“The Mad”, “The Bad”, “The Victim”: Gendered Constructions of Women Who Kill within the Criminal Justice System

Abstract

Women commit significantly fewer murders than men and are perceived to be less violent. This belief about women’s non-violence reflects the discourses surrounding gender, all of which assume that women possess certain inherent essential characteristics such as passivity and gentleness. When women commit murder the fundamental social structures based on appropriate feminine gendered behaviour are contradicted and subsequently challenged. This article will explore the gendered constructions of women who kill within the criminal justice system. These women are labelled as either mad, bad or a victim, by both the criminal justice system and society, depending on the construction of their crime, their gender and their sexuality. Symbiotic to labelling women who kill in this way is the denial of their agency. That is to say that labelling these women denies the recognition of their ability to make a semi-autonomous decision to act in a particular way. It is submitted that denying the agency of these women raises a number of issues, including, but not limited to, maintaining the current gendered status quo within the criminal law and criminal justice system, and justice both being done, and being seen to be done, for these women and their victims.

1. Introduction

This article will discuss the symbiotic concepts of labelling and agency in the context of the socio-legal constructions of women who kill. Women commit significantly fewer murders than men. Indeed, the most recent official crime statistics detail that of the 121 people convicted of murder in England and Wales in 2011/12, 108 of those individuals were men, 13 were women [1]. Therefore, when women commit violent crime, more specifically murder, some of the fundamental social structures based on appropriate gendered behaviour are contradicted and challenged.1 Due to the challenge that women who kill pose to gender discourse within a patriarchal society, through their violations of both societal and gender norms, an explanation is often sought for their behaviour. Such explanations are, not-surprisingly, reflections of the images of women as portrayed in current gender discourse surrounding appropriate feminine behaviour. As a result, both society and the law label women who commit murder as mad, bad or victims.2 Symbiotic3 to this labelling of women who kill is the denial of their agency by society, the law and the criminal justice system.

Agency is a complex, multi-faceted, interdisciplinary concept, with a multitude of definitions, ranging from the subtly different to the divergent. Therefore, for the purposes of this article a specific definition of agency will be used, that is; the recognition of the ability of an individual to make a semi-autonomous decision to act in a particular way. This reflects Messerschmidt’s definition, that ‘[a]gency refers to the behaviours in which a person chooses to engage in order to shape his or her experiences within social structures in light of his or her understanding of the social structures that surround and constrain his or her options’ [7]. This article will explore the manifestation of the three labels attached to women who kill, before examining how each of these labels deny the agency of women who kill. Finally, it will critically question the consequences that labeling and agency denial have when exploring justice for both women who kill and their victims within the criminal justice system.

2. Battered Women Who Kill—the Mad Woman and the Victim

Battered Woman Syndrome (BWS) ‘[w]as developed by the American psychologist Lenore Walker in order to dispel myths and misconceptions about domestic violence and to help establish the reasonableness of homicide by battered women’ ([8], p. 733). The syndrome consists of two elements. The first element is known as ‘the cycle theory’. This suggests that characteristically male violence against their partners has three phases:

The first involves a period of heightening tension caused by the man’s argumentativeness, during which the woman attempts various unsuccessful pacifying strategies. This “tension-building” phase ends when the man erupts into a rage at some small trigger and acutely batters the woman. This is followed by the “loving-contrite” or “honeymoon” phase, in which the guilt-ridden batterer pleads for forgiveness, is affectionate and swears off violence. But he breaks his promise and the cycle is repeated.

([8], p. 733)

The second element of BWS involves ‘learned helplessness’. Repeated, unpredictable and seemingly unavoidable abuse by their partner results in battered women becoming increasingly passive and developing a number of characteristics including low self-esteem, anxiety and depression as well as blaming themselves for the violence they suffer. This sense of helplessness traps battered women ‘[i]nto a situation from which [they are] psychologically and hence physically unable to escape’ [9].

Before the introduction of The Coroners and Justice Act 2009, case law demonstrated that the inclusion of BWS evidence in cases of women who killed their abusive partner was recognised in relation to the defences of provocation [10] or diminished responsibility [11]. However, with the introduction of the 2009 Act, BWS is now primarily a matter for the amended defence of diminished responsibility, with women no longer being able to adduce evidence of BWS under the new defence of loss of control. Therefore women who plead loss of control will now only be able to present themselves as battered, rather than suffering from BWS. As a result of this shift in the law, women who plead loss of control will be labelled as victims, whereas women who utilise evidence of BWS to support a plea of diminished responsibility will be labelled as mad.4

2.1. Loss of Control—Battered Women Who Kill as Victims

The new partial defence to murder of loss of control is found in sections 54–56 of the Coroners and Justice Act 2009. The effect of the defence of loss of control in practice in cases of battered women who kill their husbands is still unknown as there is yet to be a reported case involving a battered woman pleading the new defence of loss of control. However, as noted above, it is submitted that in theory women will not be able to use evidence of BWS to support the new defence of loss of control. Indeed, Alan Norrie has suggested that the amendments to the law, particularly the defence ground concerning the defendant having a justifiable sense of being seriously wronged, may encourage a change in how battered women are portrayed within the legal system, should their defence utilise loss of control. He notes that;

Under the old law, defendants were encouraged to provide evidence of a characteristic that could be taken into account with regard to the reasonableness of their conduct. This led to the pursuit of a medico-legal category, battered woman syndrome, which could legitimate the existence of the characteristic for legal practice … Under the new law, defendants and their lawyers will be encouraged to portray themselves as ordinary people grievously harmed and acting out of a legitimate sense of anger at what has been done to them. This may be a benefit of the new approach.

([12], p. 286)

Therefore, although women will no longer be able to use evidence of BWS when utilising loss of control, they will still be able to use evidence that they were indeed battered women. As such, it is argued that battered women who plead loss of control will be labelled as victims.

The construction of women largely being victims of crime ‘[b]egan to emerge in the 1970s with the rise of radical feminism and demands to make violence a public not private matter’ [13]. The theory of women as victims of crime was developed by many academics, focusing on women as victims of violence, particularly within their own home, but also more generally. As summarised by Carrington;

Significant and influential works include Dobash and Dobash’s (1979) study of family violence; Russell’s (1975) exposé of rape, including rape in marriage, and Brownmiller’s (1975) provocative analysis of rape to name only a few. These were followed by Stanko’s (1990) work on everyday violence and Walklate’s (1991, 2007) major and ongoing contribution to the field of victimology.

[13]

As is clear, traditionally much of the academic research surrounding women and violence has focused on women as victims of domestic violence, rather than as perpetrators of the violence themselves. This is arguably partly because women as perpetrators of violence is considered to be a relatively rare phenomenon ([3], p. 169).5 Historically, much of what was written on female criminals focused on pathological and irrational discourses to explain their involvement within the criminal justice system. However, with the development of the theory of Battered Woman Syndrome (BWS) and the introduction of the new partial defence to murder of loss of control the idea of women as both victims and perpetrators was reconciled. That is to say that women became perpetrators because they were victims.

