feminists@law, Vol 9, No 2 (2019)

Interpreting the Feminine in the Criminal Trial: Can the Insights of Rape Myth Scholarship Help Mothers Accused of Killing Their Children?

Abstract

In cases where mothers were wrongfully convicted of killing their children, both forensic and non-forensic evidence was admitted. Although the expert opinions and evidence were subsequently robustly scrutinised, the same is not true for informal evidence of maternal behaviour. The paper proposes that if we consider an analogous area of criminal justice which has seen interpretations of the feminine strongly challenged, such as in rape trials, then we might learn from rape myth scholarship how better to analyse child death cases. The article explores the difficult issues in rape myth scholarship in identifying what a rape myth is, how widely it is held, and how complex layers of functionality and connections constitute belief systems. By focussing on behavioural normativity and the deployment of fixed beliefs the article proposes a device based on the insights of rape myth scholarship with which to interrogate the behaviour evidence admitted in child death cases. Using the concept of modern mothering myths may prevent the possibility of background evidence being used as a vehicle for smuggling in prejudicial material of little probative value.

Introduction

When mothers stand trial for killing their children, the body of evidence presented has to be considered as a whole.  Given the array of its constituent parts it may be difficult to know why precisely a woman has been found guilty. In seeking better to understand how particular miscarriages of justice occurred, my research has explored a number of cases involving mothers accused of killing their children, including those of Sally Clark, Angela Cannings, Angela Gay and Donna Anthony.(1) These cases were notably characterised by expert evidence and expert witnesses later shown to be unsound and unreliable respectively, and inevitably the search for reasons to explain the wrongful convictions have concentrated on the expert evidence and its handling by the criminal justice system (CJS).(2)

But, as feminist scholar Celia Wells commented, there was ‘no attempt…to ask why it was so easy to leap to conclusions that now seem so wrong’.(3) She submitted that ‘feminist arguments’(4) could help ‘unravel these questions’.(5) Fiona Raitt and research psychologist Suzanne Zeedyk further proposed that ‘hidden factors’(6) such as ‘underlying assumptions’(7) and ‘discourses of motherhood’(8) may have ‘played a major role in the initial convictions of Cannings and Clark’.(9) Emma Cunliffe too suggested that the mechanism for achieving a mother’s (wrongful) conviction in Clark and Cannings, drew on discourses of motherhood within legal discourses.(10) Difficulties arise from such propositions however, because they are sizeable claims: for example that, in the absence of conclusive expert opinion, advocates were forced to explain children’s deaths in Clark and Cannings by filling the evidential gaps using ‘social expectations’ about ‘proper mothering’;(11) or that where the diagnostic techniques of child mistreatment are uncertain, mothers may be criminalised by deliberately portraying them as failing to conform to the dominant ideology of motherhood.(12) Such claims are, as Laura Hoyano rightly points out, ‘very large and diffuse’.(13) It is possible, that the difficulties arise because ‘hidden factors’,(14) ‘underlying assumptions’(15) and ‘discourses of mothering’(16) are as yet insufficiently defined and require further analysis in order to clarify how they are deployed by agents of the CJS in portraying maternal behaviour.

As both Hunter(17) and Redmayne(18) suggest a technically complex evidential context may hide the true operation of informal background or character evidence in practice(19) and in examining the wrongful conviction cases, it is evident that informal non-medical(20) and non-forensic material(21) such as maternal behaviour and childcare(22) was freely admitted to these trials. When further child death cases were examined where female childminders and baby-sitters were convicted,(23) where mothers remain incarcerated,(24) and where mothers were acquitted(25) or their conviction was reduced from murder to manslaughter,(26) information about maternal behaviour and childcare was admitted in all cases apart from two.(27) Consequently, it appears that information about a woman’s behaviour is normatively admitted where women are accused of killing their children. The presence of such material as part of a body of evidence that includes highly complex, contested and controversial expert opinion, supports the concerns raised by feminist commentators that portrayals of maternal behaviour may have influenced trial outcomes.(28) Thus, in particular miscarriages of justice, not only may overzealous expert opinions and flawed scientific information bolster the prospect of injustice, but interpretations of the feminine also.

Prima facie, admission of non-medical and non-forensic behaviour information at trial seems sensible. At the law-science interface however, jurors face unenviable decision making responsibilities based on contested, controversial or scant expert opinions. In such situations, behaviour evidence may become more significant as jurors may be tempted to rely on fixed beliefs.(29) If such material is then stereotypically(30) or even prejudicially(31) interpreted, then constructions, assumptions and expectations of the feminine may contribute to unsafe convictions.

In contrast to the exhaustive investigations into the use of expert opinion and scientific evidence in criminal prosecutions,(32) scant attention has been paid in the literature to the impact on criminal proceedings of admitting evidence of maternal behaviour and child care in cases where mothers are accused of killing their children. Nor is there robust research demonstrating that past female or maternal behaviours and actions before and around the time a child dies lead to reliable inferences. Consequently, it is possible that evidence of maternal behaviour generally discounted as holding merely informal, peripheral or contextual value may be influential when admitted into criminal proceedings and may need to be considered more carefully in relation to its use or even its admissibility.  

In order to explore whether and how evidence of maternal behaviour might influence trial outcomes, a comparable situation within the criminal justice system in which information about women’s conduct and personal history is admitted as informal background evidence has been sought. Two key topic areas were identified: the first when abused women have killed their husbands,(33) and the second when women are claimants in rape cases.(34) In each area feminist scholarship and judicial commentary indicates that adverse interpretations of female behaviour may occur at trial and may influence outcomes.(35) Significantly, rape myth scholars have argued that the use of stereotypical assumptions and/or the device of rape myths to interpret evidence of female behaviour, results in damaging inferences and adverse outcomes in criminal proceedings.(36) For reasons of space only one area can be considered here,  although both are considered  at length elsewhere.(37)

Rape myth scholarship (RMS) has therefore been explored to learn whether using the rape myth as an analogous device to analyse cases, perhaps by identifying mothering myths, may be helpful where mothers are accused of killing their children. Although several cases have been analysed, for reasons of space two key cases are used as illustrations, and the following brief overview shows how foregrounded expert evidence may be interwoven with informal background behaviour evidence.  

Following the consecutive deaths of Sally Clark’s  first two infants, Christopher and Harry, suspicions were raised and she was later found guilty of murdering both.(38) Her first appeal, dismissed by Henry LJ,(39) was followed by a second successful appeal granted by Kay LJ.(40) At trial, the prosecution, using flawed statistical(41) and later discredited medical opinion,(42) had argued that Clark had smothered her sons to death. The successful appeal relied on fresh evidence in the form of microbiology test results not disclosed at trial indicating that a cerebro-spinal fluid infection may have been the cause of Harry’s death.  

