Getting Away With Murder? A Review of the ‘Rough Sex Defence’

Several high-profile murders of women killed during alleged consensual sex ‘gone wrong’ have led to widespread calls for reform to prevent the use of what has been termed the ‘rough sex defence’. Concerns about the use of this ‘defence’ are located within broader concerns about the high rates of domestic abuse and fatal violence against women. Lobbyists, campaign groups and members of parliament have drawn attention to the increase in this ‘defence’ featuring in criminal cases in England and Wales and have consequently proposed two amendments to the Domestic Abuse Bill (2020), namely a statutory prohibition of consent as a defence to actual bodily or more serious harm, including death, and introducing additional scrutiny in charging decisions by requiring the Director of Public Prosecutions to authorise charges of manslaughter (rather than murder) in cases involving rough sex/sadomasochism (SM). This article provides a critical analysis of the use of rough sex/SM in female homicide cases and proposed legal reforms and concludes that the proposed reforms would fail to capture many of the ‘rough sex’ cases that have come before the courts in recent years and may not have the intended effect. We consider potential alternative approaches.


Introduction
The case of British backpacker Grace Millane, who was strangled to death in New Zealand by a man she had recently met on a dating app, dominated international headlines in 2019. The man charged with her murder was ultimately found guilty and convicted in November 2019, after a trial in which he alleged the killing was accidental and happened during consensual rough sex. 1 Despite burying Millane's body and hiding evidence of her death, watching pornography immediately after killing her and taking intimate 'trophy photographs' 2 of her dead body, the defence claimed that Grace died after consenting to being choked during sex. The case was the latest to draw attention to what has become known colloquially as the 'fifty shades defence' 3 that men who kill women during or immediately after sexual activity rely on to argue against a charge of murder. Concerns about the use of this 'defence' are located within broader concerns about the high rates of domestic abuse and fatal violence against women. It is estimated that, globally, at least a third of women will experience some form of physical and/or sexual violence during their lifetime. 4 In England and Wales, the latest data from the Crime Survey for England and Wales indicate approximately 1.6 million women (and 786,000 men) experienced some form of domestic violence in the previous 12 months. 5 Domestic violence is gendered, not only terms of prevalence but also its nature and consequences. This is most notable in relation to fatal violence. In the UK, three quarters of victims of domestic killings (typically referred to as domestic homicides 6 ) are women, equating to approximately 2.5 women each week. In the vast majority of cases, the perpetrator is a male partner or other family member. 7 The rates of domestic homicide have risen in recent years, with recent analysis indicating a five-year high in the number of domestic violence killings in 2019. 8 Violence against women is, globally, one of the leading preventable contributors to death and illness for women aged 18-44. 9 There is limited national data on the extent of homicide cases where rough sex/sadomasochism (SM) forms part of the defence case. The campaign group We Can't Consent To This (WCCTT) have been collecting data on national and international cases where rough sex or SM is a feature and have traced the claim of rough sex as a contextual feature in murder cases back to the 1970s, but observe that it remained relatively rare to see this claim made until 2010. Since then, they have identified at least a 90% increase in cases where rough sex or 'sex game gone wrong' has featured in the defence's account; they estimate at least 67 cases where this has been used and, significantly, all involved a male suspect/offender. 10 Strangulation was the most common cause of death (66 per cent of cases). Significantly, perhaps, in at least a third of cases the suspect/offender had previous convictions for violence against women, including murder, rape, attempted murder, kidnapping and assault.