Although, as noted above, battered women who kill their abusive partners cannot use evidence of BWS to support a defence of loss of control, the fact that these women can still present themselves as battered means that they are labelled as victims. The image of helplessness associated with a battered woman has promoted ‘[a] collective understanding of the battered woman as a person whose identity is predominantly that of a victim’ ([14], p. 113). Indeed, the labelling of these women as victims sits well with gender discourse, particularly the idea that women are ‘[s]ubject to control at the hands of their partners and a patriarchal society’ [4]. Therefore it is submitted that in order for a woman to present herself as battered, even if not suffering from BWS, requires her to conform to gender discourse surrounding appropriate femininity. Indeed, it is clear that a woman’s gendered behaviour is still on trial both when she commits a crime generally, and more specifically, when she murders her husband ([12], p. 277). Therefore, a battered woman must present herself as someone who is a ‘[a] faithful wife, a devoted mother, someone who tries to keep her family together at all costs and who reacts meekly and pathologically to violence’ ([8], p. 735). Women who conform to such appropriate gendered behaviour are viewed as ‘true’ victims of domestic violence within legal and social discourse. Women who do not conform are not really battered and are therefore ‘[u]ndeserving viragos’ ([15], p. 195). This therefore suggests that not only do battered women have to conform to appropriate feminine behaviour generally, but they must also conform to the appropriate behaviour expected of a battered woman.

The ‘appropriate’ behaviour expected of a battered woman is often linked to the concept of learned helplessness, the ‘[m]ost prominent component’ ([16], p. 113) of BWS. Indeed, as was noted by Kathleen Ferraro, this concept of learned helplessness established ‘[a] perception that assertiveness, strength and an outgoing personality were inconsistent with being a battered woman’ ([16], p. 115). Based on this analogy viragos are not really battered because they ‘[f]ight back’ ([15], p. 195), thus reflecting the label of victim used to describe battered women. The suggestion that women must conform both to appropriate standards of femininity as well as the behaviour expected of a battered woman is supported by a study carried out in the United States by Brenda Russell and Linda Melillo [17]. The study involved six hundred and eighteen undergraduate students from two St Louis Universities who were presented with actual case summaries ‘[t]hat included standard forms of expert testimony modelled after BWS evidence’ ([17], p. 223). The results provided persuasive evidence that women who fit the typology of a passive, non-responsive battered woman who kills were deemed to be more credible and therefore were most likely to receive not-guilty verdicts for the charge of homicide. Conversely, women who were atypical and actively responded to their partner’s violence were viewed as less credible and consequently received more guilty verdicts.6 Labelling battered women who kill as victims presumes that they are so oppressed that they are powerless and as a result they will be non-violent. However, when battered women do become violent, resulting in the death of their abusive partner, the label of ‘victim’ offers an explanation for their actions.

2.2. Diminished Responsibility—Battered Women Who Kill as Mad

The Coroners and Justice Act 2009 also amended the defence of diminished responsibility. The wording of the current definition of diminished responsibility differs considerably from that which was found in the Homicide Act 1957. In short ‘abnormality of the mind’ has been replaced with ‘abnormality of mental functioning’, there is a requirement that the abnormality ‘arose from a recognised medical condition’, the abnormality must have substantially impaired the abilities of the defendant as listed in Section 1(1A) and the abnormality must have been a significant causal factor in the defendants actions.7 Despite these changes, the Ministry of Justice in its Impact Assessment of the 2009 Act stated;

[w]e do not expect any significant shifts in the numbers or types of cases which benefit from the partial defence of diminished responsibility... We do not therefore think that there will be an impact on the courts or prison population as a result of the changes.

([19], p. 301)

Despite the government’s assertions that the 2009 Act will have little impact, it is suggested that there is potential for impact in cases where BWS is adduced to support the plea of diminished responsibility.

The main impact of the change in the law of diminished responsibility on battered women who kill their abusive partners is the requirement that the abnormality of mental functioning must arise from ‘[a] recognised medical condition’ [20]. The Ministry of Justice have made it clear that this phrase will cover both psychological and physical conditions and therefore is not just ‘[l]imited to recognised mental disorders’ ([19], p. 294). Consequently this concept covers more than was previously covered in the unamended Homicide Act 1957. Although there is yet to be a reported case of BWS being used to support the amended defence of diminished responsibility, it is submitted that evidence of BWS can now be more easily entered to satisfy this particular requirement within the amended defence. As long as the Court is satisfied that the woman suffering from BWS and the killing of her abusive partner are sufficiently connected, the defence should succeed. Discussing the requirement of a connection, the Attorney General noted;

The government consider it is necessary to spell out what connection between abnormality of mental functioning and the killing is required for the partial defence to succeed... It need not be the sole cause or even the most important factor in causing the behaviour but it must be more than merely a trivial factor.

([19], p. 298)

The use of BWS evidence to support a plea of diminished responsibility simultaneously reflects and reinforces the gender stereotypes surrounding women. Indeed, a study on cases of diminished responsibility highlighted that;

Reports written for male defendants in which this plea was possible indicate the readiness with which they were created as ‘monsters’ or ‘madmen’, yet simultaneously capable of intending their behaviour, since men are to be understood in terms of what they do.

[21]

This could be contrasted with the treatment received by female defendants. Female law-breakers were ‘[m]ore readily constructed as “normal women”’ [21] and therefore they were more likely to experience diminished responsibility than their male counterparts. The explanation put forward for this discrepancy in the treatment of men and women when pleading diminished responsibility was based on gender stereotypes, that is to say that ‘[t]hings happen to women; they do not make rational decisions or choices’ [21]. The overwhelming conclusion of the study, as noted by Sandra Walklate, was that;

[w]hen psychiatry and the law interact, the resultant effect is that men are, for the most part, attributed with a sense of agency and responsibility for their actions, whereas women defendants are denied this.

[21]

As such, it is submitted women who use BWS evidence to support a plea of diminished responsibility have their agency denied through labelling them as mad.

This labelling of battered women who kill as mad when using evidence of BWS to support a plea of diminished responsibility is also reflected in the theory of BWS itself. The use of the term ‘syndrome’ within the name BWS, is according to Schopp, itself indicative of ‘[a] psychological disorder, an abnormality in human behaviour ([3], p. 76)’. Consequently, the utilisation of BWS evidence when pleading diminished responsibility pathologises the actions of battered women who kill and reinforces the construction of women as irrational beings. Furthermore, adopting syndrome language contributes ‘[t]o an image of battered women as psychologically defective or pathological’ ([16], p. 112). This adoption of syndrome language in the context of battered women who kill their abusive partners sits nicely with the construction of femininity where women are represented in terms of their bodies. That is that ‘[t]he “normal” female body and mind are perceived as being predisposed to malfunction’ [22].

