s bottom, was very red around her vagina, had insomnia, suffered from anxiety, exhibited disruptive behaviour at school and started verbally abusing her and physically attacking her sister. She testified that the daughter’s behaviour worsened after overnight visits to G’s house and got better in the intervals between visits. During his closing address, the prosecutor relied on the complainant’s sexualised behaviour, submitting that it was ‘a sharp indicator of sexual knowledge by [the complainant] derived only from an adult’.¹¹⁴

The defence lawyer relied on a PAS theory, arguing that the mother was ‘horrifying’ and had ‘nurtured’ a ‘story’ of CSA that the daughter ‘reluctantly peddled’.¹¹⁵ The Court of Appeal explained:

Both Crown and defence, for different reasons, wanted the evidence of the child’s behaviour to be admitted in evidence. The Crown sought to introduce it in order to support the truthfulness of the complainant’s account. On the other hand, the defence wished to have the evidence introduced to support the theory that the mother’s obsessive concern for the welfare of the child had led to the complainant making a false complaint of sexual abuse.¹¹⁶

The Court found that admission of the evidence created a real risk of miscarriage of justice.¹¹⁷ The Court admonished:

The circumstances of the case were such that the jury could not safely have drawn any inference that the complainant’s behaviour was consistent or inconsistent with sexual abuse having occurred. There were a number of other potential explanations for her behaviour. It might have been accounted for by the separation of her parents, the commencement by them of new relationships, moving house, or other circumstances which might cause a child to demonstrate symptoms of stress and anxiety of the kind shown by this complainant. Her conduct may have been consistent with sexual abuse by the father, but it might not.¹¹⁸

EA revision

R v B I, Accused and Crime Appeal were decided before the United States Supreme Court’s decision in 1993 in Daubert v Merrill Dow Pharmaceuticals¹¹⁹ which is persuasive authority in New Zealand.¹²⁰ R v B I, Accused, Crime Appeal, R v B II and G v R were all decided before the enactment of EA 2006 and repeal of EA 1908 23 G.

R v A

The most recent pronouncement from the Court of Appeal on CSAAS-related evidence came in 2009, in R v A,¹²¹ although the case involved lay-witness observations of behavioural changes rather than expert evidence. A was charged with raping his seven-year-old grandson. The boy did not disclose the rape until he was 13 or 14 and did not make a full statement to police until he was 16.¹²² The prosecution led evidence describing the deterioration in the complainant’s behaviour four years after the incident. The prosecution also led evidence that the complainant wet his bed until 14 and ceased bedwetting when he disclosed the rape. The prosecutor argued in closing:

Between the ages of 7 and 11 I’d suggest [the complainant] didn’t understand the full implications of what had happened to him. It’s only when he becomes more conscious of his own sexuality in puberty that the full significance really hits home to him and that’s when his behaviour starts to deteriorate. We know from his mother that he never stopped wetting the bed from infancy and this continued right through to the age of 14. Indeed, to the week that he disclosed to his family, to his grandmother.¹²³

Defence counsel argued that the behavioural deterioration was not caused by CSA but rather was a sign that he was ‘a disturbed young man who had made a false complaint’.¹²⁴ He argued that there were reasons unrelated to CSA to explain the behavioural changes, including the arrival of a baby sibling, moving homes and financial pressures in the family.

The Court of Appeal quashed A’s conviction, in part because of the evidence relating to the timing of the complainant’s bed wetting. The Court explained that there was ‘an unacceptable risk that the jury would have speculated, as invited to do by the prosecutor’s closing submission, that there was a connection between [the complainant] making his complaint and ceasing bedwetting that somehow supported the complainant’s account of the incident’.¹²⁵

Kentucky case study

One fascinating case study involving both an appellate court’s attempt to regulate unreliable syndrome evidence and the syndrome’s resistance to reform can be found in Kentucky. In a decades-long series of decisions, the Kentucky Supreme Court has fought an epic battle to stop trial judges from allowing juries to consider expert evidence relating to CSAAS or its ‘symptoms’ as proof that CSA occurred, while trial courts have continued to admit the evidence and allow its fallacious reasoning.

The first case in which the Court addressed CSAAS was Bussey v Commonwealth in 1985.¹²⁶ Bussey was charged with the attempted sexual abuse of his 11-year-old daughter. The Commonwealth called psychiatrist Hans Kaak who testified that the daughter exhibited ‘symptoms’ of CSAAS, including secrecy, fear and guilt.¹²⁷ The Kentucky Supreme Court quashed Bussey’s conviction, finding that CSAAS was not generally accepted in the medical community and was irrelevant because Dr Kaak could not establish a causal link between the ‘symptoms’ and CSA.¹²⁸

In 1987, in Hester v Commonwealth,¹²⁹ Hester was charged with sexually abusing his nine- and seven-year-old stepdaughters. The girls reported the abuse to a teacher, social worker and police. They gave videotaped statements detailing the abuse. At trial, the girls recanted, claiming that they had fabricated their allegations because they wanted to live with their father rather than their mother and stepfather. The Commonwealth called a ‘family sociologist’ as an expert.¹³⁰ The expert testified that, when children’s accounts of CSA contain specific details, they were ‘usually’ and ‘almost universally’ true and that, if children later recant, it is because of family pressure or conflicting loyalties.¹³¹ The Kentucky Supreme Court quashed Hester’s conviction. The Court noted that, even though the expert did not use the phrase CSAAS, her evidence was inadmissible for the same reasons given in Bussey: it had not attained general medical acceptance, and there was no logical connection between the way that most children behave when sexually abused and whether a particular child was abused.

In 1989, in Mitchell v Commonwealth,¹³² Mitchell was charged with sexually abusing his young stepchildren. Mitchell claimed that they fabricated the claims at the behest of their mother. A social worker testified about ‘the psychological reactions of children who have been subjected to sexual abuse’.¹³³ She testified that CSAAS was generally accepted by ‘clinicians’.¹³⁴ She testified that CSA began in secrecy, the child was usually helpless against an authoritative figure, this caused the child to accommodate the abuse, and some children did not immediately report and later retracted accusations. The Court noted: ‘There was no testimony that all children who are sexually abused exhibit these symptoms, nor was there testimony that children who have not been sexually abused do not sometimes exhibit some of the elements of the syndrome’.¹³⁵ The Court explained:

The child sexual abuse accommodation syndrome is not like a fingerprint in that it can clearly identify the perpetrator of a crime. Even if all of the children of the appellant exhibited some or all of the symptoms of the syndrome, it would not follow that the appellant was conclusively, or even probably, guilty of child abuse.¹³⁶

This is the problem with all syndrome evidence. The ‘syndrome’ is established through correlations. In the case of CSAAS, there is a correlation between CSA and the responses that the syndrome identifies, but there are children who experience CSA who do not exhibit those responses and children who exhibit those behaviours who were not abused. The correlation itself, between abuse and associated behaviours, is irrelevant to whether a particular child exhibiting the associated behaviours was abused.