1. <https://www.theguardian.com/world/2020/feb/21/grace-millane-man-jailed-for-life-for-killing-of-uk-backpacker> accessed 7 May 2020. 2. <https://www.thetimes.co.uk/article/grace-millane-murder-jury-hears-closing-arguments-dp6hjmb6s> accessed 7 May 2020. 3. <https://www.thetimes.co.uk/article/killers-go-free-thanks-to-fifty-shades-defence-6l9s5976q> accessed 7 May 2020. 4. World Health Organisation, 'Global and Regional Estimates of Violence This method of killing as well as the broader context of death occurring during or immediately after sexual activity is thus heavily gendered and reflects wider homicide trends; strangulation as a method of killing in domestic/intimate partner homicide has remained constant over the last three decades and is the primary method of killing of a female partner in heterosexual relationship. 11 In most cases tracked by WCCTT (60 of the 67), the victim was female. In terms of outcomes, WCCTT observe that 37 of the 60 cases involving a female victim resulted in a murder conviction (62 per cent), and in 17 cases, the outcome was a manslaughter conviction (28 per cent). In the remaining cases, five resulted in either no charges (two cases) or a not guilty verdict (one case), one case was dropped due to cause of death evidence and in one case the accused was found guilty of a different offence (dismembering of the victim's body). One case was still outstanding. These figures seem to broadly reflect the national pattern for homicide indictments and convictions. 12 The Grace Millane murder is the latest in a series of high-profile cases where consensual rough sex or SM is raised by the defence. In the UK, the killing of Natalie Connolly by her partner John Broadhurst attracted widespread concern. He inflicted more than 40 injuries, including a 'blow out fracture to the left eye socket' and 'lacerations of the vagina which resulted in arterial and venous haemorrhage'. The latter was caused by the insertion of a bottle of carpet cleaner into her vagina. Broadhurst claimed he did not intend to kill Connolly, and the injuries were inflicted during consensual rough sex. This was accepted by the Crown Prosecution Service who charged Broadhurst with manslaughter, to which he pleaded guilty. 13 Women's rights organisations and lobbyists including the campaign group WCCTT subsequently called for a review into the use of this 'defence'. 14 Reforms to homicide law in England and Wales to address the use of this 'defence' have been proposed as part of the Domestic Abuse Bill currently before Parliament.
This article provides a critical analysis of the use of rough sex/SM in female homicide cases. The first part provides a brief overview of the law on consent to harm in England and Wales. The second part of this article examines the way rough sex/sex game gone wrong has been used in homicide cases involving female victims and the key concerns raised by lobbyists, the media and MPs. The third part of this article provides a critical review of the proposals for reforming the law to abolish the use of this 'defence' and offers suggestions for possible alternative approaches. This article concludes by arguing that the concerns raised about the use of the rough sex/sex game gone wrong narrative are valid, but the proposed reforms must be part of a wider, comprehensive package to change the cultural scripts and narratives which are used to blame women for the violence inflicted on them.

Consent in English Criminal Law
There is no single statutory definition or legal test of consent in English criminal law 15 ; as Simpson has pointed out, consent is a 'recurring fundamental concept in criminality that the law has struggled to positively define across the spectrum of criminal law'. 16 Consent, broadly, is relevant to a range of criminal offences; the victim's factual and legal consent or the defendant's belief in the existence of such consent can negate the actus reus of a number of criminal offences and/or provide a defence to liability.
Although consent applies in a number offences under English criminal law, it is arguably most central to two categories of criminal offences: sexual offences and offences against the person (fatal and non-fatal) and these two categories have been at the heart of debates around the law over the last three decades. 17 The Sexual Offences Act 2003 is the only statute to provide a definition of consent in criminal law. The provisions on consent only apply to offences under that Act. Under s 74 of the Act, 'a person consents if he agrees by choice, and has the freedom and capacity to make that choice'. Sections 75 and 76 set out the circumstances in which it will be presumed (conclusively or rebuttably) that consent is absent. 18 As lack of consent forms part of the actus reus of these offences, the burden is on the prosecution to prove that consent is absent and that the defendant did not have a reasonable belief in consent.