Many commentators have correctly noted that the inclusion of evidence of BWS in a plea of diminished responsibility can result in women being sentenced more leniently. Rather than being imprisoned, ‘a finding of diminished responsibility may result in a woman’s long-term medical or psychiatric treatment’ ([15], p. 192). Consequently, although BWS may offer benefit to some women offenders by offering an explanation for their actions “[B]WS obviously works within the stereotype of women as ‘crazy’” [23]. It is possible of course, that many female defendants will not care how they are stereotyped, as long as the result is a more lenient sentence. However others will care and ‘[w]ill undoubtedly perceive it to be deeply insulting to be told that, unless they accept a label of psychological abnormality, they run the risk of escaping the prison of domestic violence only to spend a long time in a less metaphorical prison’ ([8], p. 737). Therefore, although introducing evidence of BWS to support a plea of diminished responsibility may result in sentencing benefits for women who kill, it also ensures that gender stereotypes surrounding women’s mental health remain firmly entrenched. Moreover, as noted by Morrissey, evidence of BWS is;

[r]ather less useful in supporting the most appropriate defence for battered women who kill, the justification defence of self-defence. Evidence of battering and abuse is clearly useful in determining whether an individual battered woman was in fear of her life that the killing of her partner was necessary; but evidence as to her psychological state and her subscription to a debilitating syndrome actually undermines such a defence.

([3], p. 77)

Consequently utilising evidence of BWS to support a plea of diminished responsibility provides an explanation for both society and the criminal justice system when a woman murders her abusive partner, namely that she did so because she was mad. Using this explanation of madness fails to acknowledge that battered women who kill were acting in justifiable self-defence. Indeed, labelling a battered woman who kills her abusive partner as mad and denying her agency is in contrast to the feminist jurisprudence model which ‘[e]xplains the battered woman who kills as … a rational individual who defended herself under reasonable life-threatening circumstances’ ([14], p. 116).

From the above it is clear the evidence of BWS was historically used in relation to both the defences of diminished responsibility and provocation. With the recent amendments to the law it appears that the form of BWS commonly used will still be utilised to support the defence of diminished responsibility, with the new defence of loss of control requiring women to present themselves as battered, rather than using evidence of BWS. As a result women who plead loss of control and present evidence that they were battered are labelled as victims, whereas women who use evidence of BWS to support a plea of diminished responsibility are labelled as mad. It is clear that the use of both the labels, victim and mad, ‘[a]lways actively shift the emphasis from the reasonableness of the defendant’s actions to her personality in a way which confirms existing gender stereotypes [and] silences battered women’ ([8], p. 734). Indeed, the ongoing use of the labels mad and victim reinforces existing gender discourse surrounding femininity.

3. Infanticide—the Mad Woman

Throughout history, ‘the “medicalisation” of women’s behaviour has ... been a common response to female violence. Thus women are thought to become violent because they are mentally deranged or have uncontrollable “raging hormones”’ ([24], p. 425). Lombroso and Ferrero were amongst the first proponents of pathologising female offenders’ behaviour. Their work on the female criminal concluded that as a result of their biological make-up, women were less highly developed than men and therefore they were less likely to commit crime. They stated that women were ‘[m]ore primitive, the consequence of which was that they have less scope for degeneration’ ([25], p. 301). The female criminal was therefore labelled as ‘abnormal’ and ‘pathological’. Despite Lombroso and Ferrero’s work being universally criticised [26], both society and the law continue to locate women’s criminality within the ‘psy’ discourses, with ‘[e]ven the most up-to-date studies ... finding that women criminals are ... psychologically maladjusted’ ([27], p. 36). This is particularly the case for female killers, especially for women who kill their children. The pathologisation of these women is demonstrated by the offence/defence of infanticide for women who kill their young children.

The Infanticide Act 1938 repealed and re-enacted, with modifications, the provisions of the Infanticide Act 1922. The introduction of the Infanticide Act was the result of ‘[a] policy decision to promote leniency for women who kill their own children’ [28]. Section one of the Act states:

Where a woman by any wilful act or omission causes the death of her child, being a child under the age of twelve months, but at the time of the act or omission the balance of her mind was disturbed by reason of her not having fully recovered from the effect of giving birth to the child or by reason of the effect of lactation consequent upon the birth of the child, then, notwithstanding that the circumstances were such that but for this Act the offence would have amounted to murder, she shall be guilty of felony, to wit of infanticide, and may for such offence be dealt with and punished as if she had been guilty of the offence of manslaughter of the child.

[29]

Before a more detailed discussion of infanticide takes place, it is essential to note several particulars about the Infanticide Act, as outlined above. Firstly, women can plead infanticide as their defence, as well as being convicted of the offence of infanticide. To maintain cohesion and clarity within this article, the word ‘offence’ will be used when discussing infanticide. Secondly, the offence of infanticide is only available to the biological mother of the child who has been killed. Thirdly, the age limit of the victim is set at 12 months, and finally ‘it is the only offence known in English law for which a pre-condition is the possession of an abnormal mental state’ ([5], p. 664).

The offence of infanticide provides a clear example of the assumed ‘[u]nderlying pathological nature of mothers who kill their children’ ([30], p. 206). This is reflected in the legal basis for the plea of infanticide; puerperal psychosis. Puerperal psychosis is;

[A] relatively rare and severe mental disorder which affects one or two out of every 1,000 women within the first few weeks of childbirth. The symptoms span a number of categories of psychosis, but range from mania to delusions to acute depression.

([30], pp. 206–07)

Despite puerperal psychosis in theory being required to convict a woman of infanticide, it ‘[i]s very rarely the cause of a mother killing her child. Estimates are that this occurs in around five cases a year’ ([30], p. 207). As a result, in practice the requirement of puerperal psychosis is interpreted far more liberally, often to include any sort of mental illness. However, research also suggests that ‘[a]bout half of the women who ... are convicted of infanticide are not suffering from any identifiable mental disorder at all’ [31]. Statistics such as these demonstrate that women are being convicted of infanticide and having their actions pathologised despite not satisfying the required criteria.

Women who are convicted of infanticide but are not suffering from a mental disorder are therefore routinely being labelled as mad without having any evidence to support such an assertion. Such labelling ‘[i]nvolves considerable social stigma, a high degree of intra-psychiatric treatment and the reinforcement rather than the challenging of traditional gender stereotypes’ ([24], p. 425). It is submitted that the reasoning behind convicting women who are not suffering from any identifiable mental disorder of infanticide is that it offers an explanation for their actions. In the case of infanticide; she killed her child because she was mad.

Viewing filicidal women as mentally ill, regardless of whether there is evidence to support such an assertion, ‘[f]its very well with certain ideas about women, femininity and motherhood’ ([32], p. 33). According to Frigon;

At the beginning of the twentieth century ... Motherhood was ... constructed as “natural” and a consequence of heterosex. As “compulsory motherhood” was introduced, it meant more than the imposition of pregnancy and birth but also “entry into the nexus of meanings and behaviours which are deemed to constitute proper mothering”.

([32], p. 31)

The qualities and behaviours which constitute proper mothering are a reflection of those which constitute appropriate feminine behaviour; ‘the ideology of motherhood ... increasingly identifies women solely in terms of children’ [33]. Indeed, ‘women are assumed to be inherently passive, gentle, and tolerant; mothers are assumed to be nurturing, caring and altruistic’ [31]. The actions of filicidal women are so starkly in contrast with the construction of appropriate motherhood and mothering behaviour that an explanation must be sought for their behaviour. This explanation can be found in the form of the Infanticide Act that operates, as noted above, within the ‘psy’ discourses. The Act presumes that a woman ‘[m]ust have been “mad” to kill her own child’ [31]. To put it another way;

So untenable, unthinkable and inappropriate the crime, so much is it at odds with normal motherhood or the feminine predilection for surrogate motherhood that such women can only be immutably unnatural.