In 1991, in Brown v Commonwealth,¹³⁷ Brown was charged with raping his 10-year-old daughter. The Commonwealth called a social worker who testified that the daughter’s ‘deviant behaviour’ was ‘consistent with abuse’.¹³⁸ The Commonwealth also elicited evidence from Georgia Tye, the physician who examined the daughter, that the injuries that she sustained were ‘more likely in a ten year old as opposed to a fourteen year old’.¹³⁹ The inference was ‘that the injury is consistent with the offense charged: sexual abuse that occurred when the victim was ten years old’.¹⁴⁰ The Kentucky Supreme Court quashed Brown’s conviction, finding that admission of the expert evidence was erroneous.

In 1992, in Hellstrom v Commonwealth,¹⁴¹ Hellstrom was charged with sexually abusing his stepdaughter. The Commonwealth called an experienced clinical social worker who testified that the stepdaughter was ‘traumatized’ and had ‘symptoms’ of CSA, including nightmares, anxiety, anger, distrustfulness and stomach problems, but he refrained from classifying them as CSAAS.¹⁴² He testified that delayed disclosure was ‘common’ with CSA.¹⁴³ The Kentucky Supreme Court quashed Hellstrom’s conviction. The Court explained that ‘[a]voiding the term “syndrome”’ did not change the unreliable nature of the evidence.¹⁴⁴ The Court noted: ‘Neither the syndrome nor the symptoms that comprise the syndrome have recognized reliability in diagnosing child sexual abuse as a scientific entity’.¹⁴⁵

In 1996, in Newkirk v Commonwealth,¹⁴⁶ Newkirk was charged with raping his 10-year-old niece. There was no question that she had been raped. The issue was the identity of the perpetrator. The complainant originally identified Newkirk but recanted her identification two days later. At trial, she re-identified Newkirk as the perpetrator. The Commonwealth offered expert evidence from a psychiatrist that ‘the “phenomenon” of recantation in child sex abuse cases’ was common among sexually abused children.¹⁴⁷ The trial court admitted the evidence, ruling that ‘general testimony to show that it is common for children to report sexual abuse, and then retract their allegations’ was distinguishable from the CSAAS evidence that the Kentucky Supreme Court previously held inadmissible.¹⁴⁸ The Kentucky Supreme Court disagreed, finding the evidence ‘remarkably similar to’ the evidence held inadmissible in Hester.¹⁴⁹ The Court explained: ‘In an unbroken line of decisions rendered from time to time throughout the last decade, this Court has repeatedly expressed its distrust of expert testimony which purported to determine criminal conduct based on a perceived psychological syndrome’.¹⁵⁰ The Court found that the evidence remained inadmissible under Daubert, which had recently been decided, explaining that the experts ‘made no pretense of saying whether the scientific principle at issue is subject to testing for falsifiability or refutability, and certainly did not pretend to know the potential error rate. At least on these bases, the evidence in question appears to fall short of the Daubert standard for admissibility’.¹⁵¹ The Court also reiterated that, regardless of Daubert, the syndrome evidence lacked relevancy. The Court predicted the negative consequences of starting down the slippery slope of syndrome reasoning, explaining:

[I]f we should embark upon the path toward freely admitting expert testimony of this nature, it would be difficult to turn back. For instance, if an expert can testify that children who accuse adult family members of sexual abuse are probably telling the truth despite a subsequent recantation, it would follow that a step-parent so accused could introduce expert opinion that children often fabricate accusations against step-parents. Likewise, if symptomatic evidence of CSAAS is admissible to prove abuse, the absence of such symptoms would necessarily be admissible to prove that abuse had not occurred.¹⁵²

In Wilson v Commonwealth,¹⁵³ Wilson was charged with sexually abusing his stepchildren. The Commonwealth called Julie Griffey, the children’s therapist, to testify about her experience regarding the ‘profile’ of CSA perpetrators, the habits of CSA victims and her diagnoses of the children.¹⁵⁴ Griffey specialised in counseling children who experienced CSA. She testified that she diagnosed one child as a CSA victim. The Kentucky Supreme Court quashed Wilson’s conviction, characterising Griffey’s evidence as ‘impermissible habit evidence of victims of child sexual abuse to prove that the person was a member of that class of victims because he or she acted the same way under similar circumstances’.¹⁵⁵ The Court noted: ‘Although she made reference to “the literature”, she based her conclusion on her extensive experience in working with child sexual abuse victims’.¹⁵⁶ The Court concluded: ‘Griffey’s testimony as to her observation of the habits of sexually abused children, as a class, should have been excluded as irrelevant’.¹⁵⁷

In 2008, in Alexander v Commonwealth,¹⁵⁸ Alexander was charged with raping his daughter and her friend. The Commonwealth offered evidence from a counsellor regarding ‘symptoms’ that the girls reported, including age-inappropriate sexual knowledge, reluctance to discuss the abuse, inability to concentrate, anxiety and depression.¹⁵⁹ The counsellor ‘was careful not to offer an express opinion that the symptoms the girls described to her indicated, or were consistent with, sexual abuse. She limited her testimony, rather, to merely listing the symptoms’.¹⁶⁰ The Kentucky Supreme Court held that the counsellor’s ‘symptom’ testimony was inadmissible under its precedents forbidding reference to CSAAS, ‘a constellation of behaviors therapists often find associated with sexual abuse’.¹⁶¹

In Vires v Commonwealth,¹⁶² Vires was charged with raping his young stepdaughter. The stepdaughter was referred to Dr Catherine Gouldin, a pediatrician trained in performing CSA examinations. Dr Gouldin testified that the stepdaughter reported ‘that she had had prior sexual abuse with the last contact around age six, that it was genital to genital contact, and that she currently had behavioral changes, including nightmares and flashbacks for which she was in counseling’.¹⁶³ She testified that she diagnosed the stepdaughter as ‘sexually abused’ based on ‘the behavioral symptoms’ that she demonstrated.¹⁶⁴ The Kentucky Supreme Court quashed Vires’ conviction, holding that ‘a diagnosis of sexual abuse based on behavior has no probative value and is inadmissible’.¹⁶⁵