Conversely, there is no statutory (nor agreed common law) definition of consent for (fatal and nonfatal) offences against the person. As Simpson points out, 'the fact that the statutory definition of consent only applies to sexual offences, without an equivalent provision in non-sexual offences, has led to significant criticisms from criminal law academics'. 19 Likewise, the Law Commission has been critical of the lack of consistency in the conceptual boundaries and definitions of consent across criminal law. 20 It is somewhat less clear whether consent forms part of the offence in the case of offences against the person or whether it operates as a potential defence to a charge. Some 21 view lack of consent as an element of the offence, as the conduct only becomes wrong in a criminal sense if there is no consent. However, the majority view in Brown 22 was that consent is a defence, rather than its absence being a core element of the actus reus of fatal and non-fatal offences against the person. The basis for that view is that that violence towards another is prima facie unlawful, but that the victim's consent can provide a good reason for engaging in it. Brown is the leading authority for consent as it applies to offences against the person in England and Wales. The infamous case restricted 'the validity of consent by reference to the level of harm and the circumstances in which it is inflicted' 23 (original emphasis). The case established the threshold for the level of harm that an individual can give legally effective consent to is an assault or battery. If the victim suffers actual bodily harm (ABH) or a more serious levels of harm then consent does not provide a defence, unless the activity involved is one which courts or parliament have recognised to be in the public interest. These circumstances include tattooing/piercing, surgery, regulated sports, public exhibitions and religious rituals. Thus, it was established by the majority of their Lordships in Brown that at the point ABH is caused, the consent of the victim is no longer a defence to criminal liability unless the conduct falls within one of these exceptions. Subsequent case law has confirmed that these exceptions are finite and have been identified by law because of the discernible social benefits (e.g. sports and religious rituals) and/or because it would be unreasonable for the common law to criminalise the activity if the parties engaged consensually (e.g. tattooing and piercing). The recent case of R v BM 24 confirmed that new exceptions should not be created by the courts, unless the activity is closely aligned with an existing exception.
Consent to harm in these special circumstances, similar to consent to sex, signals that the behaviour (where consensual) has positive social value and as such the criminal law should not discourage it. 25 Conversely, outside those circumstances, the act is seen as lacking social value, even where consensual, 17 and so the law should discourage it. Rough sex or SM sits somewhat awkwardly between these two perspectives. On the one hand, the basic activity (consensual sex) is viewed positively and is protected from any criminal sanction. However, where the sexual activity involves (intentionally and perhaps even recklessly 26 ) some level of physical harm, the entire activity is reframed as prima facie violence and is thus subject to criminal prohibition.

The Application of Consent Laws in Homicide Cases Involving Rough Sex
The starting point then is that an individual cannot legally consent to ABH or serious bodily harm and, consequently, cannot consent to harm that results in their death. Even the dissenting judges in the Brown case were clear that some threshold as to when consent can provide a defence must apply but believed it should be set at grievous bodily harm (GBH). It is, therefore, of no surprise that in Wacker 27 the consent of the victims to being locked into a refrigerated lorry to enable immigration did not constitute a defence to a charge of gross negligence manslaughter. If their consent provided no defence to ABH, it certainly did not to death. 28 Murder, a common law offence, requires that the defendant, of a sound mind, unlawfully kills a human being under the Queen's Peace with intent to kill or cause GBH. 29 If a defendant kills during a sexual encounter and is charged with murder, they have two possible defences which need to be separated. First, they could argue that the act that caused the killing was lawful because it was consented to and, therefore, they lack the actus reus of the offence. For this claim to succeed it would need to be shown that, as well as consent, the conduct falls within one of the exempt categories set out in Brown. Then, the actus reus for the offence is unfulfilled. However, 'rough sex'/SM is unlikely to fall into one of the accepted categories and so the defence should fail.
Second, the defendant could claim that they did not intend death or GBH and so lacked the necessary mens rea. Therefore, the wider context of the conduct is relevant in establishing not only whether that conduct is unlawful but also the defendant's state of mind. Thus, in cases where consensual rough sex/ sexual activity occurs as part of the incident that results in the death of the victim, the sexual activity may be relevant in assessing the defendant's state of mind. Put simply, if the conduct was committed for the purposes of sexual pleasure and the conduct is therefore sexual, rather than violent, in nature this may be evidence that the mens rea requirement of intention to cause death or serious harm was not present. 30 However, the courts have been clear that if serious harm was intended, even if this was for sexual gratification, the mens rea requirement is satisfied. 31 The method of killing is also relevant to the assessment of mens rea. As Edwards 32 points out, in cases where a victim dies following strangulation the question for the jury to determine is whether the defendant intended (directly or indirectly 33 ) to kill or cause GBH. The mental element of the offence therefore centres on the defendant's foresight of consequences and the probability or likelihood that strangulation would result in death or GBH. This raises particular issues. First, strangulation does not always result in death, meaning it is not automatically the case that the defendant's actions can be assumed to demonstrate the intention to kill. Second, rather than using the term strangulation, the defence will reframe the action as 'pressure to the neck' or 'squeezing' or 'pushing down' in an effort to euphemise the behaviour into a less serious form of violence. As Edwards  been sympathetic to the assertion that strangulation is rarely intended and that the behaviour is more likely to be accidental or reflect careless disregard. 34 Thus, the central issue is not really consent; it is clear that consent is not a defence to murder (or harm that constitutes a level equal to ABH or GBH); the absence of consent in cases involving the death of a victim resulting from sexual conduct would be relevant only insofar as it would provide a defence to a charge of one or more sexual offences. Rather, consent is relevant to the narrative that the defence presents as to the circumstances of the death which may, if accepted, have the effect of reducing a murder charge to one of manslaughter because they lack the mens rea for murder or may even result in a not guilty verdict and thus acquittal of the defendant. This is considered further in the following section of this article.