[34]

From the above analysis it is clear that women can be convicted of infanticide even if they are not suffering from ‘puerperal psychosis’. Wilczynski highlighted this point;

[v]irtually any type of perceived psychiatric, emotional, personal or mental problem whatsoever can be interpreted (if the psychiatrists, lawyers and/or judges so choose) as the severe mental illness (puerperal psychosis) theoretically required for the Infanticide Act.

([32], p. 34)

In addition, those women convicted, either rightly or wrongly, of infanticide are more likely to be dealt with ‘[b]y more informal and “treatment”-oriented methods’ ([24], p. 423). It is suggested that it is appropriate for those women suffering from a genuine and identifiable mental illness to be charged with the offence of infanticide and therefore be treated appropriately as a result. However, it is arguably troubling to think that women who are not suffering from any identifiable mental illness whatsoever are being convicted of infanticide in order to offer an explanation for their ‘unthinkable’ actions.

Convicting a woman of infanticide when she is not suffering from the requisite mental illness often results in her being given a non-custodial sentence at the expense of her being labelled as mad. From this it seems fair to suggest that the Criminal Justice System would rather label a filicidal mother as mad, regardless of whether she actually is, in order to provide an explanation for her behaviour, than acknowledge her responsibility for her actions. The existence of such a practice within the Criminal Justice System further entrenches gender stereotypes surrounding women. That is to say that it enforces the idea that women are mad generally, but especially when they commit murder.

4. The Bad Woman

It must be noted at the outset that ‘bad’ is a word which is used throughout the literature and therefore is one which will also be used in this article. However, it is acknowledged that the use of the word bad to label women who kill is problematic, as society would view most criminals as being bad people. Therefore, when using the term bad in the context of women who kill what is actually being alleged is that these women are perceived as being wicked, an ‘extra element’ of bad that goes beyond their actual crime. This extra element of bad is as a result of the violation by these women of too many societal and gendered norms which cannot be explained through the use of the labels mad or victim. So, for example, a woman who kills her child but is not diagnosed with a recognised mental disorder which would allow her to be labelled as mad, is labelled as bad. The extra element of bad, leading to her being perceived as wicked, is her violation of the gendered and societal norm of ‘good’ motherhood for women.8

It has been shown that if the required conditions are met, or even if the facts of the case or the behaviour of the woman in question can be moulded to fit the required conditions, then women will be labelled as mad or as victims. However, if the actions of the female killer and her background cannot be moulded in such a way as to fit either label, then another explanation for her actions must be found. This explanation takes the form of labelling her as bad. The distinction between good and bad women is not a new one. In their work on the female born criminal, ‘[L]ombroso and Ferrero defined distinctive sub-species of women as “good” and “bad”’ ([35], p. 115). Indeed, the dichotomy between good and bad women is not only found within academic work but it is ‘[a] constant theme in art literature, films and other media’ ([35], p. 99). It therefore becomes clear that there is a trend to label female killers as bad when their actions cannot be explained utilising the other labels discussed above.

“‘[B]ad’ women are cold, selfish and are ‘non-women’ or masculine or even monsters” ([32], p. 34). This can be contrasted with so-called good women who, according to Pollack, ‘[a]re conventional socially and morally and if they do transgress it is in ladylike and peculiarly feminine ways’ ([35], p. 148). The immediate difference between so-called good and bad women is the way in which their lifestyle and behaviour either does or does not accord with appropriate feminine behaviour as dictated by gender discourse. A similar principle applies to women who kill. Although these women can, for obvious reasons, never be labelled as good, if their behaviour and lifestyle cannot be explained by labelling them as mad or as a victim, and they have the requisite extra element of badness, then the only other explanation on offer for their actions is quite simply that they are ‘inherently bad’. Bad women are often sub-categorised into particular types of bad women. These categories include, but are not limited to women who kill who display sexually deviant behaviour and women who kill who are considered to be bad mothers.

4.1. Sexually Deviant Women

As noted above, women who kill and also display what is regarded as sexually deviant behaviour are often labelled as bad. Labelling women as bad for this reason demonstrates an attempt by both society and the law to regulate female sexuality. During the nineteenth century the ideal woman was ‘[d]ocile, chaste, modest, pious, religious, maternal and above all obedient to patriarchal authority’ [6]. Indeed;

A recurrent feature of feminine respectability is sexual propriety ... Historically, women have been judged more harshly than men if they do not meet expectations of appropriate sexual behaviour in terms of chasteness and monogamy, and these norms have played a more important role in the regulation of femininity than masculinity.

([36], p. 64)

A similar ideal is still expected of women today; women must still conform to what is considered to be appropriate sexual behaviour; that is to say that they must not have too many sexual partners, and that they must have the ‘right kind’ of sex. Moreover, there is still the view that women’s ‘[g]reatest sexual fulfilment .... [should come] ... from having babies’ ([27], p. 278). Linked to this is the idea that women’s relationships should be heterosexual, with women engaging in lesbian relationships considered to be especially deviant, as female homosexuality is considered to be ‘[s]everely at odds with the contemporary normative ideal of marriage and motherhood for women’ ([36], p. 106). Consequently it is clear that women can be labelled as sexually deviant if they are sexually promiscuous, too sexually adventurous or are not involved in heterosexual relationships.

Many feminist criminologists have argued that patriarchy requires that women who are considered to be sexually deviant must be controlled. Heidensohn has noted that the law, particularly the criminal law, is the main control mechanism in this context. She has suggested that the law controls female sexuality in four ways;

• The courts operate a “double standard” with respect to sexual behaviour, controlling and punishing girls, but not boys for premature and promiscuous sexual activities.
• The courts—and probation officers and social workers—“sexualise” normal female delinquency and thus over-dramatise the offence and the risk.
• “Wayward” girls can come into care and thence into stigmatising institutions without ever having committed an actual offence.
• Deviant women … that is, women who do not conform to accepted standards of monogamous, heterosexual stability with children, are over-represented amongst women in prison because the courts are excessively punitive to them ([25], p. 817).

Drawing upon Heidensohn’s theory, it is submitted that women whose sexuality requires regulation by the criminal law are considered to be bad women. The behaviour of these sexually deviant, bad women is the mirror opposite of that of good women, whose sexuality does not need to be controlled by the law. Consequently, female killers who demonstrate sexual deviancy when committing their crimes, or indeed demonstrate it within their lifestyle generally, are most certainly bad and must therefore be controlled through punishment. Not only have they offended against appropriate feminine behaviour by being murderers, they have also offended against appropriate female sexuality through demonstrating sexually deviant behaviour. Therefore, the only label considered to be suitable for such women is bad.

The cases of the female serial killers Myra Hindley and Rosemary West are examples of female killers who also demonstrated sexual deviancy and were consequently labelled as bad women. Although these women were convicted in 1966 and 1995 respectively, the infamy of their cases means that they are both still regularly mentioned in the media, as well as frequenting academic research. Therefore an analysis of their cases is particularly relevant to this article.9 Moreover, the cases of both women are representative of the pervasive and enduring narratives that surround women who kill who are labelled as bad. During Myra Hindley’s trial the prosecution sexualised all of her relationships even if they were not sexual in nature. ‘For example, the prosecution sexualised her friendship with her young neighbour Pat Hodges, describing it as giving her “a kick”, “certain enjoyment” and “morbid satisfaction”’ ([39], p. 356). Before, during and after her trial, the media made much of the fact that Myra engaged in sadistic sexual behaviour with her partner in crime, Ian Brady, that she allowed him to take pornographic photographs of her [40] and that once she was in prison she began a lesbian love affair with one of the female prison wardens [41].