The way Clark’s behaviour was portrayed at trial is noteworthy. Despite being described positively  as a ‘normal, happy, caring mother’,(43) she was also characterised as a woman who resented being left alone and, who ‘tended to drink more heavily when her husband was away’.(44) Such representations may have been true, but were also prejudicial to Clark’s credibility because of the way in which mothers dependent on alcohol may be judged.(45) At trial and at first appeal Clark’s alcohol dependency was given greater evidential weight than in the second appeal judgment. Clark indicates that it is not only medical opinions that may bear contrary interpretations at different stages within a child death case, but evidence of maternal behaviour may at first instance and first appeal be perceived as highly relevant but, at a later appeal, may be considered of no probative value.

The way Sally Clark’s behaviour was portrayed, and possibly perceived,  was echoed in Cannings. Cannings had four children, three of whom died suddenly in infancy.(46) Tried for the murder of two of her children by smothering, the prosecution used circumstantial evidence and medical opinion to argue that Cannings had murdered two of her three children. She was convicted, but acquitted two years later because medical opinion suggesting that the rarity of three infant deaths in one family was evidence of murder,(47) was unsafe.(48) In addition to medical opinion the appeal report records information about maternal behaviour and childcare. For example, ‘There was no suggestion of ill-temper, inappropriate behaviour, ill-treatment, let alone violence, at any time with any one of the four children’.(49) Her behaviour as a mother was apparently exemplary; Cannings was depicted as a ‘woman of good character, described as a loving mother’.(50) Health visitors reported that she and her husband had always cared for their children properly,(51) and that Cannings had bonded with her daughter Jade, who ‘seemed to be a well-cared for and loved baby’.(52)

Her children had unexplained health difficulties, however, and at trial, prosecution counsel suggested that Cannings had smothered one of her sons ‘in an attempt to evoke sympathy’.(53) By suggesting that Cannings was mentally ill, prosecution counsel may have sought to reduce her credibility by alluding to the syndrome Munchausen Syndrome by Proxy (MSbP)(54) and suggesting that at the very least, Cannings had something wrong with her as an attention-seeking mother. Once fresh medical opinion was accepted that three sudden infant deaths in one family could occur naturally,(55) prosecution arguments included in the judicial summing up(56) seeking to syndromise or portray Cannings as a mentally ill mother, were no longer relevant.

The two brief case summaries show that both worthy and questionable maternal behaviour may be admitted and such information is unlikely to be ruled inadmissible unless there is a ‘risk of jury irrationality’.(57) Overall, the courts’ approach is to place as much relevant, or ‘more or less relevant’(58) information before a jury, and to rely on judicial directions,(59) together with ‘judicial warnings and common sense to ensure that it is properly evaluated’.(60) The admission of informal, extraneous female behaviour evidence has long been a feature of sexual assault and rape prosecutions.(61) The relevance of certain behaviours such as previous sexual history have been strongly challenged by rape myth scholars to the extent that this material may now not be admitted without leave.(62) Nevertheless, other types of claimant information may be sought by defence teams such as medical, mental health and social welfare records,(63) archived internet searches(64) and more recently mobile phone content(65) and social media(66) with text and image content.(67)

For women accused of killing the children in their care therefore, any past behaviour information including internet searches, medical and mental health or social welfare reports and counselling records may be freely admissible. Such a position echoes that of rape complainants for whom feminist and legal commentary have identified the dangers of a permissive approach to the admission of behaviour evidence, because some female behaviours(68) may be readily misinterpreted as rape myths.

The following section will examine how a rape myth has come to be conceptualised in RMS and how this concept may be used in understanding what may have been happening in rape trials. RMS has studied the admission of evidence of female behaviours and sought to show how the rape myth device works in calling out prejudicial and stereotypical beliefs. By analogy, this paper suggests that in cases of women accused of killing their children, damaging but unfounded inferences may be drawn from maternal behaviour admitted at trial, and proposes that a similar device, the mothering myth, may help in identifying unfounded and prejudicial beliefs about mothers.

There are some methodological difficulties to consider. First, the totality of RMS cannot be represented here and so only key developments in the search for definition are examined. Secondly, the analogy between rape trials and child death cases is complicated by the fact that in rape trials the woman is the complainant and in child death cases she is the defendant; in rape trials the woman has alleged that the defendant should be convicted while in child death cases the mother as defendant seeks to be acquitted. The divergences in adversarial attitude and carceral intention are put to one side in this paper, to focus on how rape myth scholars have approached the issue of defining a rape myth, and how their work might enable us better to analyse child death cases. 

Rape Myth Scholarship

The issues raised by unjust interpretations of female rape complainants’ behaviour are considerable and RMS has challenged the relevance of and constructions of behaviour evidence in rape and sexual assault trials relating to sexual history,(69) dress, alcohol and drug use, time of reporting,(70) and more recently, personal records.(71) Complainant behaviours have been subject to interpretation using normative expectations about women’s behaviour and the circumstances of alleged rape, in order to reach conclusions on issues of claimant consent and credibility.(72) Unjust acquittals have been argued to have occurred because a complainant’s credibility has been unfairly devalued as a result of the admission of information about her conduct, and its adverse interpretation.(73) Ellison has suggested that the admission of information about the complainant’s behaviour has been a function of masculine judicial perspectives on questions of relevance,(74) which may ‘at best, risk…the undervaluing of women’s experience and interests’.(75) The consequence is, as Temkin and Krahé have argued, that unjust acquittals create a justice gap in the CJS generated by rape myths.(76) But, to prove that rape complaint attrition and unjust acquittals in rape trials are due to myths – ‘widely held but false beliefs or ideas’(77) – has been problematical. It is likely that the same may hold true if without more, we were to say that wrongful convictions in child death cases were caused by hidden factors, underlying assumptions, discourses of mothering or even mothering myths.

Issues in definition, linkage and functionality of ideas about rape.

The definitional difficulties encountered in RMS have been considerable; they include determining functionality,(78) achieving agreement on who believes rape myths(79) and how many people accept them; whether they function as feared, i.e. leading to adverse decision making; and if they even exist.(80)  Burt first coined the  term rape myths to describe ‘prejudicial, stereotyped, or false beliefs about rape, rape victims and rapists’.(81) These beliefs included: ‘“only bad girls get raped”; “any healthy woman can resist a rapist if she really wants to”; “women ask for it”; “women ‘cry rape’ only when they've been jilted or have something to cover up”; “rapists are sex-starved, insane, or both”’.(82) Burt identified that such beliefs were linked to beliefs justifying sexual violence.(83) In seeking to define rape myths, Burt conducted a regression analysis of interview data to question whether ‘settled ways of thinking or feeling’ about rape or rape attitudes, were ‘strongly connected to other deeply held and pervasive viewpoints such as sex role stereotyping, distrust of the opposite sex (adversarial sexual beliefs), and acceptance of interpersonal violence’.(84) Burt’s pioneering work in developing Rape Myth Acceptance (RMA) scales demonstrated the broad extent to which such beliefs were held and that they fulfilled cultural functions by endorsing other gendered ideas.

