Domestic Violence and the Gendered Law of Self-Defence in France: The Case of Jacqueline Sauvage

Abstract

Legal responses to battered women who kill have long animated scholarly debate and law reform activity. In September 2012 after 47 years of alleged abuse, Frenchwoman Jacqueline Sauvage fatally shot her abusive husband three times in the back. The subsequent contested trial, conviction for murder, unsuccessful appeal and later presidential pardon of Sauvage thrust the French law of self-defence into the spotlight. The Sauvage case raises important questions surrounding the adequacy of the French criminal law in this area, the ongoing proliferation of gendered stereotypes in law and the need for reform. In the wake of the Sauvage case, this article provides a timely analysis of the gendered law of self-defence in France. Drawing from an in-depth analysis of the judgments imposed in the Sauvage case, this article examines the adequacy of French legal responses to battered women who kill and ignites an argument for further law reform.

Introduction

On September 10, 2012 after 47 years of alleged psychological, physical and sexual abuse, Jacqueline Sauvage fatally shot her abusive husband, Norbert Marot, three times in the back (Transcript 3/12/2015).¹ It was alleged at trial that Marot, described as a ‘violent alcoholic’, had abused his wife, three daughters and son for over four decades prior to his death.² Following a contested trial, Sauvage was convicted of murder and sentenced to ten years imprisonment, an outcome that was upheld on appeal. The Sauvage case became a cause célèbre in France (Blaise 2016), reviving debates about domestic and gendered violence, and highlighting the inadequacies of the French law of self-defence. In an extraordinary move, following a petition signed by over 400,000 persons, in December 2015 Sauvage received a partial presidential pardon and in December 2016, following the court’s earlier refusal to release Sauvage on parole, she was granted a complete pardon (Bulman 2016). The decisions are examined in further detail throughout this article.

Since described as ‘a symbol of the scourge of domestic violence in France’ (Lichfield 2016, see also Wang 2016), Jacqueline Sauvage came to represent the largely silenced victimisation of women victims of domestic violence across France. In 2014 alone, 134 women in France died as a result of domestic violence (Ministry of the Family, Children, and Women’s Rights 2015). The Sauvage case demonstrates why scholars have long argued that the criminal justice system, a system made for and by men, can be a site of re-victimisation and injustice for women victims of domestic violence (see inter alia Hamilton and Sheehy 2004; Hopkins and Easteal 2010; Hudson 2006; Naffine 1990; Smart 1989). In the wake of the case, there have been calls from feminist advocates, legal practitioners and politicians for reform of the law of self-defence to better cater for persons who kill in response to prolonged domestic violence.

Beyond contributing to a broader body of scholarship examining legal responses to battered women who kill, an analysis of the Sauvage case offers important lessons in the context of the French criminal justice system. The passions triggered by the Sauvage case emerge at a time where voters have become particularly distrustful of its government and were also situated in a highly charged environment of widespread fear caused by the January and November 2015 terrorism attacks in Paris. This article provides a timely in-depth analysis of the Sauvage case and offers a critical examination of the concept and application of the French law of self-defence in cases involving domestic violence as well as a broader critique of the gendered lens with which the law continues to operate.

This article is structured in four parts. Each part highlights how the French law of self-defence, its application and attempts to reform it, together reinforce gender stereotypes of abused women. Part 1 analyses the Sauvage case and provides an examination of the current law of self-defence in France as it is set out in legislation and applied in the Sauvage case. The article then provides a critical examination of the adequacy of French legal responses to battered women who kill. Part 2 then gives an overview of the trial, conviction and presidential pardon of the case. In Part 3 we analyse the barriers faced by battered women who kill, including the lack of police reporting, the difficulty of producing evidence of abuse at trial, and the ongoing proliferation of gendered stereotypes throughout the inquisitorial criminal court process. Finally, in Part 4 the Sauvage case and law reform in this area are placed in the political context of penal populism. Political efforts to amend the law on self-defence may have less to do with an empathetic step towards battered women than with popular demands to punish violent men. These reactions, we propose, may reinforce rather than challenge gender disparities.

Defences to Murder for Battered Women Who Kill in France

Under French criminal law, a defendant is not criminally liable for his or her actions if committed in self-defence. Article 122-5 of the Penal Code states:

a person who, faced with an unjustified attack on themselves or a third person, simultaneously commits an act necessary to legitimate defence, shall incur no criminal liability except where the means employed are disproportionate to the seriousness of the attack.

In such cases the act of lethal violence is excused from legal sanction, both conviction and sentence, because the defendant reasonably believed that he or she had no alternative but to injure or take the assailant’s life. In this respect, the law of self-defence befits the paradigmatic case of a one-time fight between two men of equal size and strength where one, acting in self-defence, injures or kills his assailant (Cass. Crim. 16 October 1979).³

French courts have been quite strict in how the different requirements of self-defence are interpreted and applied, in particular with regard to the immediacy element.⁴ Where there is a time gap between the original unlawful assault and the defendant’s response, the courts have tended to interpret the latter as motivated by revenge rather than self-defence. Self -defence has been more narrowly interpreted as requiring an immediate response on the part of the defendant. Similarly, in domestic violence cases, the proportionality and immediacy requirements as set out in Article 122-5 of the Penal Code are often difficult to meet (Herzog-Evans 2014). Why would a woman use such extraordinary violence to kill her husband? Why would she react after having endured so many years of violence and not before? These problematic questions reflect a misunderstanding of the reasons why persons kill a prolonged abuser and invite a reconsideration of the ‘vantage point’ from which reasonableness is judged, as explained by Hopkins and Easteal (2010, 132):

[v]iewed from the perspective of the average judge or juror, uninformed about the dynamics and effects of domestic violence, the killing may appear entirely unreasonable; as either irrational or retaliatory. However, from a battered woman’s perspective – having lived with serious abuse under the constant threat of violence, having developed a heightened capacity to perceive danger from her batterer for whom escape has failed or is not a realistic option- there may have been no other reasonable alternative.