In 2009, in Sanderson v Commonwealth,¹⁶⁶ Sanderson was charged with sexually abusing his stepdaughter. The child’s mother and father testified about her physical and psychological ‘symptoms’ of CSA.¹⁶⁷ Clinical psychologist Lori Brown testified that it was ‘normal for child victims of sexual abuse, like BT, to add details about their abuse after they have been in counseling for an extended period of time and to appear happy in their outward life and be able to excel in their extracurricular activities and make good grades’.¹⁶⁸

In 2012, in Jackson v Commonwealth,¹⁶⁹ Jackson was charged with sexually abusing four children whom he met as a youth pastor and baseball coach. The Commonwealth offered evidence from Jeff Terry, a child-protection worker who interviewed the complainants. Terry testified that one complainant denied any abuse during early interviews, then later admitted abuse and that such behavior ‘was fairly common among child sexual abuse victims’.¹⁷⁰ He testified that, in 50% to 60% of the CSA cases that he investigated, children initially made denials, then later divulged abuse. The Kentucky Supreme Court found that the trial court committed error in admitting the evidence, explaining:

At its worst, CSAAS testimony functions as a sort of quasi-medical or psychological bolstering of the child witness, since it effectively places the imprimatur of science on the child’s behavior and testimony. While Terry never described his observations as CSAAS or used similar terminology, his testimony was nevertheless similar in substance to other CSAAS testimony.¹⁷¹

In 2013, in Blount v Commonwealth,¹⁷² Blount was charged with sexually abusing his stepdaughter. The trial court allowed the girl’s mother to testify regarding changes in her behaviour, which the mother implied were symptomatic of CSA based upon discussions she had with Dr Brown. The changes of behavior were changes in the complainant’s ‘attitude,’ hairstyle, clothing preferences, grooming habits and weight.¹⁷³ The Kentucky Supreme Court held that admission of the evidence was improper. The Court explained: ‘The clear import of this testimony was to imply that Sally had displayed the signs of CSAAS that Sally’s parents had learned about through counseling with Brown’.¹⁷⁴ The Court noted:

The trial court incorrectly assumed that testimony regarding the child’s changed demeanor was permissible so long as it is not presented as expert testimony linking it to the sexual abuse alleged in the indictment. We note with curiosity that these ‘symptoms’ of sexual abuse (ie, being an apparently happy, well-adjusted child) which were evident in Sanderson, are remarkably different from the ‘signs’ exhibited by [the complainant]. Nevertheless, to suggest that [the complainant’s] behavior was indicative of abuse is just as misleading and prejudicial.¹⁷⁵

In 2015, in King v Commonwealth,¹⁷⁶ King, a youth minister, was charged with sexually abusing a boy who attended a church sleepover. The Commonwealth introduced evidence from the chief investigator that the complainant’s ‘five-day delay in reporting the incident to his mother was not unusual because in her experience with more than 1,500 cases, it was “very rare” for children to immediately report sexual abuse’.¹⁷⁷ On appeal to the Kentucky Supreme Court, the Commonwealth conceded that the testimony was improper, and the Court agreed. The Court noted:

The phenomenon of ‘delayed reporting’ is but one of several (usually stated as five) symptoms claimed to be characteristic of the so-called ‘child sexual abuse accommodation syndrome’ (CSAAS), a theoretical construct promoted by some social and psychological professionals as a useful tool for diagnosing young victims of sexual abuse and for verifying claims of sexual abuse. That Detective Anderkin did not use the term ‘child abuse accommodation syndrome’ and did not relate all of its symptoms to Thomas is inconsequential; omission of the term ‘syndrome’ does not transform the objectionable nature of the testimony into reliable scientific evidence.¹⁷⁸

In Sallee v Commonwealth,¹⁷⁹ Sallee was charged with sexually abusing his granddaughters. The girls’ mother testified about behavioural changes in one daughter before she reported the abuse, including soiling, refusing to bathe on her own and ‘attitude’ changes.¹⁸⁰ The Kentucky Supreme Court found that the court erred in allowing the mother to testify to the behavioural changes, describing the inference that the Commonwealth wanted to draw from the testimony – ‘to point to a group of behaviors and infer that it is probative of her having suffered abuse’ – as ‘speculative’.¹⁸¹

What is fascinating about the decisions from Kentucky is their sheer volume. The Kentucky Supreme Court reviewed the admission of CSAAS evidence on an approximately bi-annual basis for three decades. The Court announced, repeatedly, that CSAAS evidence was not relevant or reliable, but trial courts continued to admit the evidence. These cases paint a striking picture of the resistance of trial courts to reforming unsafe evidentiary practices. In 1993, in Hall v Commonwealth,¹⁸² the Court complained:

Despite the fact that in seven recent cases, this Court has reversed convictions for alleged sexual offenses against children because of improper and prejudicial expert testimony, this issue continues to resurface. By this time, the law in Kentucky should be clear. In the case of a psychologist or social worker, testimony of whether sexual abuse has occurred is impermissible, as these experts are simply not qualified to express an opinion that a person has been sexually abused.¹⁸³

In Alexander, the Court reiterated its ‘long established rule that the alleged “symptoms” of sexual abuse, whether labeled as such or not, are not admissible to suggest an expert diagnosis of abuse’ characterising the evidence as ‘the sort of implicit abuse diagnosis our case law has long disallowed’.¹⁸⁴ In Viles, the Court explained that ‘we have consistently held as inadmissible, expert testimony that a child’s behavioral symptoms are indicative of sexual abuse, on grounds that this is not a generally accepted medical concept’.¹⁸⁵ In Sanderson, the Court complained that Sanderson’s objections should have ‘adequately informed the trial court of the patent inadmissibility of Brown’s CSAAS-related testimony’.¹⁸⁶ In Blount, the Court complained about the ‘apparent belief that, despite our well-established line of cases prohibiting expert opinion on CSAAS, lay witnesses may testify about a putative victim’s behavioral changes to imply what the expert is forbidden from saying directly – that behavioral changes signify sexual abuse’.¹⁸⁷ The Court admonished: ‘Avoiding a term such as “child abuse accommodation syndrome” does not cure the evidentiary defect; it does not cloak the “signs” with the scientific validity needed to make them meaningful and relevant’.¹⁸⁸ In King, the Court complained:

A review of all of the post-Daubert decisions relating to CSAAS reveals the same thing: many times the Commonwealth has attempted to prove its case using CSAAS evidence at trial, but not once has the Commonwealth attempted to prove at a Daubert hearing the scientific reliability and validity of the CSAAS theory. Not once. In Bussey, the very first case in which this issue arose, we highlighted the need to properly establish the validity of the theory. In the thirty years since Bussey, our ruling on the issue has not changed because the evidentiary record has not changed: ‘the record [still] does not reveal any attempt made by the prosecution to establish the credibility of the child sexual abuse accommodation syndrome’.