Building on Burt’s foundational work, Lonsway and Fitzgerald reviewed the literature and also found difficulty in clearly defining rape myths, because of belief interconnections and unexpected functionality for those making decisions in criminal trials.(85) RMA they found, had at its core gender, traditional sex role attitudes, negative attitudes towards women, and a likelihood of raping.(86) They concluded, that ‘Such a configuration conveys a powerful message about how RMA relates to other beliefs about women in our society’.(87) Lonsway and Fitzgerald proposed that rape myths could be characterised as ‘false or apocryphal beliefs that are widely held’; they explain some important cultural phenomenon; and they serve to justify existing cultural arrangements’.(88) Nevertheless, because of the identification of the interrelatedness with sexist attitudes, defining a rape myth was difficult because of the ‘lack of any comprehensive articulation of the domain of rape myths’.(89) Questionnaire scales seeking to measure RMA were, Lonsway and Fitzgerald suggested, unreliable because different studies had used different scales of questions to identify acceptance (or not) of particular beliefs.(90) Lonsway and Fitzgerald sought to take into account the arguments for interconnectedness and functionality, proposing that ‘Rape myths are attitudes and beliefs that are generally false but are widely and persistently held, and that serve to deny and justify male sexual aggression against women’.(91) Their findings echo Burt’s in that rape myths could be expressed as rape supportive beliefs.(92)

Gerger et al. drew on the insights of previous studies in a later review of RMA literature seeking to  define beliefs as rape myths and to measure RMA.(93) Despite agreement with the general usefulness of an RMA measurement or construct, this study concurred with Lonsway and Fitzgerald that most scales produced skewed results as a result of methodological approaches.(94)  Gerger et al. developed a new questionnaire they denoted the Acceptance of Modern Myths about Sexual Aggression scale (AMMSA).(95) The reasoning behind the creation of AMMSA was to take account of more recent research for example by Swim et al. into modern sexism and racism,(96) which argued for a greater degree of subtlety about scale question content areas.(97) This they suggested was because of widespread and institutional denials that discrimination against women, their needs and demands still occurred.(98) Gerger et al. concluded that the AMMSA scale incorporating more nuanced versions or modern myths, was found to be both reliable and consistent.(99) But they noted that their own studies were also limited by sampling methodologies.(100)

When rape myth functionality was explored by Gerger et al., they concurred with Burt’s work on cultural functionality(101) that RMA may have meaning for both women as well as men, whereby RMA ‘allows women to reduce their subjective vulnerability to sexual assault and protect their self-esteem’.(102) So, some women believe that if they do not dress revealingly in public and do not go out alone late at night they will not be assaulted. That of course is a false belief, but functionally the belief serves a purpose in (wrongly) allaying female fears(103) and providing a platform for some to judge others, and is therefore a rape myth.  

Controversially however, arguments for the existence of rape myths have been considered to be overstated(104) as an explanation for the ‘justice gap’ in rape cases.(105) Reece in particular criticised Gerger et al.’s work by arguing that ‘some attitudes are not myths…not all the myths are about rape and there is little evidence that the rape myths are widespread’.(106)  Her paper and later public lecture questioning the existence and nature of rape myths and how they are identified and measured prompted heated debate.(107) I cannot do justice here either to Reece or her critics. My point is that the progress of RMS in seeking to demonstrate that particular beliefs are myths is not straightforward. RMA or AMMSA may be demonstrable to some extent but beliefs may change in content, become more nuanced, be connected with other gendered ideas and be functional for some, and true for others.(108)

It is also possible that the difficulties encountered are caused by the language we use to define the beliefs that many find so wrong and that the word myth itself creates problems. RMS has more recently adopted the word attitude, ‘a settled way of thinking or feeling about something’,(109) as in: ‘rape supportive attitudes’.(110) The word is acquiring greater purchase perhaps because it is less hostile than the word myth can be. Indeed Burt suggested rape supportive attitudes were connected to other gendered and questionable attitudes,(111) as did Lonsway and Fitzgerald.(112) Temkin and Krahé suggest the issues of the justice gap relate to a ‘question of attitude’,(113) and Gerger et al. also measured attitudes towards rape using the AMMSA scale, and found positive correlations with ‘pervasive cultural attitudes related to gender and violence as well as sexual harassment’.(114) Reece too echoes Burt that rape myths are rape supportive attitudes,(115) but again she questions whether ‘public attitudes …  deserve to be described as “rape supportive attitudes”, or warrant the label of “rape myths”’.(116) For Reece, RMS indicates that rape supportive attitudes may not be widespread, so therefore they are not myths. In addition, some attitudes are not proven to be false so again, they are not myths, and some ‘rape supportive attitudes’ are not about rape. Reece therefore suggests the possibility that scholars are creating myths about myths,(117) and further that Gerger et al.’s work was about ‘designing a scale to catch people (out)’,(118) and to show how awful people’s attitudes are.(119)

Conaghan and Russell have strongly defended the work of Gerger et al., critiquing the arguments put forward by Reece and asserting that the AAMSA is ‘regarded as the most advanced measure of rape myth acceptance to date’.(120) Far from seeking to ‘catch people (out)’(121) Gerger et al. sought to demonstrate that prejudicial views were just more subtly held than previous RMA scales had identified, and that some beliefs about rape were simply ‘immune against empirical falsification’.(122) For example, ‘Many women secretly desire to be raped’(123) cannot be proved or disproved and therefore showing that such a belief is a myth, is difficult if not impossible.(124) Gerger et al. had suggested that rape myths should not be defined in terms of being false, but ‘“wrong” in an ethical sense’,(125) acknowledging the possibility that some rape myths may be true for some people. But, even if  ‘the factual configurations comprising rape myths may on occasion be true’ as Conaghan and Russell accept,(126) RMS has become more concerned with addressing ideas ‘treated as generalizable truths which function normatively to shape perceptions and inform judgment’.(127)

Consequently, what RMS shows us is that there are challenges in defining beliefs about female behaviours in relation to sexual assault and rape as myths. Using either RMA or AMMSA is methodologically problematic. It is difficult to say or reveal exactly which beliefs these refer to, to prove how widespread they are, to differentiate them from other beliefs, and to unpick them from other beliefs to which they are connected. They are difficult to falsify, some beliefs that scholars designate as myths may be true for some people(128) which is both problematic and provocative, and some beliefs are held covertly.(129) Considerable difficulties therefore exist in defining rape myths, and distinguishing between beliefs, stereotypes and attitudes. The word ‘myth’ may function in this context by signalling that particular attitudes are questionable, if not ethically wrong, but using the term ‘myth’ may be counter-productive because of the problems in proving that such attitudes are false. Further, if a belief’s status as myth depended on the number of people who held it, then if that number reduced it would no longer be definable as a myth, irrespective of its perceived problematic content.(130) Including extent of acceptance within our understanding therefore, does not add to conceptual clarity and may even not be necessary, with scholars suggesting that ‘prevalence and consistency of rape myths…seem to be better treated as empirical problems, rather than matters of definition’.(131) The force of rape myth arguments is therefore bound up with difficulties of language. What is more, Lonsway and Fitzgerald consider that rape myths are best conceptualised as stereotypes,(132) and Conaghan and Russell suggest that the term rape supportive attitudes, is interchangeable with myths.(133)