An analysis of French case law suggests that an interpretation of the actions of battered women from this vantage point is rare, if not altogether missing, and that few women have successfully raised the defence of self-defence.⁵

Beyond the complete defence of self-defence, and in comparison to the criminal law in England and Wales, which legislates for a partial defence of loss of control,⁶ there is no ‘equivalent’ partial defence of loss of control or provocation in French law (Elliott 2011, 236). Traditionally known as the ‘crime of passion’ defence (Ancel 1957), provocation operated in the 1800s in France as a complete defence to murder, resulting in an acquittal, in cases where a man killed his wife after observing her committing adultery. However, a series of reforms to the Criminal Code in the 200 years since codification, first restricted the defence to reduce culpability to manslaughter (but not allow for a complete acquittal) and since 1975 the partial defence was abolished and loss of control and provocation became factors to be taken into account at sentencing (Elliott 2011). This resulted in a situation where, at the time of the Sauvage case (and indeed, for the three decades prior), there is no partial defence to murder in French law that recognizes a defendant’s loss of control (Elliott 2011).

The Trial, Conviction, Appeal and Presidential Pardon of Jacqueline Sauvage

What is the truth here? Can one actually believe that Sauvage, while holding and firing that rifle the day after her son hanged himself and after years of sexual and other abuse to herself and children, was acting rationally and in entire control of herself? (Servidio-Delabre 2016)

Jacqueline Sauvage met Norbert Marot when she was 15 years old. At 16 years old she fell pregnant and they married. Over the 40 years of their marriage they had four children, three daughters and a son. On the 9th September 2012 Sauvage’s son committed suicide. The following day, Sauvage and Marot argued over business-related matters. In the early afternoon, Sauvage decided to take a nap,⁷ following which she alleged that she was woken up violently by her husband at around 3 p.m., asking her to prepare him some food (Transcript 24/10/2014). A violent altercation ensued; Marot allegedly hit Sauvage across the face and tore off her chain necklace (Transcript 1/12/2015). It was alleged that he then confiscated her keys and credit cards (Transcripts 3/12/2015) and took a bottle of whisky out to the house terrace where he proceeded to drink (Transcripts 24/10/2014, 1/12/2015). It is unclear when exactly the killing happened; Sauvage thought it was around 3 pm but the shots were only heard around 7 pm (Transcripts 24/10/2014; Transcript 3/12/2015). Sauvage describes how she ‘lost control’ and ‘exploded’; shooting her husband three times in the back with a hunting rifle, killing him. At trial she recalled having closed her eyes at the time of shooting (Transcript 24/10/2014). Her act, the defence later explained, was the result of 47 years of domestic abuse (Transcripts 28/10/2014; 3/12/2015).

Following the killing, Sauvage was indicted for murder with premeditation. Her case was referred to an investigative judge, the juge d’instruction, who questioned witnesses, interrogated Sauvage, ordered searches, and consulted experts.⁸ At the conclusion of that investigation, the investigative judge referred the case to the criminal court, the Cour d’Assises.⁹ The trial focused on determining whether Sauvage, then 65 years old, had intended to kill Marot, whether the act was premeditated, and whether the victim was her spouse, which amounts to aggravating circumstances under French criminal law.¹⁰ If premeditation was established, Sauvage faced a sentence of life imprisonment. On 28 October 2014, after 3 days of hearings Sauvage was convicted of murder without premeditation and sentenced to 10 years imprisonment. The conviction signalled that her actions were deemed by the court to be disproportionate to the threat posed (Watkinson 2016), given that in the period immediately prior to his death Marot did not pose an immediate danger.

On appeal the question of self-defence was bought into focus. To argue that Sauvage’s criminal act was not an immediate reaction to Marot’s assault the prosecutor drew attention to the lapse of time between the early afternoon assault and later killing. It was posed that because her husband was sitting with his back turned to her at the time of his death, she was not facing imminent danger and therefore it was not fear that triggered her act. The prosecution argued Sauvage ‘executed’ her husband (Transcripts 28/10/2014; 3/12/2015). The defence on the other hand argued that Sauvage’s act was committed in self-defence following four decades of abuse.

While the fatal actions of Sauvage were not perpetrated in the ‘traditional’ self-defence circumstances prescribed in French law, the defence poised that they must be understood on the backdrop of the morning’s abuse and the abuse suffered throughout the marriage. The defence alleged that in the moment immediately prior to the attack ‘a lightning strike went off’ in her head (cited in Symons 2015). When placed in the context of over 40 years of abuse, the defence justified that the scope of the immediacy requirement for self-defence should be expanded to incorporate the actions of this abused woman. In putting this argument to the court, the defence relied on Canadian case law, R. v. Lavallée, [1990] 1 S.C.R. 852, in which the Canadian Supreme Court recognised battered woman syndrome for the first time.¹¹ Lavallée reinterpreted the law of self-defence in Canada (see Sheehy 2014 for further analysis of the case).

Encouraging the court to understand Sauvage’s actions through the lens of a battered woman, the defence argued that the law did not require the act to be committed ‘at’ the exact same time as the assault but rather ‘in’ the same time, suggesting that it did not have to be concomitant but merely close in time. Finally, the defence argued that the level and degree of danger was so great, that in order to preserve her life and that of her children, Sauvage had no other choice but to kill Marot. Put differently, her reaction was proportionate to the danger she faced. Favouring the prosecution, on 3 December 2015 the appeal jury confirmed Sauvage’s sentence of 10 years’ imprisonment for murder (Transcript 3/12/2015). The court concluded that the elements for self-defence were not met, essentially because Sauvage’s act was not proportionate to her husband’s attack.