The validity of the theory was not self-evident in 1985 and it is not self-evident today. However, the gravity of the issue is self-evident. Given the serious personal and social consequences at stake, it would seem likely that over the past three decades the theory would have been exposed to thorough and rigorous research to enable its proponents to demonstrate the validity of the theory. ¹⁸⁹

Ellis

In 2019, Ellis applied to the Supreme Court for leave to appeal against the Court of Appeal’s dismissal of his two prior appeals. For the first time, Ellis’ lawyers vigorously challenged the CSAAS testimony at his trial on the ground that it was insufficiently reliable as expert evidence, and the Supreme Court has finally completed its belated (and posthumous) review of Ellis’ conviction.

Ellis offered evidence from clinical psychologist Dr Thelma Patterson about the lack of scientific underpinning for the evidence given by Dr Zelas. Dr Patterson noted that Dr Zelas rarely considered alternate explanations for the behaviour that ‘she identified as consistent with sexual abuse’ – for example, parental separation.¹⁹⁰

A unanimous Court concluded that Ellis’ conviction should be quashed. The Court noted that Dr Zelas ‘wrongly stated that if a complainant exhibited a cluster of behaviours, this was more likely to indicate abuse than a solitary behaviour’ and ‘engaged in circular reasoning’.¹⁹¹ The Court noted that Dr Zelas’ evidence suggested that the children’s behaviour was ‘diagnostic’, even though she expressly disclaimed that the behaviours ‘were diagnostic of sexual abuse’.¹⁹² The Court noted that the ‘overall effect’ of Dr Zelas’ evidence ‘was that the presence of clusters of behaviours could support a conclusion that sexual abuse had taken place’.¹⁹³ The Court explained that Dr Zelas was dismissive of ‘alternative explanations for the behaviours’ and effectively discounted them.¹⁹⁴ The Court noted that her evidence ‘forestalled the consideration of other possible causes of the behaviour’.¹⁹⁵

The Court concluded that the CSAAS evidence ‘lacked balance, did not inform the jury of other possible causes of the behaviours (or, where she did so, discounted or minimised the other causes), and involved circular reasoning’.¹⁹⁶ The Court explained: ‘The reasoning was circular because it suggested that behaviours that could be explained in a number of ways nevertheless corroborated an allegation of sexual abuse on the basis that they were corroborative because the allegation has been made’.¹⁹⁷ The Court noted that ‘to look to the complainants’ behaviours to support an allegation and then reach back to the allegation as a reason to suggest the behaviours are reliable supports for the allegation is to engage in dangerous circular reasoning’.¹⁹⁸ The Court explained that the ‘overall effect’ of the syndrome evidence ‘was to incorrectly suggest to the jury that the presence of clusters of behaviours could support a conclusion that sexual abuse had taken place’.¹⁹⁹

The Court offered guidance on the role of experts, noting that they were obligated ‘to take care to ensure that their evidence was appropriately balanced by acknowledging any other possible explanation for the behaviours’ and to ‘be “driven by professional skill and experience, not a perceived need to support a preconceived outcome”’.²⁰⁰ The Court explained that ‘it was important that an expert commenting on that evidence fairly acknowledge other possible explanations for the behaviour’.²⁰¹

PAS in family court

On the surface, the Supreme Court’s condemnation of CSAAS evidence in Ellis may seem to hold only historical interest. EA 1908 23 G was repealed in 2006. The appellate caselaw regarding CSAAS evidence was already restrictive before Daubert and the enactment of the EA 2006.

The rejection of syndrome reasoning, however, could and should have far-reaching importance for family law. The syndrome reasoning in Ellis, which now seems anachronistic in retrospect, continues to run amok in cases under the Care of Children Act 2004 (CoCA). Each of the Supreme Court’s criticisms of the CSAAS evidence in Ellis applies with equal if not greater force to the PAS evidence still regularly admitted in the Family Court.

In ‘Endangered by Junk Science’, I documented the family court’s reliance on the PAS construct despite its lack of scientific validation.²⁰² I explained how the PAS theory relies on a purported correlation between a parent’s ‘alienating’ attitudes or behaviours and the child’s behavioural response.²⁰³ I documented how ‘the theory asserts that, when children show certain observable characteristics (signs of alienation), those characteristics are evidence that they have been subjected to ‘alienating behaviours’ by one parent targeting their relationship with their other parent, even though the alienating behaviours are not directly observed’.²⁰⁴ The correlation is established through the use of expert evidence that almost exactly parallels the use of CSAAS in cases like Ellis.

As demonstrated in Appendix 1, the parallels between the use of CSAAS in CSA prosecutions in the 1990s and the use of PAS in the family court today are striking. Like the behavioural ‘symptoms’ of CSAAS, the behavioural symptoms of PAS include vague and conclusory ‘symptoms’ such as a child resisting or refusing contact with a parent, holding similar opinions or using similar language to the ‘alienating’ parent, exaggerating complaints about the ‘rejected’ parent, eavesdropping on parents, fluctuating between friendly and hostile behaviour, giving ‘vague reasons’ for rejecting a parent, experiencing relationship disturbances, ‘acting out’, resisting authority and having a rigid world view.²⁰⁵ Experts are allowed to opine that a child’s rejection of contact with a parent – even an abusive one – is caused by the ‘alienating behaviour’ of the other parent by discounting or minimising and ignoring other causes. The PAS evidence suggests to the family court that ‘clusters’ of child behaviours can support a conclusion that alienating behaviour by their preferred parent has occurred. PAS also suffers from the same problem with circularity: a child’s expression of fear or rejection of a parent are not autonomous because the child’s views are the result of ‘alienation’; a child has been ‘alienated’ because they express fear or rejection of a parent.