Language and normativity

It is I suggest Conaghan and Russell’s focus on or ‘turn to attitudes’(134) which may be most helpful. To cut through the difficulties, RMS language is moving towards words such as ‘norms’ about behaviour and the suggestion that a key issue in understanding rape myths is not whether they are true or false, but whether they are ‘normatively infused’.(135) How this necessary but troubling discussion of the RMS debate might help us in analysing child death cases is through appreciating that we need to be concerned with our approach to language and normativity and in order to move forward, should consider consolidating the language used. Significantly, Gerger et al. recognised that beliefs about women were ‘prescriptive in nature’,(136) and it is this aspect that is central not only to the content of a modern rape myth as we now perceive it, but to the way in which we can define beliefs about mothers and the characterisation perhaps of a modern mothering myth.  

Gerger et al. finally proposed that ‘rape myths are descriptive or prescriptive beliefs about rape (i.e., about its causes, context, consequences, perpetrators, victims, and their interaction) that serve to deny, downplay or justify sexual violence that men commit against women’.(137) By characterising rape myths in this way, the focus is on the consequences of the belief, which takes the heat out of all the problematical difficulties of whether the beliefs are true or false or widespread, etc.  What this may then mean for mothers accused of killing their children is that the language we use matters, and the difficulties faced by RMS are best avoided.

If we wished to suggest that perhaps myths, expectations, assumptions, or even ideologies of mothering were implicated in contributing to wrongful convictions, then first we would need to define all the beliefs about mothering, determine their prevalence, and reveal their functionality, interconnectedness and whether they were held covertly. We would need to establish empirically that some beliefs are simply false and although that may be possible for particular beliefs about maternal behaviour, for example, that mothers should always stay at home and look after their children themselves until school age, they may be true for some people. What I suggest we can use in the meanwhile as a device to analyse child death cases is the definition of a rape myth as proposed by Gerger et al. applied to mothering, to produce the modern mothering myth. A proposed definition of a modern mothering stereotype or myth (MMM), is accordingly: a fixed belief about mothering or normative behaviour for mothers, which serves to support or justify adverse decisions about mothers within the criminal justice system.

The following section considers maternal behaviours identified in child death cases,(138) and tests the definition as a device, to assess whether it helps us analyse if maternal behaviours admitted as evidence may have been probative, or myths.

Applying the Modern Mothering Myth

Is alcohol abuse probative or a mothering myth?

Drinking alcohol has long been regarded as harmful especially if carried out during pregnancy.(139) Health promotion programmes and research studies by public organisations such as the Royal College of Paediatricians and Child Health (RCPCH),(140) have sought to publicise the dangers and disseminate normative understanding that maternal alcohol dependency harms children. Drinking excess alcohol in pregnancy may harm the unborn child, ‘resulting in “foetal alcohol syndrome”, (FAS) and “foetal alcohol effects (FAE)”’, characterised by growth retardation and central nervous system impairment.(141) The combination of alcohol and certain cardiac arrhythmias such as Long QT Syndrome (LQTS) may also be fatal in infants and adults because ‘alcohol abuse is associated with an increased incidence of cardiac arrhythmias’.(142) Clark(143) had an alcohol dependency for which she had received treatment prior to her pregnancies.(144) However the facts of whether she drank in pregnancy or as a mother are not known, but were inferred by the prosecution. But for her previous dependency, there is little doubt that her defence could have argued that she was unquestionably a person of good character as evidenced by her health visitor,(145) GP(146) and nanny.(147)

As reported in the first appeal against conviction, Clark’s health visitor had observed a close attachment to and bond between Clark and her first ‘responsive’ baby Christopher.(148) Clark attended a mother and baby group ‘where she appeared as a normal, happy, caring mother’,(149) her babies ‘were well cared for, loved by their parents and happy and content’(150) and she was reported as a ‘loving, caring mother’.(151) She is reported as welcoming visits from health visitors as part of the CONI programme,(152) indicating she was a responsible mother of a next infant. Prior to the deaths of her children, health professionals praised her as exemplifying the ideal of the good mother, by her caring, nurturing and compliant behaviour.

In contrast to Clark’s behaviour as a mother that satisfied normative expectations of mothering, counsel for the prosecution told the court that on the day of her second son Harry’s death, she ‘visited the off-licence on two occasions to buy some wine saying (falsely, it would appear) that they were having a dinner party that evening’.(153) The implications of such a fact may have raised concerns in jurors’ minds (although the CPS could not use the fact of Clark’s alcohol dependency in argument due to a pre-trial ruling) and beliefs in the wrongness of maternal alcohol dependency may have served to support if not justify a guilty verdict. Such views were amply reinforced by media portrayals following the trial when Harrison J reversed his pre-trial ruling. The BBC described Clark as a ‘35-year-old lawyer who drank through both her pregnancies… a lonely drunk… a depressed alcoholic’(154)  who had received hospital treatment for ‘bouts of severe binge drinking’,(155) and she was even described by The Lawyer, as ‘driven by drink and despair, the solicitor who killed her babies’.(156)

Clark’s alcohol dependency was suggested at trial, confirmed after conviction, and affirmed in her first appeal where it was reported that she ‘tended to drink more heavily when her husband was away’.(157) A belief on the part of both jurors and judiciary that a mother may have abused her child is understandable if she purchases alcohol covertly on the day her second son dies suddenly and unexpectedly and she has received treatment for alcohol dependency in the past. However, holding such a belief in a fixed way within criminal proceedings to support a finding of guilt is at least questionable. The judgment in her second (successful) appeal did not mention alcohol at all in acquitting Clark.(158) Her dependence on alcohol and her possible consumption on the day her second child died, therefore did not justify either a belief or a decision that she was guilty of murder, nor her continued imprisonment.

The reasons for using Clark’s alcohol dependency as part of legal argument in criminal proceedings, although understandable, may reasonably be expected to be linked with fixed views about the harm that alcohol may cause to children both in pregnancy and when caring for a small child. As Ann Oakley reflected, ‘of all the things women are supposed to be, mothers come first’.(159) Helena Kennedy suggests that society expects women ‘to embody nurturance and protectiveness associated with mothering’,(160) and consequently when women are accused of harming their child, Kennedy suggests there is a ‘heightened outrage’.(161) Not only may such a mother be perceived as selfish, contravening expectations that she should forego self-centred behaviour in favour of altruism, such beliefs may induce prejudicial perceptions of the maternal behaviour.