Similarly, in the case of Rosemary West, the judge used his summing up to condemn her deviant sexuality. During the summing up Rosemary was labelled a prostitute and was described as being either bisexual or a lesbian. The judge also noted that she; ‘[p]ossessed a collection of ‘dildos, rubber underwear, pornographic videos, a rice flail, and a whip and a suitcase which contained a quantity of leather straps and buckles’ ([39], p. 359). This collection of sex toys was depicted as solely belonging to Rosemary, despite the fact that it could have just as easily belonged to both her and her husband, Fred West. In fact it is submitted that it should not have mattered who they belonged to, as their existence had limited legal relevance, despite the judge suggesting otherwise. Media reports and academic writing on Rosemary and her crimes also highlighted her sexual deviance, particularly her sexual relationships with other women [42] and the sexual abuse she inflicted on her own children [43]. Undeniably, ‘Rosemary West’s persecution was primarily based on her sexual crimes and her violent, debauched sexuality, thereby contravening the strictest social taboos of “normal” heterosexuality’ ([44], p. 22).

4.2. Bad Mothers

Another subcategory of bad women is that of bad mothers. Women who kill their children are routinely considered to be bad mothers if the specifics of their case cannot be moulded in such a way to allow them to utilise the plea of infanticide. These women are bad because not only have they committed murder, they have murdered their own child, thereby demolishing the construction of motherhood for women. An example of this is the case of Susan Poole10, who allowed her son to starve to death. Despite suffering from depression, she was found culpable for her actions. Susan Poole was charged alongside her partner, Frederick Scott, with the murder by starvation of her 10-month-old son, Dean. Susan pled guilty to manslaughter on the grounds of diminished responsibility. At trial, four psychiatrists and one doctor gave evidence that Susan was suffering from a personality disorder and severe depressive illness at the time of the offence ([30], p. 212). It must be noted that at the time of the trial, Susan had made a substantial recovery from her mental disorder. Consequently, the judge found Susan responsible for her actions; ‘when all is said and done, you killed your one son’ ([30], p. 214).

The judge also portrayed her as a bad mother; ‘when one thinks of the extraordinary maternal sacrifice and care shown by lower animals, one has to wonder at her apparent selfishness’ ([30], p. 213). Despite a probation order with the requirement of mental treatment being recommended, the judge instead sentenced Susan to seven years imprisonment ([45], p. 382). She successfully appealed against her sentence and it was reduced to five years. When considering her appeal, the Court of Appeal concluded ‘[t]hat a sentence of seven years was excessive in all the circumstances of this case. There was the appellant’s unstable background, her age, her previous good character and her plea of guilty’ ([45], p. 388). They also noted that Susan’s depression accelerated rapidly and ‘[t]hat it played a very substantial part’ ([45], p. 388) in Dean’s death. However, the Court clearly still felt that Susan needed punishing for her actions. They agreed with the trial judge’s verdict on her responsibility, as well as refusing to issue the recommended probation order with mental treatment instead of the continuation of Susan’s prison sentence. Therefore, it is reasonable to infer that the Court of Appeal also felt that Susan was a bad mother and deserved imprisonment.11 Indeed, as was noted by Morris and Wilczynski; ‘it is difficult to avoid the conclusion that it was the negative portrayal of her as a woman and as a mother which was the determining factor in her treatment within the criminal justice system’ ([30], p. 214).

The reasoning behind the labelling of filicidal women as bad when they either fail in pleading, or cannot utilise the plea of, infanticide is a consequence of society’s construction of motherhood. The status of women, both socially and legally is determined by motherhood. Women are not only expected to be mothers, but they are also expected to be good mothers;

The single defining characteristic of iconic good motherhood is self-abnegation. Her children’s needs come first; their health and happiness are her primary concern. They occupy all her thoughts, her day is constructed around them, and anything and everything she does is for their sakes. Her own needs, ambitions, and desires are relevant only in relation to theirs. If a good mother takes care of herself, it is only to the extent that she doesn’t hurt her children.

[46]

When mothers do not meet the standards of behaviour prescribed above without a reasonable and rational explanation, they are labelled as bad mothers. This dichotomy between ‘“[g]ood” and “bad” mothers serves as a means of patrolling, controlling and reinforcing the boundaries of behaviour considered “appropriate” for ALL women and mothers’ ([30], p. 217). ‘Society considers women who fail to meet the ideal of motherhood deviant or criminal’ ([47], p. 98). Consequently the law often treats mothers who commit crimes against their children, without the explanation of suffering from a recognised mental disorder, harshly for violating the traditional gendered role.

Bad women are considered to be ‘[e]specially difficult to construct in relation to acceptable performances of femininity’ ([36], p. 8). This is in contrast to women who can be ‘[p]erceived as victims or their actions explained through mental illness’ ([36], p. 8). This is because women represented as being mad or victims are ‘[m]ore recognisably feminine’ ([36], p. 8) and consequently ‘[t]hey do not cross the boundaries of gender’ ([36], p. 8) in the same way that bad women do. Consequently, female killers who are constructed as bad, either because they are sexually deviant, or because they are bad mothers, are harshly punished ([32], p. 34). These bad women are viewed as being doubly deviant; not only have they broken the law but they have also violated appropriate gender behaviour. They are punished more severely than women whose behaviour can be more readily constructed within feminine discourse;

[i]t is clear that it is only certain types of women—those who are perceived as conforming to gender stereotypes—who benefit from these more informal means of social control. Women who resist more informal mechanisms of social control can also be punished by being moved “up-tariff” and subjected to more formal means of social control such as a prison sentence.

([24], p. 431)

This harsh treatment is particularly true for women who murder either their own or other women’s children. As noted by Dorothy Roberts;

Professor Daly found that familied women who committed crimes that made them “bad” mothers, such as sexual abuse of children or prostitution, did not receive the courts’ mercy. These women not only break the law, but by breaking the law they transgress their own female nature and their primary social identity as a mother or potential mother.

([47], p. 107)

5. Labelling and Agency Denials

As noted at the beginning of the article, the labelling of female killers is symbiotic to their agency denial. That is to say, labelling women denies the recognition of their ability to make a semi-autonomous decision to act in a particular way, and vice versa. More specifically, labelling women who kill as mad, bad or victims, denies the recognition of their ability to have made the semi-autonomous decision to kill their victims. All three of the labels used for women who kill deny the agency of these women in slightly different ways.

Labelling women who kill as victims denies their agency because the concepts of agency and victimisation are understood in opposition to one another. As explained by Mahoney;

In our society, agency and victimisation are each known by the absence of the other: you are an agent if you are not a victim, and you are a victim if you are in no way an agent. In this concept, agency does not mean acting for oneself under conditions of oppression; it means being without oppression, either having ended oppression or never having experienced it at all.