Although the primary charge in prosecuted cases is murder or manslaughter, in several cases, sexual offences are also charged. For example, in Broadhurst, 35 D was charged with assault by penetration alongside murder and s 18 GBH with intent. Legal scholars and practitioners have voiced concerns about the application of consent in such cases, given the definitional and conceptual diversions across different areas of criminal law. The different consent rules create conceptual and operational confusion in cases where there are both offences against the person and sexual offences on the indictment relating to the same incident. Consent may be raised by the defence for both charges but must be legally and conceptually distinguished by the court. In cases where a sexual offence is charged, lack of consent forms part of the actus reus and thus the prosecution must prove that consent was absent/that the defendant did not have reasonable belief in the victim's consent. However, where consent is raised as a defence for the offence against the person charge, the defence must establish not only that the victim/complainant 36 consented but also that the case falls into one of the categories where consent is available as per the authority in Brown. As Murphy explains, the jury must be directed to consider consent in relation to sexual assault, but under Brown, must be directed to disregard it in relation to assault occasioning ABH. 37 Simpson concludes that, 'the divergence of the application of consent, both within and between different categories of offences, has had a negative impact on clarifying the concept, not only within sexual offences but within criminal law as a whole'. 38 This becomes potentially complex for the jury in cases where murder and sexual offences, for example rape, are on the indictment. If the jury accept that the victim/complainant could and did legally consent to the sexual activity, thus finding the defendant not guilty of the sexual offence, they then have to consider whether the circumstances of consent to the sexual activity provides sufficient evidence that that the defendant did not intend to kill or cause really serious harm to the victim. In practice, it may be difficult to find there was consent to the sexual activity but that the presence of consent does not amount to a lack of mens rea for the murder of the victim. To convict of murder, the jury would need to reason that although there was consent to the sexual activity, the activity involved ABH (or worse) and the defendant intended to cause GBH (or worse). As others have pointed out, the current consent principles if not 'absurd and unworkable' 39 require mental gymnastics from the jury.
A good example of how the law can operate is Broadhurst. 40 There the judge accepted that consent of the victim to the beating that caused injuries to her bottom and breasts was not a defence as the injuries involved ABH, based on the Brown ruling. 41  trigger mechanism was not unlawful on the basis of Slingsby. 42 In that case, the defendant had inserted his hand into the victim's vagina and caused a serious injuries as a result of a signet ring he was wearing. His defence, accepted by the Court of Appeal, was that he was not aware of the risk of harm and so, he believed he was engaging in a battery to which the victim consented. Although the consent itself was not a defence, the belief the act was a consensual battery (with no risk of more serious harm) was. It seemss astonishing the court in Broadhurst believed the defendant thought the insertion of the carpet cleaner and his unsuccessful attempt to remove it would only involve a battery. The outcome as the law applied in that case was that beating victim's bottom to cause bruises was unlawful but inserting a container of carpet cleaner into their vagina was not.