That alcohol dependency is not problematic for any person including mothers, is not my position. But, I wish to suggest that raising the fact of Clark’s past alcohol dependency at trial and at first appeal, without evidence of her having drunk excess alcohol immediately prior to the children’s deaths, risked providing the jury with a behavioural cue and engaging a fixed belief (a MMM), that justified the view that a mother with a previous alcohol dependency was guilty of murder. 

Are emotional over-reactions probative or mothering myths?

The consequences of trauma whether through physical violence(162) or sexual assault(163) can result in significant ‘emotional disorganisation’(164) that may affect the behaviour of otherwise rational women. How such emotions including ‘fear, shock, disbelief, anger, self-blame and embarrassment’(165) may be expressed varies according to the individual. But such expressions may diverge from expected norms, and be perceived as bizarre or unexpected and interpreted according to prescriptive beliefs such as rape myths(166) for example, that have not allowed for the impact of trauma on behaviour.(167) Similarly in child death cases, the appeal reports provide evidence of mothers’ behaviour on the days their children died, and the following discussion identifies that this behaviour may have been interpreted according to fixed beliefs about how traumatically bereaved mothers should behave, leading to adverse trial outcomes.

Clark was on her own at home when she noticed that her son Christopher aged nearly three months was ‘a “dusty grey colour”(168) and she knew something was wrong. She picked him up and dialled 999’(169) and asked for an ambulance. There is no mention of whether she tried to resuscitate the baby. When the ambulance arrived only two minutes later according to the appeal report(170) the house was locked on the inside with Clark unable to find the keys. Paramedics entered the house after a ‘neighbour arrived with the spare keys’,(171) to find Clark holding the baby who was already ‘pale, cyanosed, cold and quite rigid’.(172) Clark’s behaviour is described at home, in the ambulance and at hospital; the ambulance driver stated Clark was ‘very distressed, crying and screaming’,(173) she was ‘on the verge of hysteria’(174) and was so distressed the paramedic could not put the child on the resuscitator.(175) On being told that Christopher was dead, Clark’s ‘reaction was described by a hospital doctor as very dramatic and hysterical’.(176) Further, the doctor branded the behaviour as ‘atypical and the over-reaction made her feel quite uncomfortable’.(177) In addition, a staff nurse stated that Clark had ‘said that her husband would blame her and would not love her any more’.(178) The evidence provided by professional witnesses suggests doubts that Clark’s grief was normal, indicating concern that Clark may have harmed Christopher.

Misgivings may have further increased because of discrepancies between Clark’s accounts to ambulance personnel and doctors concerning Christopher’s whereabouts when he died.(179) Clark stated that he was in a Moses basket to ambulance crews,(180) but in a bouncy chair to paediatricians.(181) When the police visited the home at 02.00 am on the night of the baby’s death they questioned the parents and removed both pieces of baby equipment,(182) having already noted on the coroner’s form that Christopher had been found in a bouncy chair.(183) Clark failed to later challenge that discrepancy, and the first appeal report states, ‘The fact that the appellant gave inconsistent accounts of where she found Christopher adds to its significance rather than detracting from it’,(184) as she was unable to remember whether the child died in a bouncy chair or the Moses basket.(185)

It is difficult to tell whether not remembering which place the child was in when discovered lifeless was probative, as the prosecution suggested or, whether Clark’s memory may have been impaired by the shock of Christopher’s death.  Ellison suggests that the impact of trauma in sexual assault cases may have a significant effect on memory. ‘Significantly, research suggests that the normal variability of memory can be exacerbated by the impact of trauma, such as that experienced by victims of sexual assault’.(186) It is therefore possible that the risk that Clark may have suffered post-traumatic stress disorder (PTSD) as a result of the death of her child, resulting in impaired memory, may have been overlooked by both the defence and judicial summing up. Such a shocking moment of discovery is unlikely to be forgotten. But the view that detailed memories are accurate and can be recalled indicates that evidence of Clark’s faulty memory may have been interpreted in accordance with a fixed belief that all mothers should remember the factual circumstances of a child’s death. Any inconsistencies therefore would lead to support for an adverse conclusion.(187)

Clark’s second child Harry also suddenly stopped breathing in the evening; she called the ambulance whilst her husband commenced resuscitation.(188) Again the first appeal court judgment records professional witness evidence about Clark’s behaviour. Paramedics said when they arrived, Clark was ‘running up and down the street outside the house, barefoot, in pyjamas and very distressed’;(189) and that she had behaved in a ‘“very dramatic and almost hysterical”’(190) manner, described as ‘“such an over-reaction”’.(191) To compound the concerns about Clark at the time of Harry’s death, she could not accurately recall to police in interviews at home the time that her husband had returned home on the night the second baby died, as she said she had confused the night the second child died with the night the first had died.(192) In addition, a few days later the coroner(193) said that Clark had stated ‘she and her husband would try for another baby’.(194) Mrs Hurst said she felt that comment ‘most unusual’,(195) and realised then that Clark had lost two babies. This observation led her to contact a senior police inspector and request a Home Office pathologist to conduct the post-mortem on the second baby Harry.(196)

Clark’s behaviour and comments were therefore appraised by professionals and an adverse interpretation was made that her behaviour was not normal. Whether Clark’s comments support an adverse interpretation is uncertain. Newly bereaved mothers must surely behave as individuals and not according to preconceived essentialised normative understandings. Nevertheless, a coroner is likely to have witnessed many bereaved parents and possibly sufficient to form a view that Clark’s behaviour was aberrant.  But such interpretations about behaviour based upon experience are not the same I suggest, as objective large scale research studies on bereavement behaviour, which are notably lacking in this area.(197) 

The points identified in this section about Clark’s behaviour are taken from Henry LJ’s judgment dismissing her first appeal. Whereas Clark’s hysteria and distress is mentioned twenty times by Henry LJ in his legal reasoning, in Kay LJ’s judgment of the second successful appeal,(198) each word is mentioned only once. It is possible that Henry LJ was persuaded that Clark’s behaviour around the time of her sons’ deaths was so abnormal it supported if not justified upholding her conviction and dismissing her appeal. However, in the judgement of her second appeal such factors were barely mentioned. One cannot know the extent to which any member of the court may have believed that the evidence of Clark’s overwhelming distress, confusion and inappropriate comments justified a guilty verdict. But it is clear that in Kay LJ’s judgment, such factors were of no relevance or weight. Accordingly, there is a question whether professionals’ fixed beliefs about normal behaviour of mothers confronted with a dying child supported if not justified a guilty verdict, without expert evidence from a psychiatrist to support such perceptions, and there is also a question as to why defence counsel failed to adequately challenge what may have been mothering myths.  