The decision to sentence Sauvage to 10 years’ imprisonment for murder provoked an unprecedented public outrage. Over 430,000 people signed a petition denouncing the court decisions, and called on the French President to pardon Sauvage (Johnston 2016). Sauvage’s three daughters led the request for pardon and the petition was signed by a number of media representatives, as well as politicians from the entire political spectrum. On 31 January 2016 President François Hollande partially pardoned Sauvage under Article 17 of the French Constitution.¹² The partial pardon did not quash the conviction for murder but did serve to reduce the remainder of Sauvage’s term of imprisonment by 2 years and 4 months and remove her prison tariff, thereby ensuring she would be entitled to parole in April 2016 (Watkinson 2016). Hollande’s decision to partially pardon Sauvage was particularly unusual given that the President stated during his 2012 election campaign that he would not use the power of presidential pardon, describing it as belonging to ‘a different concept of power’ (cited in Johnston 2016).¹³

Despite the partial presidential pardon, in August 2016 the sentence enforcement court refused to release Sauvage even though this decision was supported by both the defence and the prosecution (Agence France-Presse 2016b). The court held that Sauvage had not sufficiently ‘acknowledged’ and ‘reflected’ on the act she had committed. Her parole was further denied on the grounds that, in staying near her former home where she was getting support, she maintained a position of ‘victimhood’ (Libération 2016). Three months later, in December 2016 President Hollande granted Sauvage a full pardon ensuring an ‘immediate end’ to her imprisonment (as cited in Bulman 2016).

Legal Barriers for Battered Women Who Kill and the French Inquisitorial System

The Sauvage case is illustrative of the difficulties that domestic violence victims encounter in proving ongoing abuse and having their actions understood within an inherently masculine criminal justice system. As described by one of Sauvage’s daughters:

The court did not understand the distress, the powerlessness and despair of our mother. The court did not accept the actions of our mother, who could not find any other way to put an end to this daily violence (cited in Symons 2015).

While the limits of the law in understanding and adequately responding to battered women are well documented in adversarial criminal justice systems, such as the United Kingdom, the law’s response to women who kill in the context of prolonged family violence have been less explored in inquisitorial systems, such as France. In exploring the legal barriers that confront battered women who kill in France, we seek to fill this gap in current scholarship on the merits and limits of the ‘inquisitorial’ system.

To this end, and while countries no longer strictly belong to either an accusatorial or an inquisitorial system but rather include elements of both, the Sauvage case is a powerful illustration of France’s predominant inquisitorial style in criminal cases, in two specific ways. First, in the inquisitorial tradition, the court has an active and leading duty to seek the truth and in particular establish the defendant’s guilt, as opposed to the limited role the court plays under an adversarial system to decide whether the prosecution has proven the accusation. Second, a central difference between an accusatorial and an inquisitorial system lies in the place the defendant occupies in the criminal process. In France, the defendant is expected to contribute to the discovery of the truth. It follows that questioning the suspect plays a more central part in the proceedings than it does in an accusatorial approach. In cases of battered women who kill, the law on self-defence requires that two elements be proven: the domestic abuse and the reactions thereto. Yet, it may be harder for self-defence to be effective in an inquisitorial system because of the part played by judges and the place defendants hold in the criminal system. As is argued throughout this article, the part played by the court and the position in which the defendant is placed facilitate the proliferation of misunderstandings about responses to prolonged domestic violence of those abused, as well as the production of gendered stereotypes that are used to blame women victims.

Situated within this context of an inquisitorial approach to justice, the following section brings to the fore the problems faced by women in raising claims of domestic violence at trial, as crystallised in the law’s response to the actions of Sauvage. This discussion draws heavily on our in-depth analysis of the transcripts from the Sauvage case which reveals the ongoing relevance of long held concerns surrounding the ability of women to provide official evidence of their domestic violence and the difficulties faced by battered women who do not conform to traditional femininity and gendered stereotypes. The implications of both challenges in terms of undesirable legal outcomes and problematic narratives are explored.

Proving Domestic Violence: Police Records and Medical Evidence

Mirroring the challenge faced by many battered women, there were no official reports of the violence committed against Sauvage and her children. Subsequently at trial the defence were unable to produce any record of police complaints, attendance, an arrest or prior conviction on the part of Marot to support Sauvage’s alleged history of abuse victimisation. Evidence presented showed that Sauvage was however, known to the local hospital where she had been on four occasions between 2007 and 2012 for abuse-related injuries (Symons 2015). Throughout the trial Sauvage’s daughters and neighbours also provided evidence of abuse (Transcripts 27/10/2014; 1/12/2015; 2/12/2015). In the absence of official reports of domestic violence, the presiding judges and the prosecution during both trials questioned why Sauvage had never reported the abuse to the police, filed a complaint nor asked for help from friends or other family members.