[48]

The consequences of utilising victimology theory when labelling women who kill are noted by Belinda Morrissey, who remarks that;

[I]n emphasising victimhood, intentionality or agency is neglected. Representations of the murderess as victim, then, function to deny her responsibility, culpability, agency and often her rationality as well, in their bid to explain her behaviour ... While undeniably often successful in securing reduced sentences, the disadvantages of such a strategy outweigh the benefits in terms of improving general societal attitudes to, and challenging negative myths and stereotypes of, women.

([3], p. 25)

This denial of women’s agency can be seen in the discourse surrounding battered women who kill.

Battered women are just that; battered. Therefore they ‘[a]re not seen to act, on the contrary they are the battered, the products of the batterer’ ([3], p. 96). The utilisation of the phrase battered women who kill removes the agency of such women because these women killed their partners only as a direct response to being battered by them. ‘The woman herself is neatly elided by the clash of the terms “battered” and “kill”’ ([3], p. 96). Labelling battered women who kill as victims and foregoing their agency, not only makes it easier to control them, but perhaps more importantly, it ensures the maintenance of the appropriate gender behaviour status quo. Indeed, as noted by Morrissey;

The campaign to allow BWS evidence into court may well have begun with the best of intentions, then, but the theory now seems to fast be becoming a straitjacket which tries to confine the realities of battered women and domestic violence within rigid parameters which do little to challenge society’s or the law’s understanding of spousal abuse, women’s violence, female agency and femininity itself.

([3], p. 78)

Women killers who plead infanticide or use BWS evidence to support a plea of diminished responsibility and are labelled as mad have their crime acknowledged ‘[w]hile removing the agency and responsibility for its commission’ ([3], p. 34). Indeed within the law more generally, the utilisation of pathological discourses often does not recognise the ability of an individual to make decisions for themselves. For example, under the Mental Capacity Act 2005 ‘[a] person lacks capacity in relation to a matter if … he is unable to make a decision for himself in relation to the matter because of an impairment of, or a disturbance in the functioning of, the mind or brain’ [49]. The use of ‘psy’ discourses presents women who kill ‘[a]s not intending the deed, as not knowing or understanding that they are committing it, as experiencing nothing in relation to it’ [50]. Using the mad label for these women relies on the discourse of irrationality that is readily associated with femininity. Indeed as noted by McColgan; ‘[f]emininity has traditionally been associated with irrationality, impulsiveness and weakness’ [51].

It is submitted that by focusing on the influence of women’s mental state or their biological functions, the mad label denies these women’s agency over their actions, rendering them harmless. When women are labelled as mad, their responsibility and agency is automatically rejected. The denial of agency for filicidal women labelled as mad is reflected in sentencing. ‘[O]f the 49 women convicted of infanticide between 1989 and 2000, only two were jailed; the rest were given probation, supervision or hospital orders’ [52]. As explained by Wilcynzski, this lenient sentencing reflects the belief that filicidal killings by women who plead infanticide are ‘[a]bherrant “tragedies” for which they are not responsible ... they need “help to come to terms with” what they have done’ ([24], p. 424).

Labelling women who kill as bad denies their agency in a subtly different way to constructing them as victims or as mad does. The agency of bad women is denied;

[b]y insisting upon the evil nature of the murderess, thus causing her to lose humanity. She is transformed into a monster from outside society threatening the mainstream, rather than one of its members, produced and enabled by her social and cultural milieu. The agency denial which takes place in this technique is specifically that of human agency. The murderess is considered to have acted, but not as a human woman.

([3], p. 25)

As explained by Morrissey, a murderous woman labelled as bad is ‘[n]ot just monsterised but transformed into the living embodiment of mythic evil through her relation to figures traditionally interpreted in this way’ ([3], p. 25). Therefore her agency as a human and as a woman is denied, with any agency that she is afforded being ‘[t]hat of a character from a familiar stock story’ ([3], p. 25). Therefore, bad women who kill do not have human agency.

The agency denial of Myra Hindley is perhaps most illustrative of this point, with her portrayal as the icon of evil and more specifically ‘[t]he feminine face of evil’ ([44], p. 14). As a result she was considered to be ‘[b]eyond femininity and humanity [and was consequently placed] into a realm of mythical monstrosity’ ([36], p. 42). Indeed, writing on Hindley continually utilises the monster imagery to describe her, with headlines such as; ‘Myra Hindley, the Moors Monster, dies’ [53], ‘The Monster Body of Myra Hindley’ [54] and descriptions of her as being ‘[M]edusa-like’ [54]. It is clear then that that the vivid dichotomies of the good and bad, human and inhuman woman and the continued reference to lack of adherence to ‘[i]deological norms of female behaviour’ [55] combine to refuse bad women agency.

It is clear that each of the labels; mad, bad and victim, deny the agency of women who kill in slightly different ways. However, it is submitted that despite these slight differences in how these women’s agency is denied, there is only perhaps one acceptable explanation for why these agency denials occur. One contentious explanation put forward for these continued denials of female agency is that;

[f]emale criminals are relatively unusual when compared to the numbers of male criminals, and concepts of a “reasonable woman” have, therefore, been deemed unnecessary. This means that women’s responsibility and agency is not automatically presented, as is the case with men.

([3], p. 169)

This explanation is difficult to digest, not least because it suggests that as women are not ‘major-players’ in the criminal justice system their experience is somehow of less importance.

Another perhaps more realistic explanation was given by Morrissey. She explained that;

Denials of female agency ... are crucial to decreasing the threat women killers pose to the dominant male-dominated institutions of heteropatriachy. If a woman can be found to have been so victimised that she did not know what she was doing when she killed, or if she is portrayed as a mythic, inhuman personification of wickedness, then the radical implications of her acts are muffled, her challenge to oppression nullified, at least as far as the dominant purveyors of cultural meaning are concerned. She is returned to her place of passivity and silence.

([3], p. 170)

This makes it clear that it is easier to give explanations for the actions of murderous women than to recognise their ability to have made a semi-autonomous decision to act in the way that they did. Indeed, it is certainly arguable that giving women agency over their murderous actions would disturb and challenge established gender norms. However, it is submitted that continuing to deny the agency of female murderers arguably presents far more serious issues than merely challenging gender stereotypes.

6. Problems with Denying the Agency of Women Who Kill

The discussion on the different labels applied to women who kill demonstrates how women who have committed essentially the same crimes can be viewed differently depending on the construction of their crime, their gender and their sexuality. It has become obvious that there is a correlation between the label given to female killers, their treatment within the criminal justice system and more broadly the social responses to their actions. Despite the differences in the treatment of these women depending on how they are labelled, it is clear that all three of the labels deny the agency of, and are consequently uniformly damaging to, the women they are attached to. As noted by Frances Heidensohn;

What is so striking about all of these images of deviant women is how profoundly damaging they are, once attached to any particular woman or group of women. Amongst them all, there is no conception of the “normal” exuberant delinquency characteristic of males. Any women would be damaged by being portrayed as a witch or a whore; and while a “sick” female deviant may be less punitively treated, she will attract other stigma.