How a mother should behave following the death of a child may be impossible to state without over- simplification. However as Judge LJ suggested in Cannings,(199) if a fixed and over-simplified view is held that ‘lightning does not strike three times in the same place’(200) then however a mother behaved, ‘might be thought to confirm the conclusion that lightning could not indeed have struck three times’.(201) If the children’s deaths were natural then ‘virtually anything done by the mother on discovering such shattering and repeated disasters would be readily understandable as personal manifestations of profound natural shock and grief’.(202) Judge LJ suggests that maternal behaviour in Cannings was therefore adversely interpreted within the context of and as a result of flawed expert evidence,(203) and the same could be said of Clark.(204)The judicial comments indicate that prejudicial interpretations of maternal behaviour may be very persuasive, especially where expert evidence on the interpretation of pathology findings such as ‘petechial or pinpoint haemorrhages’,(205) and ‘intra-retinal haemorrhaging’(206) are complex and unfamiliar.

In Cannings also there is little evidence that the normative interpretation of maternal behaviour was robustly challenged by her defence.(207) The appeal judgment records that evidence of Cannings’ behaviour and emotional reactions when her young children experienced apparent or acute life threatening events (ALTE) was submitted in great detail,(208) together with evidence of her use of the apnoea alarm,(209) and who she called when she realised a child was in danger.(210)

Failing to use an apnoea monitor: indicative of guilt or a mothering myth?

Cannings was regarded by health professionals as a good mother and care-giver, with the appearance of an affectionate and caring mother.(211) She had four children of whom three died. She was charged with the murder of two. All three suffered from what were referred to as acute or apparent life threatening events (ALTE) where they apparently stopped breathing, and the appeal transcript identified ALTEs as SIDS in which no death had actually resulted.(212) Prosecution counsel argued that the ALTEs were the result of Cannings attempting to smother the children by obstructing their upper airways,(213) and the reasoning was supported by evidence that Cannings frequently forgot to use the apnoea monitor.(214) 

At the time when the mothers in this study were having their families, those with a new baby where there had been a previous sudden infant death (SID), were offered support from the Care of the Next Infant (CONI)(215) programme managed by the University of Sheffield’s Child Health Unit.(216) The worry for parents about how to care for a next infant was considerable, as Frances Rose, who was monitored as a baby explains:

I know my parents went through the CONI scheme with me, ending up with a year or so of sleepless nights due to apnoea monitors (23 years ago these were less than accurate!), which gave them a certain amount of peace of mind, but was coupled with countless false alarms.(217)

Apnoea is the term used when there is no respiratory effort for greater than 20 seconds or for a shorter period if accompanied by cyanosis(218) or bradycardia,(219) as in an acute life threatening event (ALTE).(220) Apnoea monitors are electronic devices activated by sensors attached to a baby’s chest or abdomen that respond to a baby’s respiratory movements and were provided for families to use when the baby was asleep or at night. Waite et al. found that most families (86%) used them.(221) The monitor beeped with respirations and sounded a continuous alarm if the chest or abdomen stopped moving, indicating that respirations could not be detected. A variety of monitors were issued under the CONI programme for home use,(222) but they always had ‘serious drawbacks’(223) because they were unable to ‘reliably detect life threatening events, their high rate of false alarms…failing to reliably detect when babies stop breathing’.(224) Hence, as in Frances Rose’s example, apnoea monitors often sounded an alarm for no apparent reason, and confidence in monitors ‘gradually declined’(225) as parents became more aware of the ‘limitations of the apnoea monitors’.(226) As Judge LJ pointed out in Cannings, ‘it is not, as some think, a machine which prevents an infant death’.(227)

Cannings was issued with a monitor,(228) but her behaviour was argued by prosecution counsel to be anomalous because she often forgot to ensure that it was attached and working,(229) and she reported being unable to remember whether she had heard the sound of the alarm when her babies had stopped breathing.(230) She stated in evidence that ‘the police believed I had never used them [apnoea alarms] at all’,(231) and that police had sound engineers test the alarms.(232) Consequently, prosecution counsel argued that ‘the appellant had not told the full truth about the workings of the apnoea alarm’.(233) Evidence of her inconsistent memories, and emotional reactions was also presented in terms that suggested her behaviour, described as distressed, very shocked, sobbing, retching and vomiting,(234) may like Clark’s have been perceived as too much, and therefore indicative of guilt.(235)

Whether the jury believed that the strength of Cannings’ emotional reactions and the fact that she did not attach the monitor and listen for it at all times supported a finding of guilt, is difficult to know for sure. However, the prominence given to such factors in the appeal judgment suggests that at trial, such considerations were significant. Hallett J directed the jury to ‘look at all the evidence’,(236) and therefore maternal behaviour would have formed part of that appraisal, especially as there is no mention in the appeal report of a defence challenge to such evidence. In addition, although Clark and Cannings may be distinguished by Cannings having lost three babies and Clark having lost two, both were part of the CONI programme and issued with monitors, however Clark did not use the apnoea alarm at all during the day,(237) and this fact was not raised in evidence. A belief might be held that in Cannings’ home where the young infants suffered repeated ALTEs, twenty-four hour monitoring should have been in place. However, monitors were known to be unreliable, infants were under continuous observation and monitor use as a decisive factor in criminal proceedings was inconsistent.

In any event, it is possible that heuristics may have played a part in juror decision making.(238) As Temkin and Krahé suggested in relation to rape trials, counterfactual thinking has been observed to occur when mock jurors are invited to re-imagine a situation such as a rape, and ask themselves what could have been done differently. In such circumstances, mock jurors are more likely to blame the person they have just imagined acting differently. If instead of a rape, the mock jurors were to imagine an ALTE and imagined what could or should have been done differently, then theoretically jurors might blame the mother for not making sure the child was attached to a working apnoea monitor. Of course that may be a very reasonable belief, but whether the belief supports or justifies a finding of murder given the known difficulties of monitoring instruments is a different matter. Nevertheless it is possible that failing to use an apnoea monitor may have been used by the jury as a key behavioural cue in attributing responsibility to the mother.

As Judge LJ later suggested, it was possible that given the large number of experts called and the complexity of the evidence given, the jury may ‘inadvertently, unconsciously, have thought to itself that if, between them all, none could offer a definitive or specific explanation for these deaths, the Crown's case must be right’.(239) Or, if evidence of maternal behaviour was interpreted using fixed beliefs combined with heuristics, a guilty verdict was supported and justified because the expert evidence was so inconclusive, thus presenting the possibility of a mothering myth.  

Not calling for an ambulance: probative of guilt or a mothering myth?

When Cannings found her first baby Gemma ‘lying on her back, looking very, very white. She tried, unsuccessfully, to revive her. She called an ambulance’,(240) but the baby could not be revived. The second baby Jason had an ALTE when the health visitor was present who resuscitated the child prior to his admission to hospital.(241) Jason had a further ALTE at home a few days later, whereupon Cannings dialled 999 and the paramedics arrived.(242) The baby died subsequently in hospital and, following a review of both deaths by leading paediatricians and neuropathologists, no cause of death was identified.(243) The third baby Jade had an ALTE (whilst not connected to the apnoea alarm); Cannings called her GP who attended, and the child was taken to hospital(244) and survived until adulthood.