The psychological expert testifying explained that it ‘was probable that justice was never perceived as something that could provide help but rather only as something that could sanction’ (Transcript 27/10/2014). On appeal, Sauvage’s son’s former partner sought to draw attention to the difficulties of reporting domestic violence:

I would like to add, with regards to domestic violence, you may believe it is easy to file a complaint. The son reproduced the same pattern as his father. I’ve never been as much beaten up as [Jacqueline]. We do not get much help. He had threatened me to death. The police station in Orléans recorded my complaint. But it is not true to claim that we get help! When the complaint is recorded, the men are released…and the husbands come home, and they hit you again. I left because I am from a younger generation. Mrs Marot, she’s from the older generation. There is so much fear… These people destroy you from within. It is moral harassment, they condition us… (Transcript 1/12/2015)

Similarly, in her case, Alexandra Lange reminded the court that the police refused to record her complaint on the grounds that the claims of insults and abuse were ‘insufficiently violent’ (Benetti 2015). This reflects a long-held recognition in research on domestic violence that an absence of evidence of abuse should not be interpreted as evidence of an absence of abuse. On average the Ministry of Social Affairs (2015) estimates that 223,000 women face domestic violence each year in France, while the organisation Fédération Nationale Solidarité Femmes (2015) reported receiving 38,972 telephone calls in 2014 from women who had experienced different forms of abuse. According to the Ministry of the Family, Children and Women’s Rights (2015), only 14 per cent of victims of domestic violence file complaints with the police. Of those that do engage the police, few abusers are convicted (Ministry of Social Affairs 2015, 7; see further Herzog-Evans 2014).

The French inquisitorial system provides an additional layer of investigation led by an investigative judge. This special judge may dismiss the case on evidential grounds; a discretionary decision which research suggests can present particular barriers to justice for women victims of sexual and gendered violence. Jo Lovett and Liz Kelly’s (2009, 52) comparative research on rape cases, for example, stresses that the level of attrition at this pre-trial stage is significant. For those who have killed their abusers and who seek to rely on the law of self-defence, proving a history of domestic violence is key and an investigative judges’ discretion to dismiss domestic violence claims may impair their eligibility to the complete defence of self-defence.

Outside of the French context, research has long documented the reasons why women may not engage the police or other authorities when victimised by a current or former intimate partner. Indeed, recent research confirms that ‘victims of domestic and partner abuse remain among the least likely to report their victimisation to the police’ (MacQueen and Norris 2016). Reasons cited for why a victim may not report domestic violence include fear of gender bias and discrimination, a fear that the violence will escalate following criminal justice intervention, feelings of isolation, a prior negative interaction with the police and/or perceived lack of support from the police (Bradfield 2002; Corrigan 2013; Douglas 2008; MacQueen and Norris 2016; Meyer 2011; Stewart 2001). Furthermore, Australian research examining women victims of domestic abuse reveals a perception among victims that the police will only help persons who fit within the strict confines of the archetypal victim (Douglas 2012). Consequently, persons who fall outside of this ‘ideal victim’ stereotype remain hesitant to engage a justice system that they perceive will be unwilling and/or unable to provide a meaningful response. The impact of negative police interactions evident in the Sauvage case, suggest that even though it has been over a decade since domestic violence was criminalised in France, the policing experiences of women like Jacqueline Sauvage demonstrate that while a change in legislation has occurred, there is still a need to ensure this is translated into a change in practice.

The Sauvage case also provides an important reminder that the difficulties of engaging the justice system for victims of domestic violence are not overcome following police and other pre-trial investigations. While acknowledging that Sauvage as well as her children had been exposed to Marot’s violent behaviour, judges and prosecutors shared their doubts as to the frequency and intensity of his violence, notably because of the lack of physical and medical evidence (Transcripts 24/10/2014; 28/10/2014; 3/12/2015). Two of Sauvage’s daughters testified they had been raped but their testimonies were called into question for failing to provide scientific evidence of such physical assaults (Transcripts 27/10/2014; 1/12/2015; 2/12/2015). During the initial trial, the presiding magistrate noted the expert had found traces of a blow to the defendant’s lip but ‘nothing else significant’ on the rest of her body had been identified (Transcript 28/10/2014). Similarly, the lawyer representing Marot’s sister’s interests highlighted that Sauvage had ‘only’ been hospitalised twice in the 47 years of her marriage, which, according to the witness, called into question the reality of the risk of violence she claimed to have experienced (Transcript 28/10/2014). In addressing the jury, the prosecutor relied upon this evidence to highlight the scarcity of the ‘traces’ of such violence (Transcript 28/10/2014).

The difficulty of evidencing domestic violence in the criminal court system is not unique to the Sauvage case or the French context. Similar concerns have been raised in previous research (see, inter alia Mayaud 2006; Schneider 2000). The tendency for women to ‘explain away’ injuries as accidental to health professions and family and friends compound the difficulties that battered women face in explaining their actions to the court. In France courts require strict causal evidence between proof of bruises and physical violence when domestic abuse is alleged. In 2006 for instance, the Cour de Cassation confirmed the lower court’s decision to acquit a man accused of domestic violence on the grounds that the medical certificate, while evidencing bruises and contusions, did not demonstrate the injuries had been committed by the accused.¹⁴ It is also particularly difficult to demonstrate such violence when it is psychological rather than physical and for this reason, in the 5 years since the introduction of a psychological abuse offence in 2010,¹⁵ there have been few convictions secured (Benetti 2015). This is a particularly interesting point given the spread of similar offences globally, whereby other comparable jurisdictions have in recent years introduced separate offences for psychological abuse in an attempt to improve legal responses to the myriad of behaviours that can constitute a domestically abusive relationship. In England and Wales, for example, in December 2015 a new offence of controlling or coercive behaviour was legislated¹⁶ with the aim of improving the likelihood of successful prosecutions in cases where physical violence was not present or was not the only form of domestic abuse experienced by the victim (Fitz-Gibbon 2016). Within this context, the French experience broadly and the Sauvage case more specifically, highlights that legislation for psychological abuse in and of itself is unlikely to improve victim experiences and/or access to justice. Such reform must arguably be accompanied by specialist training and broader evidentiary reform.