The psychologists who offer PA evidence in family court exhibit the same ethically problematic characteristics as the CSAAS experts criticised by the Supreme Court in Ellis. Their evidence constitutes advocacy rather than a balanced approach to causation, fails to acknowledge other possible explanations for a child’s rejection of a parent and is directed at supporting a preconceived outcome: shared care as the only appropriate option.

PA(S) has been rejected as lacking in scientific validity by the American Psychological Association,²⁰⁶ the American Psychiatric Association, the American Medical Association, the National Council of Juvenile and Family Court Judges in the United States,²⁰⁷ the Association of Clinical Psychologists in the United Kingdom²⁰⁸ and the New Zealand Psychological Society.²⁰⁹ Proponents of PA lobbied for its inclusion in both the Fourth and Fifth Editions of the APA’s Diagnostic and Statistical Manual of Mental Disorders (DSM-4 and DSM-5), but the APA rejected the proposals because the syndrome lacked scientific validation.²¹⁰ Paul Fink, the former president of the American Psychiatric Association, explains: ‘PAS has not been recognized by the American Psychiatric Association, the American Medical Association, or the American Psychological Association as a legitimately researched, evidence-based condition’.²¹¹

There is no meaningful distinction between parental alienation syndrome (PAS) and parental alienation (PA) as a dynamic construct or occurrence. Proponents of PA(S) theory dropped the word ‘syndrome’ from their construct after the American Psychiatric Association rejected the inclusion of PAS in the DSM-4. In their proposal for the DSM-5, they employed the new terminology of ‘parental alienation’ rather than PAS. It is clear from their proposal, however, that they still understood PA to be a psychiatric condition capable of identification by a psychologist based on a child’s behaviour (‘strong alignment’, ‘rejection without a legitimate reason’) thought to correlate with un-observed parental conduct (‘accommodating’ the child’s ‘attitude’ and ‘embracing’ the child’s suffering). In their proposal to include PA in the DSM-5, the proposers explained:

[O]ur collection of authors proposes a decidedly less polemical new definition of parental alienation. From now on, we define parental alienation as ‘a particular psychological condition in a child (typically when their parents are engaged in a highly conflictual separation) who strongly aligns himself with one of his parents (the preferred parent) and rejects a relationship with the other parent (the rejected parent) without a legitimate reason’.

The first key idea in this definition is: ‘a child who aligns’. The idea of alliance shows that the ‘fault’ of the favored parent is not having manipulated the child but more often having accommodated his attitude, embraced his suffering and resistance to join with him in a fortress of solidarity. The other key term in the definition, undoubtedly the main psychiatric issue, is the idea of the ‘absence of a legitimate reason’: it is evidence that shortcomings or mistreatment prior to rejection must exclude a diagnosis of parental alienation.²¹²

This ‘new’ definition of PA was again rejected by the American Psychiatric Association.

In 2019, the World Health Organization removed the terms ‘parental alienation’ and ‘parental estrangement’ from the eleventh edition of its International Classification of Diseases. The WHO issued an explanatory note stating: ‘During the development of ICD-11, a decision was made not to include the concept and terminology of “parental alienation” in the classification, because it is not a health care term. The term is rather used in legal contexts, generally in the context of custody disputes in divorce or other partnership dissolution’.²¹³ The WHO expressly disclaimed endorsement of the term ‘parental alienation’ due to concerns about ‘the misuse of the term to undermine the credibility of one parent alleging abuse as a reason for contact refusal’.²¹⁴ They noted: ‘There are no evidence-based health care interventions specifically for parental alienation’.²¹⁵ This is a key point of this article: PA(S) developed entirely as a subject matter of expert evidence for use in family courts and has no counterpart in mainstream psychology or psychiatry.

Conclusion

CSA is a serious and widespread problem. Children’s reports of CAN should always be investigated. Generally, plausible, cogent accounts of abuse should be believed. The purpose of this article is not to argue that children’s accounts of CAN should be treated with skepticism. On the contrary, the temptation for prosecutors to bolster children’s testimony about abuse often arises from a climate in which survivors’ accounts, standing alone, are not considered believable. For example, in R v B I, the Court of Appeal noted that the complainant’s evidence ‘if accepted by the jury, would be sufficient to substantiate a verdict of guilty on all the charges’.²¹⁶ This, it seems, is the real problem: that cogent and credible testimony of a complainant with no reason to fabricate claims of abuse somehow requires bolstering. It is the perceived unwillingness of juries to believe complainants’ evidence without the buttressing of syndrome evidence from experts that creates the temptation to ‘syndromise’ complainants’ experiences. It is the family court’s unwillingness to credit cogent and compelling evidence of FV that makes it susceptible to the ideological PAS evidence of court evaluators.

The purpose of this article is to bring attention to one type of unreliable pseudo-scientific evidence: evidence that attempts to infer from the similarity of one child’s behaviour to a purported profile or pattern the existence of abuse. Both CSAAS and PAS arose out of the same era and have similar origins. Both syndromes were developed by clinicians and based on psychodynamic theory and anecdotal experience rather than academic research employing correlational studies and statistical controls. Summit was concerned CSA victims’ reports were not being believed. Gardner was concerned that children were being encouraged to fabricate claims of CSA and were being believed too readily.²¹⁷ Summit thought that he observed patterns of behaviour in child victims that corroborated their claims. Gardner thought that he saw a pattern in similar cases that indicated pathology in their mothers. Neither theory was validated scientifically, let alone validated for reliable forensic use. Both nonetheless ended up infiltrating the justice system. Summit’s CSAAS theory ran amok in criminal proceedings, offered by prosecutors who believed that ‘symptoms’ of CSAAS could bolster kids claims of CSA. Gardner’s theory still runs amok in the family court, offered by abusive parents to divert responsibility away from their abuse and onto the protective parent.

This is not an issue of ‘qualifications’. It is an issue of reliability and methodology. The mere fact that an expert may have extensive specialised experience or an alphabet soup of letters following their name says nothing about whether an opinion that they offer is derived from valid methodology applied reliability in a particular case.