Following the fourth child Mathew’s birth, Cannings was taught ‘advanced resuscitation techniques’.(245) When Mathew too suffered an ALTE in contrast to the three previous occasions, Cannings did not call 999, but called her husband to come home; no ambulance was called until after he rushed home from work.(246) At the hospital when Mathew was confirmed to have died,(247) Cannings’ husband ‘asked her in the presence of the staff nurse why she had called him before she had called an ambulance, as indeed she had. She was quiet for a few minutes, and then told her husband that she had panicked’.(248) When interviewed by police on this question, she said she had wanted her husband to be present, that she wanted his help, ‘so that he could see Matthew and see what he was like’.(249) The prosecution considered Cannings’ behaviour to be irregular, because although she had commenced resuscitation after ringing her husband, she ‘had not herself directly and immediately sought help either from the emergency services or indeed from neighbours, at least one of whom was a nurse who had offered to help’.(250) In response, Cannings said of Mathew that she ‘“couldn't believe the way he was”. She wanted “Terry to be there to support me. I had always been on my own”’.(251) It is difficult to understand how a mother may have felt in Cannings’ situation at the moment of discovering Mathew, faced with the prospect of losing a third child, and suspicion from her husband and family, and investigations by doctors and police. It is understandable that she did not want to be alone. But, it was ‘Mathew’s death that triggered the investigation which culminated in her conviction’(252) for the murder of both Jason and Mathew. The question whether Cannings’ behaviour in not calling the ambulance immediately in Mathew’s case was so prejudicially interpreted by the court at her trial, cannot be answered. However, within the framework of  expert evidence that ‘lightning does not strike three times in the same place’,(253) then a fixed belief that Cannings should have called the ambulance immediately before commencing resuscitation and before calling her husband, may have supported and justified a finding of guilt.

Nonetheless, if each child’s death is considered separately, Cannings behaved correctly throughout her challenging time as a mother, apart from the very last occasion, when she said she panicked and wanted her husband there. It is therefore also possible that the jury took her long tribulations into consideration and neither failure to use the apnoea alarm nor failure to call the ambulance were significant in a finding of guilt. But the jury needed to decide whether the child deaths were natural or unnatural,(254) and they heard expert evidence that three infant deaths in one family is very rare.(255) In the absence of direct or indirect evidence,(256) this opinion was key in determining that Cannings had a case to answer.(257)

Judge LJ held that the ‘expert evidence was absolutely critical to these convictions’(258) and that the fresh evidence regarding Long QT syndrome undermined the original expert evidence.(259) Accordingly, it is possible that the jury came to a guilty verdict on the basis of a belief in the certitude of expert evidence (later considered to be unreliable in Patel,(260)) rather than because a fixed belief was held about maternal behaviour. If this is so, however, it is unclear why defence counsel were unable to convince the court of the reliability of expert evidence of Long QT syndrome presented at trial, unless other factors were more persuasive, such as adverse interpretations of maternal behaviour. The manner in which carers behave at that critical moment of realising that a child needs help is mentioned in other cases too.(261)

The last section to be examined here on evidence of behaviour relates to mental health, and whether evidence of behaviour can rightly indicate poor maternal mental health that then explains a child’s death. There are two areas of maternal mental health that are referred to in child death cases such as Cannings and Anthony.(262) The first relates to the possibility of child killing whilst a mother’s mind was disturbed,(263) and the second relates to child killing as a result of attention seeking behaviour such as MSbP.(264)

Maternal mental ill-health: probative of guilt or a mothering myth?

Cannings, was described as a ‘woman of good character’,(265) who had no previous convictions,(266) and was ‘a loving mother, apparently free of personality disorder or psychiatric condition’(267) such as depression or post-partum psychosis, and who ‘consistently denied harming any of her children’.(268) Judge LJ explained in the appeal transcript that ‘Without medical evidence about the appellant’s mental state, a verdict of infanticide was not open to the jury’.(269) Judge LJ’s comments indicate that at trial, the court was concerned that if the jury returned a verdict of murder, she should not receive a mandatory life sentence for homicide for each child, but should be sentenced for the offence of infanticide ‘a specific, lesser offence of homicide’.(270) ‘Conviction for infanticide is usually followed by a noncustodial sentence’(271) albeit often subject to a treatment or hospital order.(272)

Cannings was criticised at trial for not calling the ambulance when Mathew was critically ill, for not calmly proceeding to resuscitate Jason or Jade herself, and for not using and being alert to the apnoea monitor at all times. The appeal transcript suggests however that Cannings was ‘faced with recurring disasters which made comprehensible any form of response which, on cold forensic analysis, would otherwise appear strange’.(273) But the jury were told by the prosecution that she would not ‘have killed the children (as the jury found that she had) unless she was suffering from some form of personality disorder or psychiatric condition’.(274)

It is therefore possible that a settled or fixed belief that Cannings’ behaviour was abnormal, together with a prescriptive belief that only a mental health diagnosis could account for the child deaths, justified a guilty verdict, notwithstanding that ‘there was no evidence to sustain any such diagnosis: indeed it was to the contrary’.(275) Hallett J was moved to say after the guilty verdict was returned that, ‘“I have no doubt that for a woman like you to have committed the terrible acts of suffocating your own babies there must have been something seriously wrong with you”’,(276) which she considered was the only way to explain why Cannings could have murdered her children.(277) The judge may have been trying to frame the case in order to allow for a compassionate legal response had an application for an infanticide defence been made. If made at trial, then the mandatory imposition of a life sentence that a murder conviction requires, may well have been averted or, had Cannings later admitted to killing her children, she may have been considered for an expedited appeal.(278) A confession is however required for a defence of infanticide and mothers may be unable to admit to a killing at trial;(279) they may be guilty, or have an undiagnosed psychiatric illness as suggested in Kai-Whitewind,(280) or even be innocent. If a fixed belief exists that mothers such as Cannings should have said and done things differently, together with a belief that mental ill health may explain the behaviour, the killing and a denial of having caused harm, then such beliefs may have supported contested expert opinions, and justified a guilty verdict.

Attention seeking behaviour: probative or a mothering myth? 

The prosecution also contended that Cannings had smothered one of her sons ‘in an attempt to evoke sympathy’,(281) suggesting the ‘need to draw attention to herself, a manifestation of factitious disorder by proxy, a condition which, in her case, was excluded’.(282) The appeal court ‘had difficulty following this suggestion’,(283) and it is unclear why it was made, unless to damagingly associate Cannings with a destructive diagnosis of MSbP.(284) In Anthony(285) however, evidence of an MSbP diagnosis was put forward for admission, but excluded following defence submissions arguing that Professor Meadow’s diagnosis ‘amounted to no more than evidence of propensity’.(286) Evidence of ‘behavioural tendency or propensity’,(287) indicates that a defendant is more likely to behave in a particular way than another,(288) but propensity cannot describe what really happened,(289) only what the defendant may have done.