In the Sauvage case disbelief over the extent and frequency of the abuse experienced permeate the transcripts and reflect a lack of understanding of women’s options and ability to safely dissolve a violent relationship (Mahoney 1991). On a number of occasions, the magistrates presiding over both courts, the prosecutors and the legal representative for Marot’s sister, questioned why Sauvage had remained with her abuser through decades of abuse inflicted upon her and her children (Symons 2015). Those involved in the case repeatedly questioned why she would persist in living and working with a man who physically and psychologically assaulted her and sexually abused her children (Transcripts 24/10/2014; 3/12/2015). For these legal practitioners, the defendant had sufficient power and agency to put an end to the abuse. She could, and even should have left. According to this representation by those within the criminal justice system, Sauvage willingly chose to remain a passive victim. The presiding judge and prosecution shared their confusion and discontent with the jurors in the case (Transcripts 24/10/2014; 28/10/2014).

Within an inquisitorial system, the power lies in the hands of the court, and the defendant holds a central place in the search for the truth. At the heart of the court’s inquisitorial investigation into the actions of Jacqueline Sauvage and her entitlement to a defence of self-defence lie the proliferation of gender stereotypes about domestic violence victims and a problematic construction of who qualifies as a ‘good’ domestic violence victim and who does not. Through this lens only those who record and report are legitimate to be treated as true victims (see also Gotell 2002). The Sauvage case is illustrative of the ways in which the reactions of an abused woman become the very standards by which her criminal actions and eligibility to self-defence are measured. Protections in law place the responsibility on the victim to perceive and report criminal victimisation, and to participate actively in the enforcement of proceedings (see also Bumiller 1987; Gotell 2002). It is assumed that those deemed to be protected can and do accept the burden of proving they are victims. In essence, the system operates from the position that those who deserve the protection are those who have sought it. The desire to protect abused women goes hand in hand with a certain broader gendered understanding of women in society; the weak, the ones in need of protection. Only those women who conform to certain gendered conceptualisations become eligible to a more lenient and protective treatment under the law, which includes the eligibility to self-defence.

Commenting on the case, Honorary Magistrate Bilger (2016) argued there was ‘no shame in being surprised’ by Sauvage’s lack of reaction. He also wondered why Sauvage had remained ‘passive’ and never tried to leave or resist, but instead chose to kill her husband after years of abuse (Bilger 2016). Similarly, criminal lawyer Florence Rault (2016) questioned the defendant’s inertia and delayed reactions to years of psychological abuse and violence. It is ‘difficult’, Rault (2016) writes, ‘to envisage that a woman could have forgotten during 47 years what she would have endured’. Such remarks are highly concerning and illustrate a lack of understanding of the experiences of domestic violence victims by those within the French criminal justice system. They also contrast significantly with the image of Sauvage advanced by the defence, who described a woman under the control of her abusive husband, who had suffered the ‘irreversible consequences of violence against women’ (cited in Symons 2015).

While it is not uncommon for evidentiary questions and credibility doubts to arise in battered women cases, the questions raised in Sauvage point to the ongoing difficulty that those within the legal system demonstrate in grasping the nuanced dynamics created between the abuser and a victim. The coercive grip and hold that an abusive partner gains over his victim is hard to conceptualise and does not fit into easily defined categories of behaviours (Stark 2007; Johnson 2008; Herzog-Evans 2014). Moreover, the normalisation of the brutality and its recurrence becomes a source of doubt. A common (but misinformed) assumption is that, had these women really been beaten, or at least as badly as they claim, they would have left the relationship (Herzog-Evans 2014).

In the Australian jurisdiction of Victoria, the government has acknowledged and attempted to address such problems through the introduction of evidentiary law reform and detailed juror directions on the nature and dynamics of family violence. The reforms, first introduced in 2005¹⁷ and updated in 2014,¹⁸ sought to ensure that evidence of family violence relevant to a homicide case where a person has killed their prolonged abuser can be heard by the courts. Evidence which can be admitted under the legislation includes that pertaining to the:

general dynamics of abusive relationships, the cycle of violence, the complex reasons women stay in violent relationships and why some women do not report violence and why a woman might plan to kill in order to protect herself (VLRC 2004, 173).

Through this legislation the Victorian evidence reforms aim to ensure the circumstances surrounding a woman’s use of lethal violence and the impact that those circumstances may have had in the case to be better contextualised for the court (Douglas 2012). The benefits of doing so are well captured by Hopkins and Easteal (2010, 132) who note that the Victorian reforms enhance court responses by ‘requir[ing] judges and jurors to walk in the shoes of battered women who kill in order to evaluate the reasonableness of their actions’. The Sauvage case highlights why such evidentiary law reform is needed in France to ensure that women’s experiences of violence are bought within the confines of the legal system. Beyond the experience in Victoria, scholars have long pointed to the value of expert evidence and family violence evidence provisions¹⁹ (see, inter alia, Ayyildiz 1995; Bradfield 2002; Hopkins and Easteal 2010; Sheehy et al. 2012b) and initial reviews of the Victorian approach have been overwhelmingly positive (see, inter alia, Douglas 2012, Hopkins and Easteal 2010).

The Emotionless Woman and the Bad Mother

The evaluation of certain aspects pertaining to Sauvage’s personality point to the highly gendered nature of the consideration of self-defence. Women who do not conform to traditional gendered stereotypes are unlikely to draw the sympathy and empathy of prosecutors, judges and jurors (Christie 1986; Fitz-Gibbon 2014; Medlicott 2007). In such cases a woman’s claim of prior victimisation can be undermined through the uptake of denigrating and victim blaming narratives that serve to reallocate responsibility along uncomfortable gendered lines. The power given to the court and the place the defendant occupies in the French inquisitorial system arguably exacerbate such gendered assumptions.