([35], p. 95)

Indeed, it cannot be denied that using these labels to depict female killers, whether using them correctly or not, perpetuates and entrenches feminine gender stereotypes within both society and the law. The use of these labels may allow individual women, in particular circumstances, to win their battle but they do little to allow women to win the war against having to conform to appropriate feminine behaviour or asserting their individual agency.

6.1. Issues of Justice for Women Who Kill

Another issue that arises from the use of the above labels and denials of agency is that of justice. That is whether justice is actually being done, or indeed whether it can be seen to have been done [56], when female killers are labelled in this way and have their agency denied. When women commit violent crimes more questions are asked of, and simultaneously more explanations are made for, the violent actions of these women. This is because women are processed by the criminal justice system ‘[i]n accordance with the crimes which they committed and the extent to which the commission of the act and its nature deviate from appropriate female behaviour’ ([25], p. 306). This is particularly the case with women who kill. When these women are tried for their crimes there is ‘[a] tendency for [their] trials to be turned into trials of their character and the extent to which they accorded with appropriate femininity’ ([57], p.16). This gendered dimension to the trial process reinforces gender stereotypes and denials of women’s agency, in turn creating a form of gendered criminal justice. This form of gendered justice does not just focus on the murder committed by the woman in question, but also the degree to which her behaviour and often her lifestyle have deviated from appropriate feminine gender behaviour.

This gendered justice was most recently evident in the sentencing of Magdelena Luczak and her partner for the murder of her son, Daniel Pelka. In her sentencing comments, although the judge acknowledged that both Magdelena and her partner breached their position of trust as parents to Daniel, she explicitly referenced Magdelena’s failings as a mother. She emphasised; ‘Your breach of trust Magdelena Luczak is wholly irreconcilable with the loving care that a mother should show towards her son’ and ‘[y]ou, Magdelena Luczak, were fully complicit in these acts of incomprehensible cruelty towards your own son …’ [58]. Although both Magdelena and her partner were given the same prison sentence, the fact that particular focus was placed on Magdelena’s deviance as a mother demonstrates how the concept of justice for women who kill takes a gendered form. Magdelena was not just being sentenced for murder, but arguably also for breaching her primary social identity of a mother.

Justice also differs for women who kill depending on the label attached to them and the way in which their agency is denied. This is most prevalent in cases of women who kill their children. As Huckerby explains; ‘[n]ot all criminal mothers are subject to the same treatment by the criminal justice system … more punitive treatment is delivered to those women who do not meet the ideal norms of “motherhood”’ ([59], p. 151). Filicidal women who successfully plead infanticide and have their actions pathologised are generally treated with a degree of leniency and sympathy. A mad mother has her agency denied as she is not considered to know or understand what she was doing when she killed her child. Therefore her ability to have made the semi-autonomous decision to act in the way that she did cannot be recognised because she was acting in a moment of madness. As a result, her actions ‘[a]re characterised as isolated and contained incidents that can be easily altered through medication and therapeutic treatment’ ([59], p. 166). It is important to re-emphasise here that despite the Infanticide Act being specific as to the requirement of puerperal psychosis for a successful plea of infanticide, the ‘[c]oncept and scope of madness in infanticide cases is deliberately nebulous, so that judges, juries, and the media can selectively draw upon it to provide leniency for women whom they believe deserve sympathetic treatment’ ([59], pp. 160–61).

In contrast, bad mothers are often treated much more punitively within the criminal justice system. The agency of bad mothers is denied through their placement within a realm of monstrosity which denies their humanity and thus their human agency. A bad mother is ‘“[d]epraved” … “ruthless, cold, callous, neglectful of [her] children or domestic responsibilities, violent …”’ ([59], p. 158). Her actions cannot be pathologised and therefore the act of killing her child which is “[c]onsidered so antithetical to the behavioural norms of motherhood [is used] to justify the “demotion” of status from “mother” to the prematernal state of “woman”’ ([59], p. 151) and finally to that of monster, thus denying her agency.

The selectiveness with which the justice system can draw upon the concept of madness in cases of women who kill their children means that if a filicidal woman’s case either cannot be constructed, or is not perceived in such a way that she has her agency denied through being labelled as a mad mother, it will be done through labelling her as a bad mother. It is clear then that the way in which filicidal women are labelled and how their agency is denied directly affects their treatment within the criminal justice system. Consequently, a woman who kills her child would arguably fare better being diagnosed with a recognised psychological disorder, and having her actions pathologised (even if she does not meet the threshold of puerperal psychosis) in the hope of being treated more leniently within the justice system. If she does not succeed in her quest to be labelled as a mad mother, the alternative label of a bad mother awaits, with the potential for a harsher punishment and an altogether different agency denial.

It is not just for women who kill their children that justice differs depending on how they are labelled and the way in which their agency is denied. The consequence of labelling and agency denial often results in either arguably very lenient, or extremely harsh punishment for any women who kill, with no clear middle ground existing between these two extremes. The recent case of Nicola Edgington is perhaps most illustrative of this point. Nicola Edgington killed her mother in 2005 and was consequently diagnosed with paranoid schizophrenia, with a prominent mood disorder. As a result she successfully pleaded guilty to her mother’s manslaughter by reason of diminished responsibility. She was detained indefinitely under the Mental Health Act 1983, a clear acknowledgment that Nicola was suffering from a mental disorder at the time she killed her mother. Despite her sentence of indefinite detention in a psychiatric facility, Nicola was released three years later as she was no longer considered a danger to the community. In October 2011, Nicola attacked Kerry Clark and killed Sally Hodkin and was subsequently found guilty of murder and attempted murder after the jury rejected her plea of diminished responsibility. On 4^th March 2013 Nicola was sentenced to a minimum of 37 years in prison [60].

During her trial for murder and attempted murder in 2013, psychiatric evidence was presented declaring that Nicola was indeed suffering from an abnormality of mental functioning. However, the jury concluded that any such abnormality did not meet the requirements for diminished responsibility. Therefore the court concluded that her mental abnormality did not substantially impair her ability to form a rational judgment, or to exercise self-control. Sentencing Nicola, the judge acknowledged that she suffered from a ‘mental disability’, but accepted the jury’s findings that there was not a convincing case ‘[t]o conclude that the abnormality reduced [her] culpability to any significant extent’ [60]. This seemingly drastically reduced any weight that the judge attached to the mitigating factor of Nicola suffering from a mental disorder. Moreover, in his sentencing report the judge recognised several aggravating factors, including ‘[p]remeditation, and a determination to overcome failure in order to achieve [her] ends’ and the fact that the attacks were ‘unprovoked and random’ [60]. He also explained that he could not ‘ignore the fact that Nicola had killed before’ [60]. In contrast to the case against Nicola in 2006, the judge in 2013 made it clear that Nicola was more culpable for her actions.

Comparing the two homicide cases brought against Nicola, several things become apparent. In the first case in 2006, Nicola was arguably labelled as a ‘mad woman’ by the court, as she was suffering from a mental abnormality which ultimately denied her culpability for killing her mother. Consequently the court felt that she needed treatment, rather than punishment. In contrast, in the 2013 case, Nicola was labelled as a ‘bad woman’ who was legally culpable for her murderous actions, and consequently needed punishment rather than treatment. This is despite her obvious on-going mental disorder, which in itself presumably required further treatment. What is clear then is that the responses in both cases are at the opposite ends of the spectrum. It seems then that the current law on murder and manslaughter, when being applied to cases of women who kill, sits best when working at extremes, rather than focusing on a more measured middle ground.