Interestingly, in R v B I, Justice Casey appeared to suggest that this type of expert evidence that was too unreliable to use in criminal cases was somehow nonetheless helpful in the family court, noting: ‘We were told that material of this nature is commonly accepted under the very wide discretion conferred by [the predecessor to s 178 of the Oranga Tamariki Act 1989] … . [S]uch reports can be relevant in resolving the very different issues raised in the Family Court jurisdiction’.²¹⁸ This reasoning makes no sense. The requirements of logical relevancy in EA 2006 s 7 and reliability of expert evidence in EA 2006 s 25 apply equally in family and criminal cases. Justice Casey’s reasoning is particularly non-sensical given that the next sentence in his opinion was: ‘The differing expert opinions on credibility suggest that in this area techniques have not reached the stage where the accuracy of the results can be relied upon’.²¹⁹ There is no magical reason why junk science should somehow become permissibly reliable in family cases. On the contrary, permitting unreliable expert evidence in family court that a child’s claims of CSA are untrue while prohibiting unreliable expert evidence in criminal court that a child’s claims of CSA are credible would turn the standard of proof on its head. In criminal cases, where defendants face imprisonment and their guilt must be proven beyond a reasonable doubt, perhaps children’s evidence should face scrutiny. As Justice Hardie Boys admonished in R v Crime Appeal:²²⁰ ‘[I]n the anxiety to ensure justice by protecting the interests of child complainants, it is not to be overlooked that justice equally demands care for the interests of the accused. He must be protected against assumptions too readily made, and against generalisations too facilely applied to the particular case’.²²¹ In family court cases, children’s welfare and best interests are supposed to be the paramount consideration, and the standard of proof is balance of probabilities. The family court should require less evidence to establish CAN than a criminal court. There is no ‘accused’ whose rights to natural justice outweigh the child’s right to protection. In either arena, claims of CAN need to be resolved based on evidence that is directly relevant to the credibility of the claims. In either situation, shortcuts based on psychodynamic speculation are tempting but lead to unreliable results.

The parallel appellate treatment of CSAAS evidence in criminal trials and PAS evidence in CoCA proceedings is striking. The senior courts have been fastidious in trying to eliminate the unreliable application of psychodynamic syndrome reasoning in criminal cases. At the same time, they have permitted the same fallacious reasoning and unfair psychological profiling to flourish in the family court. This has created a pernicious feedback loop. PAS-inflected decision making in the family court is self-reinforcing. Every time the court finds that a child’s resistance to contact with one parent was caused by ‘alienating behaviours’ of the other, it reinforces both the existence of the syndrome and the circularity of its underlying assumptions: PA is a rampant and pervasive problem. How do we know? Look how often the family court finds that it has occurred. The feedback loop repeats, the assumptions become more entrenched and the syndrome reasoning results in less reliable decision making.

The Supreme Court was right to quash a conviction based on the syndrome, reasoning that is logically irrelevant, statistically fallacious and scientifically unreliable. Now, it should follow its reasoning to the logical conclusion and rein in the unreliable syndrome reasoning that still dominates the family court. An ‘expert’ opinion regarding whether a particular individual has had any experience (CSA, PA) cannot reliably be derived from the mere fact that they exhibit a cluster of vague and common ‘symptoms’ believed to correlate with that experience. Even the Crown’s expert in Ellis conceded: ‘I am very aware from my own experience as well as from my reading that it is inappropriate to diagnose sexual abuse solely on the basis of behaviour symptoms and that is not something that I would do’. As I have previously documented, the identification of children as ‘alienated’ in the family court, on the other hand, routinely occurs solely based on an assessment that they behave in ways that are consistent with having been exposed to ‘alienating behaviours’.²²²

Notes

1 Ellis v R, No CA 120/98, 9 June 1998, 3.

2 Peter Jaffe and others (eds), Protecting Children from Domestic Violence: Strategies for Community Intervention (Guilford Press 2004).

3 Robert L Emans, ‘Psychology’s Responsibility in False Accusations of Child Abuse’ (1988) 44 Journal of Clinical Psychology 1000; Robert E Emery and Lisa Laumann-Billings, ‘An Overview of the Nature, Causes, and Consequences of Abusive Family Relationships: Toward Differentiating Maltreatment and Violence’ (1998) 53 American Psychologist 121, 125.

4 New Zealand Law Commission (NZLC), The Justice Response to Victims of Sexual Violence: Criminal Trials and Alternative Processes (NZLC R136, 2015).

5 Douglas J Besharov, ‘Unfounded Allegations – A New Child Abuse Problem’ (1986) 83 Public Interest 18.

6 Daniel G Saunders and others, Child Custody Evaluators’ Beliefs About Domestic Abuse Allegations: Their Relationship to Evaluator Demographics, Background, Domestic Violence Knowledge and Custody-Visitation Recommendations (2011) 14.

7 Anja Heilmann and others, ‘Physical Punishment and Child Outcomes: A Narrative Review of Prospective Studies’ (2021) 398 The Lancet 355.

8 R Kim Oates and others, ‘Erroneous Concerns About Child Sexual Abuse’ (2000) 24 Child Abuse and Neglect 149.

9 Saunders and others (n 6).

10 Cindy L Miller-Perrin and Robin D Perrin (eds), Child Maltreatment (Sage 2012) 309.

11 ibid.

12 Carole Jenny and Thomas A Roesler, ‘Quality Assurance – A Response to “the Backlash” Against Child Sexual Abuse Diagnosis and Treatment’ (1993) 2 Journal of Child Sexual Abuse 89; Robin D Perrin and Cindy L Miller-Perrin, ‘Interpersonal Violence as Social Construction: The Potentially Undermining Role of Claims Making and Advocacy Statistics’ (2011) 26 Journal of Interpersonal Violence 3033, 3041.

13 John R Conte, ‘Child Sexual Abuse: Awareness and Backlash’ (1994) 4 Sexual Abuse of Children 224, 228.

14 John EB Myers (ed), The Backlash: Child Protection Under Fire (Sage 1994).

15 David Finkelhor, ‘The “Backlash” and the Future of Child Abuse Advocacy’ in Myers (n 14).

16 April L Girard, ‘Backlash or Equality? The Influence of Men’s and Women’s Rights Discourses on Domestic Violence Legislation in Ontario’ (2009) 15 Violence Against Women 5.

17 World Health Organization, Preventing Intimate Partner and Sexual Violence Against Women: Taking Action and Generating Evidence (WHO 2010).

18 Rashida Manjoo, Report of the Special Rapporteur on Violence Against Women, its Causes and Consequences (United Nations, 2011).