An analysis of the Sauvage transcripts reveal the emphasis on Sauvage being an experienced hunter, capable of handling rifles and bullets, and killing a prey (Transcripts 24/10/2014; 28/10/2014; 1/12/2015). ‘This is a woman who is used to handling guns’, the prosecution argued during the first trial (Transcript 28/10/2014). Not only was her credibility as a legitimate victim of domestic violence repeatedly questioned but Sauvage was also portrayed as someone capable of violence herself. For example, at both trials testimony of Marot’s former mistress was presented which claimed that Sauvage had assaulted her when she discovered her husband was having an affair. Further countering the defence’s positioning of Sauvage as the real victim in the case, the prosecutor reminded the court she had also been disciplined while in pre-trial detention.

Throughout the legal processes Sauvage is represented as someone who is determined and authoritarian and who valued her work success and social status. When addressing the jurors on appeal, the prosecutor emphasised this:

Madam Sauvage is someone who is determined. Since her youngest age, she opposes her mother, her father and her five brothers. She continues to have this determination. She has a family. Then a house. But this determination is also the source of her unhappiness. She cannot go back. She has four children, she continues to work. In this company, she becomes assistant collaborator. She is in charge of deliveries. She then launches a wine trade activity. (Transcript 3/12/2015)

Further legitimating this characterisation, the presiding magistrate on appeal underscored that prison staff had described Sauvage as an authoritarian person. The implication here was that a woman who is authoritative, hard-working and determined cannot then lay claim to being a passive and helpless victim of decades of abuse.

Cementing her characterisation as an illegitimate victim Sauvage is described as emotionless. She is critiqued for neither shedding a tear nor showing emotional distress throughout the court process (Transcript 28/10/2014). For Marot’s sister’s lawyer, ‘revealing one’s emotion is not a sign of weakness but rather that of one’s humanity’ (Transcript 28/10/2014). The implication was Sauvage did not possess the latter. But perhaps even more condemnatory than her representation as an illegitimate victim, is Sauvage’s representaton on repeated occasions throughout both trials as a bad mother, one who repeatedly failed to protect her children from abuse. The magistrate presiding over the first instance trial questioned why Sauvage had kept her children within her husband’s reach ‘in making them work with you, you brought them back to the “tyrant”, as you describe him. This is something I struggle in understanding’ (Transcript 24/10/2014). The magistrate then raised the tone of her voice and asked: ‘is it protecting your children when remaining in such a situation?’ (Transcript 24/10/2014). The lawyer for Marot’s sister directly blamed Sauvage for keeping her children within their abuser’s grasp. To the court, she openly accused Sauvage of being a bad mother, ‘[t]he duties of Mrs Sauvage were to protect her children. She is an accomplice in failing to report the aggressions committed on her children’ (Transcript 28/10/2014). Similarly, the prosecutor asked ‘either Mrs Sauvage did not believe her children, as suggested by some testimonies, or she did believe them and that’s terrible. What should be thought of a mother who does nothing?’ (Transcript 28/10/2014). This response reflects a problematic narrative of victim blaming and responsibility transfer that has been long documented in child protection cases, where a battered mother is accused of failing to protect and/or neglect amidst allegations of willingly exposing a child to prolonged domestic violence (The “Failure to Protect” Working Group 2000). The misrepresentation of responsibility in such narratives is, we would argue, highly concerning and signals another point at which the French criminal justice system fails to adequately understand the gendered dynamics of family violence and the experiences of women victims. Interestingly, while the prevalence of gendered victim blaming narratives in criminal and civil trials has been documented extensively in other jurisdictions, there is little French based research on this topic. The Sauvage highlights the proliferation of such narratives in the French context and the need for further research to examine the extent to which this occurs more broadly.

Drawing this together, we posit that the courts in the Sauvage case express messages not just about certain forms of proscribed behaviours but also about perceptions of women in French society. As an experienced hunter, determined worker and worst still, a bad mother, Sauvage transgressed traditional female gender roles and failed to meet the ‘normal’ standards of conventional femininity (see, inter alia Carlen and Worrall 2004). Each of these gendered stereotypes—which, in passing, were expressed by a majority of women legal practitioners²⁰—arguably adversely affected the full consideration of her claim to have acted in self-defence in killing her husband as her actions were represented and understood through a lens of victim blaming and denigration. The problematic nature of such gendered legal narratives is tied to the inquisitorial style of seeking the truth in criminal cases. The way Sauvage was interrogated and the place she was expected to hold during the oral debates eclipsed her abuse. The system makes it harder to appreciate the nature and extent of the violence she experienced; the defendant stands before the court first and foremost as a criminal, not a victim. The creation of problematic gendered legal conceptions extends beyond the application of the defence of self-defence in the courtroom. They also emerge within a broader socio-political context. Within this setting, to date, efforts to reform the law of self-defence fail to cater for, and understand women’s experiences.

Responses to the Sauvage Case: Penal Reform and Penal Populism

Penal Reform and Gender Misconceptions

In the wake of the Sauvage case, legal advocates, political stakeholders and members of the community have called for reform of the law of self-defence to better accommodate the circumstances within which battered women kill. Several approaches to reform have been proposed, including re-examining the notion of ‘reasonableness’ and reviewing the immediacy requirement in French self-defence law. This includes a proposal for new legislation by Parliamentarian Valerie Boyer, aimed at improving legal responses to battered women who kill (Blaise 2016; Boyer 2016). The bill provides that, subject to medical expertise, a battered woman who kills shall not be held criminally liable if they, ‘at the time, and due to repeated domestic abuse, suffered a mental disorder that altered her discernment or impeded the control of her acts.’ From a broader principled perspective, the French activist group Osez le Féminisme [Dare to be Feminist] has called for an expansion of self-defence to better represent cases of ‘female victims of violence’ (Watkinson 2016), noting that the current requirement for proportionality is unjust when applied in battered women cases (Wang 2016). Critiques have qualified these political moves as attempts to promote a form of feminism, one that treats every woman as a victim and blankly rejects that some may be violent (Rault 2016). The purpose of justice for Rault (2016) is not to assist a particular cause. Rather it is expected to judge and eventually condemn transgressions of the public order. Not all women are victims, Rault claims. In her words, ‘women are not systematically victims of everything and responsible for nothing’ (Rault 2016, also cited in Wang 2016).