For Nicola, this had the consequence that her actions were pigeonholed in such a way that although her agency was denied in both instances, she was either labelled as ‘mad’ and arguably treated leniently, or as ‘bad’ and was treated punitively. The ‘bad’ label does not seem to be prepared to acknowledge or incorporate, to any significant degree, a defendant with some form of mental disorder. Similarly, the ‘mad’ label arguably fails to acknowledge any significant degree of culpability for the defendant’s actions and limits any punishment. Pigeonholing Nicola into being labelled and treated as either ‘mad’ or ‘bad’, when she arguably falls into both categories to some degree, arguably demonstrates the need for a clearer middle ground for female defendants in cases such as these. This middle ground could go some way to being filled with an approach by the criminal justice system which acknowledges the agency of women who kill.

Battered women who kill their abusive partners face specific justice based issues when they are labelled as victims. Although labelling them in this way denies their agency over their murderous actions, it simultaneously emphasises the responsibility these women have in becoming victims in the first place. Indeed, as noted by Lorraine Radford;

The topsy turvy justice of patriarchal law puts women on trial for their own victimisation. Thus … questions asked in courts of battered women who kill emphasise women’s own responsibility for prolonged victimisation. Why don’t battered women leave their abusers? Why are they abused so many times?

([15], p. 177)

Therefore, it is argued that although these women do not have agency over their own actions, they are deemed to have some responsibility for the actions of their abusive partners. Focusing on battered women’s responsibility in this way refutes ‘[s]ociety’s complicity in the killing and the situation which helped precipitate it’ ([8], p. 735), as well as diverting attention away from the criminal justice system’s responses to these women.

As well as being held responsible for their own victimisation, battered women who kill must also conform to prescribed forms of ‘victim appropriate’ behaviour in order to secure justice, as noted earlier in the article. As explained by Radford, this appropriate behaviour and the life-history scripts which are written for these women are done so by ‘[p]rofessionals and medical experts within and behind the scenes of the courtroom’ ([15], p. 195). Women who are truly the victims of their abusive partners must arguably have their agency denied in their life script, before it is denied through being labelled as mad or a victim. Therefore, the deserving, and arguably non-agentical victims include, ‘[t]he upper middle class man’s ideal bride … “good mothers”, “good wives”, “good housekeepers”, “good heterosexual servicers”…’ ([15], p. 195). In contrast, women who may be perceived as asserting some agency within their life script by attempting to fight back against, or resist their partners’ abusive behaviour are not really battered. These ‘virago’ women have their agency ultimately denied when they are labelled within the criminal justice system.

6.2. Issues of Justice for Their Victims

Denying the agency of women who kill also presents issues regarding justice both being done, and being seen to be have been done, for the victims of the crimes committed by these women. One such example, which highlights the point most dramatically, is that of filicidal mothers who are able to plead infanticide, despite not suffering from the required puerperal psychosis. These women have their agency denied and consequently often receive a non-custodial sentence, usually a probation order, despite the fact that they have murdered their child. This does not sit well with societal expectations of justice, which usually requires those who commit murder to be imprisoned for a significant period of time. Indeed research has found that ‘public support for the life sentence [increases] in relation to the seriousness of the crime’ [61]. It is submitted that women who are erroneously able to utilise the defence of infanticide are quite literally ‘getting away with murder’ as a result of their agency being denied. Therefore their victims are not getting the justice that they and the rest of their family deserve. It should be noted here that it is not being suggested that these women should not be able to utilise another defence, such as diminished responsibility. It is simply being suggested that they should not be able to utilise the defence of infanticide if they are not suffering from puerperal psychosis, or as a minimum, suffering from a serious mental disorder, akin to that of puerperal psychosis, which allows for more lenient treatment within the criminal justice system.

Linked into this issue of victim justice is the fact that denying female criminal agency directly denies the existence of female violence. Despite the fact that female killers ‘[a]re relatively unusual when compared to the numbers of male criminals …’ ([3], p. 169) case studies such as those outlined throughout this article demonstrate that women are indeed capable of extreme violence. As it is therefore impossible to say that such cases do not exist, denying the agency of these women through labelling them allows an explanation to be invoked which goes some way to denying the propensity of women for violence. These labels and the consequent denials of agency which occur fails to give credence to the notion that women’s violence ‘“[i]sn’t always personal, private, or impulsive, that sometimes it is … a means … of furthering an ambition … a vehicle to her own empowerment”’ ([3], p. 153). This, as Patricia Pearson notes, has the effect of demeaning;

[t]he right our victims have to be valued. And it radically impedes our ability to recognise dimensions of power that have nothing to do with formal structures of patriarchy. Perhaps above all, the denial of women's aggression profoundly undermines our attempt as a culture to understand violence, to trace its causes and to quell them.

([3], p. 176)

7. Concluding Remarks

This article has acknowledged the symbiotic relationship between labelling women who kill as either mad, bad, or a victim and the continuous denials of their agency. Labelling female killers as a victim denies their agency by portraying them ‘[a]s so profoundly victimised that it is difficult to regard them as ever having engaged in an intentional act in their lives’ ([3], p. 25). Female killers who are labelled as mad have their agency denied by acknowledging the crime they have committed ‘while removing the agency and responsibility for its commission’ ([3], p. 34). Labelling female killers as bad eliminates their agency by suggesting that ‘[a]lthough the action took place, the actor was not a human woman but a personification of evil’ ([3], p. 34). This denial of female agency presents a number justice based issues for the women themselves, their victims and the criminal justice system. It also ensures the continued reinforcement of gender norms within both legal and social discourse.

Focusing on the symbiotic relationship between labelling and agency denials in the context of women who kill allows a largely under-utilised approach to be taken when engaging with women who commit violence. Typically the focus is on men who commit violence, however acknowledging the problems that labelling and agency denials have for these women, highlights and simultaneously admonishes the belief that women are not capable of extreme violence. Therefore, acknowledging women’s agency solidifies the notion that women are indeed capable of violence through recognising their ability to make a semi-autonomous decision to commit violent acts, in the context of this article, to commit murder. Doing so is important not just for securing justice for their victims, but also for the women themselves whose treatment and position within the criminal justice system is arguably often overlooked and underplayed.

In order to take account of some of the concerns raised within this article surrounding issues of justice which are raised when labelling and agency denial occur, it is submitted that reform is required within both the criminal justice system and the criminal law. Initially the criminal justice system needs to end the judgment of women according to their adherence to, or deviance from, social and gender norms, instead focusing only on the crime that they have committed. In turn this would allow for less focus to be ascribed to the labels which are currently attached to women who kill and which deny their agency. It is submitted that the concept of agency within the criminal law and particularly the relationship between women and agency needs further exploration and analysis within the academic literature. This could be done through reviewing a range of case studies of women who have been convicted of murder, considering the labels which were attached to these women, the way in which their agency was denied and the consequences that this has had for both these women and their victims. Doing so will affirm the premise that acknowledging women’s agency can, and indeed would, exist in harmony alongside the aims and principles of the criminal justice system and the criminal law.