19 Jane Anderson, ‘The Impact of Family Structure on the Health of Children: Effects of Divorce’ (2014) 81 Linacre Quarterly 378.

20 MA Straus, ‘Gender Symmetry and Mutuality in Perpetration of Clinical-Level Partner Violence: Empirical Evidence and Implications for Prevention and Treatment’ (2011) 16 Aggression and Violent Behavior 279.

21 Nico Trocmé and Nicholas Bala, ‘False Allegations of Abuse and Neglect when Parents Separate’ (2005) 29 Child Abuse and Neglect 1333.

22 Carrie Leonetti, ‘Endangered by Junk Science: How the Family Court’s Admission of Unreliable Expert Evidence Endangers Victims of Family Violence’ (2023) 43 Children’s Rights Law Journal 17.

23 Lenore EA Walker, ‘Nonjudicial Influence on Family Violence Court Cases’ (2020) 64 American Behavioral Science 1749, 1756.

24 ibid 1756.

25 Leonetti, ‘Endangered’ (n 22) 20–21.

26 Molly Dragiewicz, Equality with a Vengeance (Northeastern University Press 2011).

27 Carrie Leonetti, ‘Sub Silentio Alienation: Deceptive Language, Implicit Associations, Cognitive Biases, and Barriers to Reform’ 62 Washburn Law Review (forthcoming 2023).

28 Leonetti, ‘Endangered’ (n 22) 21, 43.

29 ibid.

30 Joan S Meier, ‘US Child Custody Outcomes in Cases Involving Parental Alienation and Abuse Allegations: What Do the Data Show?’ (2020) 42 Journal of Social Welfare and Family Law 92.

31 706 NYS2d 611 (NY Sup Ct 2000).

32 ibid 613.

33 Leonetti, ‘Endangered’ (n 22) 18.

34 Ellis v R [2022] NZSC 115 (‘Ellis Exoneration’) [28]–[30].

35 ibid [33].

36 R v Ellis (1994) CRNZ 172 (‘Ellis Appeal’) 174.

37 ibid 175.

38 ibid 191.

39 Ellis Exoneration (n 34) [219].

40 ibid [139].

41 Ellis v R, SC 49/2019, Criminal Appeal [Submissions], 12–14 October 2021, [2021] NZSC Trans 17 (‘Ellis Submissions’) 96.

42 Roland C Summit, ‘The Child Sexual Abuse Accommodation Syndrome’ (1983) 7 Child Abuse and Neglect 177.

43 ibid 179.

44 Roland C Summit, ‘Abuse of the Child Sexual Abuse Accommodation Syndrome’ (1992) 1 Journal of Child Sexual Abuse 153, 154–56.

45 Summit, ‘CSAAS’ (n 42) 190.

46 ibid 179.

47 ibid 189.

48 ibid 179–80.

49 ibid 180.

50 Summit, ‘Abuse of CSAAS’ (n 44) 154.

51 ibid 153.

52 ibid 156.

53 ibid.

54 ibid 157.

55 ibid 158.

56 ibid 157.

57 ibid.

58 ibid 159–60.

59 Ellis Exoneration (n 34) [146], [217].

60 ibid [1], [26], [59].

61 Ellis Appeal (n 36) 187.

62 Ellis v R [2019] NZSC 83 (‘Ellis Leave’) [50 (viii)].

63 ibid.

64 ibid.

65 ibid.

66 Ellis Exoneration (n 34) [156].

67 Tony Deverson and Graeme Kennedy (eds), The New Zealand Oxford Dictionary (OUP 2005).

68 NZLC/Te Aka Matua o te Ture, Issues Paper 42: Second Review of the Evidence Act 2006 (Wellington, March 2018) 184 n 1107.

69 EA 1908, s 23G (repealed by EA 2006, s 215).

70 EA 1908, s 23G (2) (c) (repealed by EA 2006, s 215).

71 NZLC, Evidence: Report 55 – Volume 2: Evidence Code and Commentary (August 1999).

72 ibid 67–69.

73 NZLC, Issues Paper 42 (n 68) 185 n 1108.

74 Cabinet Policy Committee ‘Evidence Bill (Paper 5: Spousal Witness Immunity and Children’s Evidence)’ (18 March 2003) POL Min (03) 8/15 [40].

75 Ellis Submissions (n 41) 252 (Winkelman CJ).

76 ibid 265 (Billington QC).

77 ibid 272.

78 ibid 273.

79 ibid 273, 306 (Billington QC).

80 ibid 275 (Billington QC).

81 Leonetti, ‘Endangered’ (n 22) 40–42.

82 ibid 315 (Glazebrook J).

83 ibid 309 (Williams J).

84 (1987) 2 FRNZ 697.

85 ibid 699–701.

86 ibid 698.

87 ibid 701.

88 ibid 703.

89 Evidence Amendment Act, 1989 No 104, s 3.

90 [1989] 1 NZLR 714.

91 ibid 715.

92 ibid 717.

93 ibid.

94 ibid 718.

95 ibid 720.

96 ibid 720.

97 ibid 721.

98 CA 244/91, 20 December 1991.

99 ibid 11.

100 ibid 12–13.

101 ibid.

102 ibid 16.

103 ibid 14.

104 ibid 19.

105 [2003] 2 NZLR 777 (CA).

106 ibid [6].

107 ibid [11]–[12].

108 ibid [12].

109 ibid [25].

110 ibid [26].

111 ibid [29].

112 ibid.

113 CA 414/03, 26 October 2004.

114 ibid [25].

115 ibid [26].

116 ibid [27].

117 ibid [5].

118 ibid [49].

119 509 US 579 (1993).

120 Lundy v R [2018] NZCA 410 [241].

121 [2009] NZCA 250 (R v A I).

122 R v A [2010] NZCA 124 (R v A II) [5].

123 R v A I (n 121) [23].

124 ibid [32]–[33].

125 ibid [39].

126 697 SW2d 139 (Ky 1985).

127 ibid 140–41.

128 ibid 141.

129 734 SW2d 457 (Ky 1987).

130 ibid 457.

131 ibid 458.

132 777 SW2d 930 (Ky 1989).

133 ibid 932.

134 ibid.

135 ibid.

136 ibid.

137 812 SW2d 502 (Ky 1991).

138 ibid 503–504.

139 ibid 504.

140 ibid.