Narratives of Consent in Rough Sex Homicide Cases
Clearly in cases of murder there are particular evidential difficulties, as the victim is unable to give evidence about the incident and whether they consented. This, it has been argued, gives the defence an advantage, as they are able to present a narrative of sexual libido and desire 43 to frame the incident as sexual rather than violent, and it is difficult to challenge this because of the victim's inability to provide evidence. Laura Farris MP has argued that the victim has 'no voice' 44 and is unable to challenge the narrative presented by the defence. In most cases, the 'evidence' put forward by the defence is impossible to verify, as there are unlikely to be no witnesses to the alleged sexual activity (as the victim is deceased) and as such the evidence proffered by the defence concerning the victim's sexual preferences and history is purely inferential. 45 Edwards argues that the defendant is able to rely on the sexual consent narrative by manipulating and appropriating SM narratives 'to disguise what is essentially cruel and misogynist conduct'. 46 Harriet Harman MP has raised similar concerns, stating that 'men are using the narrative of women's sexual enjoyment of being injured to escape murder charges and face only manslaughter charges'. 47 Essentially, the argument that is being made here is that the defendant is able to reframe the incident in a way that obscures the seriousness and, crucially, their legal responsibility for the outcome. 48 Feminist scholars have described this in broader terms as 'euphemising'-a technique which enables male violence to be labelled and presented in a 'misleading way such as to obscure the seriousness or responsibility of whoever has committed it'. 49 This forms part of a broader 'identity work' strategy where, in the criminal justice system, 'accounts, vocabularies of motives, techniques of neutralization, and narratives-while distinct theoreticallycan all be used to construct and negotiate nondeviant identities'. 50 This includes discursive tactics to position the victim as responsible for their victimisation (victim blaming) while seeking to neutralise or legitimise the defendant's behaviour. These narratives and patriarchal stories 51 have been widely studied in the context of sexual and domestic violence cases 52 but less so in the context of fatal violence.
Several tactics are argued to underpin the production of this narrative as part of the identity work undertaken by the defendant and their legal representatives. First, previous sexual history is often relied on by defence to argue consent was given by the victim in the case, as part of the defence narrative. Feminist activists and scholars 53 have paid significant attention to the problems associated with the use of sexual history evidence (SHE) as it relates to sexual offence cases, leading to significant legal reform 54 which aimed to restrict the circumstances in which such evidence would be deemed relevant and subsequently adduced. The effects of SHE in sexual offence cases are multiple and significant. A range of studies examining public attitudes, professional attitudes and mock-juror decision-making have observed that SHE makes cases less likely to be prosecuted and, where they are, acquittals more likely as victims are considered less credible and more likely to have consented. 55 While the legal credibility of complainants may be harder to challenge due to contemporary procedural rules, legal scholars have argued that the moral credibility of complainants is powerfully strategised to reduce responsibility of the defendant and increase the responsibility, and blame, of the complainant. As McGlynn points out, in the context of sexual offences cases, the complainant's sexual history 'contributes to shifting the focus of the trial from the defendant's actions to those of the complainant, thereby also shifting legal and moral blame and responsibility from the defendant to the complainant'. 56 This narrative, that women are to blame for male violence, that they provoked it, asked for it, enjoyed it or should have done more to avoid it happening, are deeply woven into the patriarchal society and, despite legal and policy reform, continue to frame the public and, in some cases legal, discourse around violence against women. In homicide cases, it is argued that even where the defendant is found guilty of murder, the use of rough sex/sex game gone wrong as a mitigating narrative means that the reputation of the woman becomes the central focus and continues to stain the character of the victim beyond the conclusion of the trial. 57 Both scholars and activists have argued that consent is person-and situation-specific, and thus consent to an activity cannot be inferred from previous sexual activities or broader sexual preferences. Despite introduction of law aimed at limiting the use of SHE in sexual offences cases, the available evidence indicates that SHE is adduced in around one-third of trials with a high success rate for applications under s 41. 58 Multiple studies have documented that juror (and judge's) perceptions of complainants are influenced by the admission of SHE based on wider rape myths and stereotypes about 'real' rape victims and offenders and ideal/deserving complainants. 