The proposed bill’s focus on medical expertise and a pathologization of battered women’s experiences of violence may be troubling for many. This critique does not in and of itself, however, suggest a need to revert to the status quo. Rather, we would argue, it belies the need to look to other jurisdictions’ experiences of reforming the law of self-defence to better cater to women’s experiences of violence. To this end, in reforming the law of self-defence, France is fortunate that it can draw on the experiences of international comparative jurisdictions who have already implemented reform to counter similar concerns. For example, law reform enacted in Canada in March 2013 broadened the notion of self-defence and moved the law away from a justification-based approach.²¹ Under the new approach while the timing between the killing and the conduct that caused the defendant to act in self-defence remains a factor to be considered by the court in assessing the reasonableness of the accused’s actions, it is not a necessary requirement. In Australia reforms introduced over the last 15 years have sought to both clarify and extend the law of self-defence to better accommodate the circumstances within which persons kill a prolonged abuser (for reform details see further Crofts and Tyson 2013; Hopkins and Easteal 2010; Fitz-Gibbon and Stubbs 2012). While neither jurisdiction is inquisitorial like France, law reforms introduced signal ways in which traditionally masculinised justice systems can reform to meet the needs of battered women. Across both countries, reformed laws aim to better recognise that domestic violence cases often involve recurrent and continuous acts of violence resulting in continual fear. To understand the critical moment at which battered women revert to violence, the law must acknowledge the cumulative effect of months or years of brutality. Through this lens, whether the victim gauged the requisite amount of force needed to repel the attack and act accordingly, is determined from a broader temporal perspective.

The Sauvage case also demonstrates why any reform to the laws of evidence or homicide in France must be accompanied by judicial and legal practitioner education and training on the dynamics of domestic violence and the nuances of women’s experiences of violence. An analysis of the Sauvage transcripts reveal the proliferation of misunderstandings of the impact of domestic violence, the subsequent behaviours of those abused and the permeation of gendered stereotypes that serve to blame the battered woman. Beyond the Sauvage case the importance of practitioner and judicial professional development to ensure that those who operate within the criminal court system are well resourced to understand the circumstances within which men and women commit violence has been recognised by a range of legal and criminological scholars (see, inter alia Fitz-Gibbon 2015; Hopkins and Easteal 2010). As Crofts and Tyson (2013, 893) argue:

Legal change without social change will be ineffective. Clearly more education around the dynamics of family violence is needed if both the legal profession and wider community (e.g. juries) are to fully utilize important changes to the law in a way that more justly recognises women’s human rights.

Without such judicial and practitioner education, there is a real concern that while the laws may be passed in France, they will fail to achieve a meaningful change in legal practice. As Sheehy (2014, 304) notes in her analysis of Canadian self-defence laws, the implementation of the law post-reform is beholden to ‘advocacy and judicial interpretation’.

Beyond the Sauvage case specifically, this call for reform of French legal responses to battered women who kill comes at a time when violence against women has emerged as a critical global issue. The 2014 Istanbul Convention, the Council of Europe Convention on preventing and combating violence against women and domestic violence, affirms the need for signatory countries to prioritise responses to violence against women and girls to better protect women and improve gender equality. The Convention was ratified by France in July 2014 and came into effect from 1 November 2014. Recent political attention to the issue builds on over five decades of feminist activism in France and elsewhere, which has been long focused on improving the law’s response to violence committed by men within the ‘private’ confines of an intimate relationship.

Since the mid-2000s, a number of reforms in France have been introduced to address domestic violence and to improve legal responses in this area (Huguin 2013). Since legislation introduced in July 1992 to legally recognise that domestic violence is a crime, successive governments have introduced removal orders for violent partners, a new offence of psychological violence, reforms to expand the definition of domestic violence to include relationships beyond married spouses and the 2012 legislation on sexual harassment between partners (for further information on these reforms, see Huguin 2013). Outside of legislation, efforts have also been made to improve the training of first responders, the police for example (Sénat 2015). The Government Action Plan of 2015 also aims to improve services providing assistance, guidance and support, and to ensure better coordination between social workers.²² As part of this initiative, 70 therapists were recruited to work alongside police officers in different stations across the country (Ministry of Interior 2016). On top of these reforms, in May 2012 the government introduced a Ministry for Women’s Rights, demonstrating their commitment to addressing violence against women nationally (Huguin 2013). This range of activity suggests an increasing political desire to appear active in preventing and responding to domestic violence. This is a significant improvement from the pre-1992 era when such violence was not even considered a criminal offence.

These efforts contrast with neoliberal approaches to tackling sexual violence (Gotell 2008; Bumiller 2008). At the heart of neoliberalism lies the idea that individuals are rational and fully responsible actors. Rather than being considered a social and political problem, that is intrinsically tied to gender inequalities, sexual and gender based violence has been redefined as an individual issue that is best managed through responsibilization and self-regulation. This has created new forms of victim blaming and reduced the power of feminist voices and activists. In Hollande’s socialist France, efforts have been made to tackle sexual violence as a political and social problem. However, the drive behind such reforms, as illustrated by the Sauvage case, is essentially populist. They have paved the way to reinforce, rather than fundamentally change, gendered conceptions about domestic violence. The increased attention, we argue, has been less about the abused than about the abuser, and in so doing, further eclipses victims of gender based violence.