141 825 SW2d 612 (Ky 1992).

142 ibid 613–14.

143 ibid 613.

144 ibid 614.

145 ibid.

146 937 SW2d 690 (Ky 1996).

147 ibid 691.

148 ibid 693.

149 ibid 694.

150 ibid 690–91.

151 ibid 695.

152 ibid.

153 2006 WL 2707451.

154 ibid *1.

155 ibid *5.

156 ibid.

157 ibid.

158 2008 WL 4291541.

159 ibid *1.

160 ibid *2.

161 ibid.

162 2008 WL 4692362.

163 ibid *4.

164 ibid.

165 ibid *5.

166 291 SW3d 610 (Ky 2009).

167 ibid 612.

168 ibid 612–14.

169 2012 WL 3637159.

170 ibid *9.

171 ibid.

172 392 SW3d 393 (Ky 2013).

173 ibid 397–98.

174 ibid 395.

175 ibid 397.

176 472 SW3d 523 (Ky 2015).

177 ibid 527.

178 ibid 528.

179 2020 WL 2831939.

180 ibid *2.

181 ibid *6.

182 862 SW2d 321 (Ky 1993).

183 ibid 322.

184 Alexander (n 158) *2–*3.

185 Vires (n 162) *5.

186 Sanderson (n 166) 612.

187 Blount (n 172) 395.

188 ibid 397.

189 King (n 176) 530.

190 Ellis Exoneration (n 34) [194]–[195].

191 ibid [160].

192 ibid [168].

193 ibid [174].

194 ibid [172], [196].

195 ibid [184].

196 ibid [11].

197 ibid [11] n 11.

198 ibid [184].

199 ibid [11].

200 ibid [125].

201 ibid [182].

202 Leonetti, ‘Endangered’ (n 22).

203 ibid 30–31.

204 ibid 31.

205 ibid 43–44.

206 American Psychological Association, Violence and the Family: Report of the American Psychological Association Presidential Task Force on Violence and the Family (1996) (‘Although there are no data to support the phenomenon called parental alienation syndrome, in which mothers are blamed for interfering with their children’s attachment to their fathers, the term is still used by some evaluators and courts to discount children’s fears in hostile and psychologically abusive situations’).

207 National Council of Juvenile and Family Court Judges, ‘A Judicial Guide to Child Safety in Custody Cases’ 12 (2008) (‘Any testimony that a party to a custody case suffers from “parental alienation” should “be ruled inadmissible and/or stricken from the evaluation report”’).

208 ACP-UK, The Protection of the Public in the Family Courts, December 2021.

209 New Zealand Psychological Society, Submission on Behalf of the New Zealand Psychological Society on the Review of Family Violence Law, 18 September 2015 (‘It is a deep concern and a major threat to the safety of women and children that the New Zealand Family Court continues to apply the doctrine of Parental Alienation Syndrome, which has long been discredited in the United States, from where it originated. It is now accepted in the United States that there is no scientific evidential basis for Parental Alienation Syndrome. No research conducted in the United States has ever been able to produce valid evidence of Parental Alienation Syndrome’).

210 J Mercer and Margaret Drew, Challenging Parental Alienation: New Directions for Professionals and Parents (Routledge, New York 2022) 34–35.

211 Maurice Berger, ‘Le Refus de Contact d’un Enfant avec un Parent dans un Contexte de Divorce Conflictuel – Partie I’ (2021) 69 Neuropsychiatrie de l’Enfance et de l’Adolescence 32, 33; Paul Fink, ‘Parental Alienation Syndrome’, in Mo Therese Hannah & Barry Goldstein, eds, Domestic Violence, Abuse, and Child Custody (2010).

212 Berger (n 211) 32, 34.

213 WHO, ‘Parental Alienation’ <https://www.who.int/standards/classifications/frequently-asked-questions/parental-alienation> accessed 15 January 2023.

214 ibid.

215 ibid.

216 R v B I (n 84) 698.

217 Richard A Gardner, The Parental Alienation Syndrome and the Differentiation Between Fabricated and Genuine Child Sex Abuse Cases (Creative Therapeutics 1987).

218 R v B I (n 84) 706.

219 ibid.

220 CA 244/91, 20 December 1991.

221 ibid 19.

222 Leonetti, ‘Endangered’ (n 22).

Appendix 1

Table

	CSAAS (Ellis)	PAS (family court)
Stimulus	Sexual abuse	Parental alienation
Response	secrecy, withdrawal, estrangement^a^bcdef helplessness^a ambivalence^b fluctuating anger and affection^b^d soiling and bed wetting^ghij insomnia and nightmares^d^fghk tantrums, lashing out^a^dghl headaches and stomach problems^f^g anxiety, exhaustion, depression^c^fghm guilt, shame, embarrassment^b^ce sexual preoccupation^g^hm imaginary friends^a substance use^a self-harm^a^c academic over-achievement^a decline in school performance^c	disclosing abuse^n^opy fear of or hostility to one parent^o^pqrsz ambivalence^s^u fluctuating anger and affection^p^rsvw similar thought/language to parent^p^rs wanting to live with one parent^q^rsx# defiance, acting out, anti-authority^r^s spying and eavesdropping^r anxiety and depression^s^t rigid world view^r

^aSummit ‘CSAAS’.

^bR v B I.

^cR v Accused.

^dR v B II.

^eBussey v Commonwealth.

^fHellstrom v Commonwealth.

^gEllis Appeal.

^hG v R.

ⁱR v A.

^jSallee v Commonwealth.

^kVires v Commonwealth.

^lBrown v Commonwealth.

^mAlexander v Commonwealth.

ⁿC v B [2018] NZCA 322.

^oRMJ v BJG [2017] NZHC 1159.

^pDaviau v Desrosiers [2015] NZFC 7922.

^qLowe v Way [2017] NZFC 6365.

^rFinn v Poole [2015] NZHC 1362.

^sJF v DF 112 NYS3d 438 (2018).

^tBush v Johnson [2019] NZHC 186.

^uMann v Armstrong [2020] NZFC 1319.

^vBaker v Everill [2015] NZFC 1036.

^wJ v M (2004) 24 FRNZ 1019.

^xHenderson v Morgan [2013] NZHC 1010.

^yPip Cobcroft and Catherine Townsend (2022).

^zApril Trenberth, ‘A Brief Overview’ (2022).

^#Sarah Calvert, ‘Talking to Kids’ (2022).