59 Thus, the prejudicial effects of SHE in sexual offence cases are well established. Similar research on the effects of SHE in homicide cases has not been conducted, however, the potential prejudicial impacts of the consensual rough sex 'defence' which relies on the victim's sexual history is a central concern of lobbyists and MPs seeking reform. Although s 41 may have reduced the use of SHE in sexual offence cases, in the context of homicide, s 41 does not apply. Consequently, there are few procedural limitations on the use of SHE in homicide cases involving consensual sex as part of the defence case. Although in some cases there is a lack of evidence, or weak evidence, that the complainant was an instigator of rough sex/SM, Edwards argues that the defence is able to rely on broader cultural scripts about women and sexuality to position the complainant as an enthusiastic and willing participant. 60 Indeed, such concerns have been raised in relation to previous cases. In both Wilson and Emmett, the defendant claimed the acts had been consented to yet neither partner was prepared to give evidence to support these claims. 61 In both cases, the police were alerted by the women's doctors (in breach of medical confidentiality). We note that one of the few reasons justifying a breach of confidence is that there is a fear of ongoing domestic abuse. 62 In Wilson, the act of a husband branding his initials on his wife, with the overtones of slave ownership, bears all the hallmarks of coercive control. As Herring 63 points out, the fact that both of these cases are referred to throughout legal textbooks and scholarly writing as cases of SM rather than domestic violence 'shows the ease with which a case of domestic abuse can be presented as sadomasochism'. This illustrates that, at the very least, 'caution' should be exercised before 'accepting an argument that the case is one of BDSM'. 64 The limited available evidence suggests participation in Bondage/Discipline, Dominance/Submission, and Sadism/Masochism (BDSM) is not equal; research has consistently shown that more men than women report arousal for fetishism and sadism and significantly more men than women report engaging in BDSM activity. 65 For women who have experienced BDSM behaviours, a significant proportion did not consent. For example, in a recent study commissioned by BBC Radio 5 live 66 asked 2,002 UK women aged between 18 and 39 if they had experienced various acts during sex. The majority (59%) had experienced slapping, 38% had experienced choking, 34% had experienced gagging, 20% had experienced spitting and 59% had experienced biting. Under half of the women (44%) said these acts were always wanted. However, 29% said they were unwanted some of the time, 14% said they were unwanted most of the time and 10% said they were unwanted every time. Clearly the amount of unwanted 'rough sex' is considerable.
Moreover, the number of men admitting they have engaged in behaviour not consented to (assault and potentially sexual assault) or have not obtained consent prior is significant; a recent study found a third of men admitting slapping, choking, gagging or spitting on their partner without first asking for consent. 67 Thus, there is insignificant evidence that SM/rough sex is instigated and enjoyed by equal numbers of women as men, but there is also evidence that women are routinely victims of violence during sexual activity that they did not consent to. As Edwards argues: As to the question of erotic asphyxia there is no evidence that it heightens women's sexual libido but there is evidence that men routinely use strangulation as a method of assault, that it is a trope and a reality in pornography, that women die in the course of it and that it is part of the misogyny narratives 68

A Critical Review of Proposals to Reform the Law
The emergence and use of consensual rough sex/SM by the defence in homicide cases has led to widespread calls by the public, campaign groups and MPs to review and reform the law to prevent the rough sex 'defence' being used in future homicide cases. In England and Wales, among those calling for action are Harriet Harman MP, Laura Farris MP and Mark Garnier MP who are pushing for the defence to be outlawed in the Domestic Abuse Bill under consultation in 2020. 69 Harman, along with others, has argued that the 'defence' is the 21st-century version of 'she was asking for it', a reference to previous gendered justifications for killing women that were facilitated in law by the previous provocation defence. Harman has argued that men are, in an ironic twist of fate, using advancements in gender equality and sexual liberation as a justification for killing women 70 and that 'men are now, literally, getting away with murder by using the "rough sex" defence'. 71 Laura Farris MP argued in a recent Parliamentary debate that 'acts of extreme violence are given a different complexion because they occurred during sex and it is said that the victim wanted it' resulting in a 'veneer of complicity through the sexual element'. 72 Following the latest parliamentary debate on the Bill, the government confirmed they have 'committed to ensuring the law is clear that this [rough sex] "defence" is unacceptable and are looking at ways to achieve this'. 73 We will explore the two amendments to the Bill sponsored by Harman, before exploring other proposals. At the time of publication of this article, the government have committed to outlawing the rough sex 'defence' through the Domestic Abuse Bill but have not yet indicated how they will do this and whether this will involve acceptance or amendment of the proposals put forward by Harriet Harman and other MPs, though early indications suggest there are concerns about the proposed wording allowing 'wiggle room' for defendants. 74 We will explore the two amendments to the Bill sponsored by Harman, before exploring other proposals.