Domestic Violence and Penal Populism

The decision to sentence Sauvage to 10 years’ imprisonment for murder provoked an unprecedented public outrage. Over 430,000 people signed a petition denouncing the court decisions, and called on the French President to pardon Sauvage and change the law on self-defence (Johnston 2016). Here we consider the extent to which the wider political climate of penal populism in responses to domestic violence is gendered. We argue that the intense public reaction and ensuing political reforms in response to the Sauvage case sheds light on the ties between public demands and gendered penal politics in the context of domestic violence.

Penal punitiveness or penal populism is often described as the introduction of excessive and disproportionate penal policies (Simon 2000; Garland 2001; Pratt 2007; Wacquant 2002). One key feature in this rise of punitiveness is that penal politics are no longer shaped by criminal experts or practitioner knowledge but rather dominated by the voice of the ill–served and long-serving public. The fears and emotional reactions of the public to crime have often served as a justification for the implementation of coercive measures. Taking crime seriously means ensuring public safety and offering well-deserved retribution to the victims of a crime. To this end, politicians, essentially conservative, have sought to introduce ways to imprison entire categories of offenders, whether violent or not, and for extended periods of time (Feeley and Simon 1992; Bottoms 1995).²³

Against this hegemonic backdrop, French penal reformers have generally remained insulated from popular pressures and demands. In 1981 for instance, Robert Badinter, then Minister of Justice under socialist president François Mitterand, introduced a bill to abolish the death penalty, disregarding the widespread public support for this extreme sentence (Badinter 2000). After the Second World War, human rights and humanist conceptions of punishment and justice predominate penal reforms (Baranger and Bougeneaux 2005). In the late 1990s, a shift towards increasingly punitive penal policies anchored in victims’ rights rhetoric was observed. Lévy (2004) speaks of the ‘sacralization’ of the victims and its effects on penal reform in France. French politicians, like their counterparts across the Channel, have relied on and instrumentalised popular fears to introduce policies to gain electoral support. Attempts from the right and left of the political spectrum to criminalise migrants in a context of globalised terrorism attests to such penal populist politics.²⁴

While it is beyond the scope of this article to explore whether French politics are increasingly embracing a form of penal populism, the political reactions to the Sauvage case raise interesting questions in terms of the ties between penal reforms and popular anxieties. Some may perceive these political efforts as providing support, recognition or retribution for those who stand up to their abusers—a positive step forward in populist responses to violence against women. When the public requires Sauvage to be pardoned or when politicians suggest improving legal responses for battered women, they are essentially asking that such men not benefit from the protection and full retribution of justice. To punish these men, the public proposes to pardon their killers or to improve the defence of self-defence. Put differently, to punish violent men more, reformers propose to punish women less. In this sense, these public reactions to the Sauvage case, and to domestic violence more generally, depict a form of penal populism, one that calls for less rather than more, and is intrinsically gendered. Indeed, these penal politics illustrate an instrumentalization of the abused women’s culpability to achieve something else, namely to punish violent men. The emphasis and centrality given to violent men, to the fact they are underserving of justice’s full retribution, further contributes to eclipsing victims of abuse. This form of penal populism may thereby reinforce rather than challenge gender disparities in the context of domestic abuse. Contributing to works on the framing of sexual violence by neoliberalism (Gotell 2008; Bumiller 2008), this argument illustrates how the recognition of sexual violence as a political problem may nonetheless exacerbate gendered inequalities. Political efforts to tackle the issue as a global rather than individual problem may still distract the attention away from the victim. In unveiling the driver behind calls for reform, namely a growing intolerance towards violent men, our analysis of the Sauvage case illustrates how progress and reform eclipse the very minority that is sought to be protected and how problematic gendered assumptions about domestic violence are thereby entrenched.

Conclusion

One week after the partial pardon was granted in the Sauvage case, a French criminal court imposed a suspended five-year jail sentence on Bernadette Dimet, who was convicted of committing voluntary violence leading to accident death for the 2012 shotgun killing of her abusive husband (Agence France-Presse 2016a). The Dimet case mirrored that of Sauvage, Dimet had married her partner when she was 16 years old and they shared two children. It was alleged that over the course of the marriage he psychologically abused her and forced her to have sex with him and raped other women known to the family (Agence France-Presse 2016a). While the prosecutor requested an eight-year term of imprisonment be imposed, the suspended sentence and its timing points to a renewed willingness of the French criminal justice system to bestow sympathy and leniency upon persons who kill in response to prolonged family violence. It is not however enough. Such defendants must be protected by law and by a framework of law that can adequately recognise and understand the nature of violence committed between intimate partners. To achieve this, we argue reform of the French law of self-defence and evidence is needed.

In Sauvage a presidential pardon was required to circumvent a legal process that arguably did not, and could not, justly respond to the actions of Jacqueline Sauvage. The community’s outrage and political recognition of the injustice in this case must now be used as a momentum for reform of a law that continues to privilege and represent only the circumstances within which men kill. While Parliamentarian Valerian Boyer has proposed new legislation to improve legal responses to battered women who kill, it remains to be seen what direction reform will take and whether it will curry favour political and public forums. Drawing from an in-depth analysis of the transcriptions from the Sauvage case this article demonstrates why such reform is urgently required. It highlights the impact of practitioner misunderstandings of family violence and gendered narratives of victim blame evident in French criminal law. In doing so it builds on a significant body of prior scholarship that has demonstrated the ways in which male centric criminal justice systems fail to accommodate women’s experiences of violence.