Anthony had two children, Jordan who died aged almost a year, and Dean who died aged just over four months. Anthony denied harming either child and submitted that both deaths were natural.(290) Post-mortem findings on Jordan showed no evidence that the death was unnatural, nor any indication that the death was natural either,(291) and the cause of death was concluded as unascertained or SUDI. In Michael’s case although features of SID were also identified, the pathologist stated that ‘the possibility of one mother having two unexplained deaths, in other words lightening striking twice, was most unlikely and outside his experience’.(292) He therefore concluded that the babies had both been suffocated.(293)

Professor Meadow who had also given (flawed) evidence in both the Clark and Cannings(294) cases, reviewed the post-mortem findings in Anthony, concluding that the deaths were typical of smothering because of the ‘“incredibly long odds” against two children in the same family dying of natural unexplained causes.’(295) Natural cot death he said happened every 1 in 1000 births, therefore he suggested, the ‘“chance of a natural cot death occurring twice in the same family is 1 in 1000 x 1 in 1000 which is 1 in 1,000,000. It is extraordinarily unlikely…”’.(296) As a result of this conclusion, Anthony was alleged to have killed the children ‘to bring attention to herself’,(297) in line with MSbP,(298) although no expert opinion on MSbP was permitted. Anthony was sentenced to life imprisonment.

Her first appeal in 2000 was refused; at that appeal she sought to exclude expert witness Meadow’s opinions, because he had been of the view (although it was not admitted in court), that she suffered MSbP, and therefore his opinions submitted in court would have been prejudiced by that view.(299) In addition, Anthony’s defence sought to argue that she was suffering from a severe personality disorder at the time of the children’s deaths and that she should have access to the defence of diminished responsibility.(300) Both grounds for appeal were rejected, but five years later, following referral to the CCRC(301) a second appeal accepted that as in Cannings,(302) ‘the occurrence of a second unexpected infant death within a family is not a rare event and is usually from natural causes’.(303) Meadow’s evidence on statistical probabilities was acknowledged to have been flawed.(304) The Appeal court drew on fresh evidence(305) in Anthony to conclude that if the case were to have proceeded at the time of the second appeal then ‘the medical evidence for the Crown would have appeared less compelling than it must have seemed at trial’.(306)

It is therefore difficult to know the extent to which Anthony’s alleged attention seeking behaviour influenced either the trial or the first appeal outcomes. However, given the presence of the leading proponent of MSbP at trial, even though he could not give evidence of MSbP, considerations of mental health may have been significant in both judicial and juror considerations. Beliefs in the actuality of difficult and unlikeable parents,(307) fictitious illness,(308) and mothers who kill their children,(309) may have been encouraged by Meadow’s discourses at JSB seminars,(310)  his publication record,(311) and his argument that if no medical reason could be established for a SID then MSbP should be considered. The RCPCH has published guidance on the diagnosis of FIIC(312) or FIIP (by Proxy), as the syndrome has been variously referred to in the UK, and it remains controversial in both English(313) and Australian courts.(314) It is not appropriate either to dismiss the existence of behaviour that amounts to FIIC, or to suggest that the Crown were wholly wrong to raise the possibility of MSbP in both Cannings and Anthony.  The issue that arises however, is whether a fixed belief is held by agents of the court that where there is no conclusive pathology, nor irrefutable evidence of mental ill health, but there are indications that may be interpreted as attention seeking behaviour, that a mother must be guilty of abuse or murder. 

Conclusion: Does behaviour matter?

Judging character plays a ‘central role’(315) in our ordinary everyday decision making but, within the criminal trial reliance on ‘character becomes more controversial’,(316) particularly as character as indicated by ‘past behaviour and behavioural tendencies’(317) may be introduced at trial as a way of impugning the credibility of a defendant or witness.(318) Personal background information however, together with conduct and character evidence has long been admitted in criminal trials,(319) in order that the courts have an opportunity of hearing from defence character witnesses about the nature of the accused, and whether the allegation was “out of character”.(320)

The reason for the admission of character evidence is that it indicates dispositions which persist over time, and aids interpretations of behaviour around the time of a criminal event based on past behaviours.(321) But, behaviour as character could also be used to denote a person’s ‘bad character, previous convictions, disposition, or reputation’(322) and such evidence has been considered inadmissible at common law; bad character evidence was considered prejudicial to a defendant because it ‘encouraged mistaken inferences and faulty fact finding’(323) and demonstrated  merely propensity.(324)

The Court of Appeal has voiced its concerns relating to the inclusion at common law of material that demonstrates propensity or reduces credibility because such background evidence may be used as a vehicle for smuggling in otherwise inadmissible evidence.(325) This is important, because the difference between how the legal system considers evidence should be interpreted,(326) and how evidence is interpreted by the public,(327) is an area that ‘we know little about’.(328)   If the legal profession and the judiciary have concerns about the role of background evidence and its permissive admission, the use of warning devices such as the modern rape myth and the modern mothering myth become indispensable in highlighting and preventing possible injustices. 

That background evidence may be used in criminal trials is not novel, but we should be cautious about using non-criminal behaviour evidence because as Redmayne signals, the trial may then become an examination of the ‘defendant’s life and attitudes’(329) with the possibility that for all of us, and especially mothers accused of killing their children, ‘the state might dredge up any unpleasant behaviour in our lives and use it to convict us’,(330) or of course in rape trials perhaps, to wrongly acquit the defendant.

RMS tells us clearly that female behaviour and its interpretations matter. It also suggests that further research is needed into understanding where there is meaning, relevance or even probative value in particular female behaviours if they are to be used as evidence, and into the ways in which the public interpret behaviour evidence. RMS has had success in challenging the permissive admission of evidence of complainant behaviour in rape trials and by using the device of a rape myth has sought to isolate, define and measure the prevalence of beliefs which may be used in criminal trials to show propensity and/or bad character. In so doing, RMS has uncovered multiple layers of complexity comprising beliefs, attitudes and norms relating to women, with varying functionalities and interconnections.

The same complexities may be true also of beliefs about mothers and their behaviours when they are accused of killing their children. By being more aware of the pitfalls of language, and the firmer ground of normativity we may be able to identify the operation of prescriptive beliefs in child death cases. The modern Mothering Myth may provide us with a device with which to challenge the admissibility of evidence of a mother’s conduct and behaviour prior to and around the time that a child dies. It is essential for justice that we distinguish between beliefs going only to propensity and actual conduct relating to criminal liability when mothers are accused of killing their children. We can and should challenge modern mothering myths.