Amendment One: Banning the 'She Was Asking for It' Defence
The proposed amendment reads: No defence for consent 1. If, in the course of any behaviour which constitutes domestic abuse within the meaning of this Act, a person ("A") wounds or assaults another person ("B") causing actual bodily harm, more serious injury or death, it is not a defence to a prosecution that B consented to the infliction of injury. 2. Subsection (1) applies whether or not the actual bodily harm, more serious injury or death occurred in the course of a sadomasochistic encounter.
Essentially, the argument here is that consent to the sexual activity that contributed to the death should never form the basis of a defence to homicide (nor indeed any offence involving ABH or GBH). This appears to reflect the law as established by the common law as set out in Brown and so it might be questioned whether there is any point in confirming it. Two reasons may be given.
First, some MPs including Harman have argued that Brown should be made statutory law to ensure it is applied, because 'Statute law is much more under the noses of the judiciary and the prosecutors and the defence'. 75 It might be argued that, especially given the controversy surrounding Brown and the extensive criticism is has received for its reasoning, a statutory confirmation of its standing would ensure the legal principle emerging from it still stands, even if much of the reasoning in the case has been decried. Second, it might be argued that there are some question marks over the extent of Brown. In R v Wilson, 76 one of the reasons given for overruling the conviction of ABH for a husband who branded his initials onto the buttocks of his wife was that consensual activity between husband and wife, in the privacy of their matrimonial home, was not a matter for criminal investigation or prosecution.
This appeared to take the view that the criminalisation of consensual SM in Brown did not apply to, at least, heterosexual married couples. That reading was challenged in R v Emmett 77 where Wright J held 'we can see no reason in principle, and none was contended for, to draw any distinction between sadomasochistic activity on a heterosexual basis and that which is conducted in a homosexual context'. Indeed, the alternative explanation that Wilson was a case of tattooing rather than SM became the dominant reading. However, the Court of Appeal in BM 78 recently seemed to confirm the suggestion that marital SM may be lawful saying of Wilson, 'This court concluded that consensual activity between husband and wife in the matrimonial home was not a matter for criminal investigation and prosecution under section 47'. 79 There is, therefore, some genuine doubt as to the extent Brown applies to married couples. The proposed amendment could, therefore, be justified in clarifying this uncertainty.
The problem, however, is that the Harman amendment only applies to behaviour which constitutes domestic abuse within the meaning of this Act. Most significantly, it does not, therefore, apply to couples who are not 'personally connected' as defined in clause 2. That includes, inter alia, those in 'an intimate personal relationship with each other' 80 but would not include those who are have met casually, such as the Grace Millane case. Indeed, the concern may be that the amendment implies that outside the context of domestic abuse (as defined in the Bill) consent is a defence in a case of sado-masochism.

Amendment Two: Limiting Prosecutorial Discretion
The other amendment promoted by Harriet Harman to the Domestic Violence Bill was as follows

Consent of Director of Public Prosecutions
(1). In any homicide case in which all or any of the injuries involved in the death, whether or not they are the proximate cause of it, were inflicted in the course of domestic abuse, the Crown Prosecution Service may not, in respect of the deatha. charge a person with manslaughter or any other offence less than the charge of murder, or b. accept a plea of guilty to manslaughter or any other lesser offence without the consent of the Director of Public Prosecutions. (2). Before deciding whether or not to give consent for the purposes of subsection (1), the Director of Public Prosecutions must consult the immediate family of the deceased.
It would appear that, for most of the advocates calling for complete removal of this 'defence', there is a belief that the narrative around rough sex/SM is influential in terms of prosecutorial decision-making (i.e. charging the defendant with manslaughter rather than murder and/or accepting a guilty plea to manslaughter) and/or juror decision-making. This amendment is clearly designed to challenge that. Any prosecutor minded to charge a case where there has been a death in the course of domestic abuse with murder rather than manslaughter. This provision is clearly wider than rough sex/SM cases and can apply in a case where the killing is the course of a domestic abuse case. Our discussion will be limited to its application in rough sex/SM cases.
The first point to note is that the clause is not as wide as some activist would seek. By limiting it to domestic abuse cases the clause does not apply to people who have no established relationship. It also may not apply to cases where the defendant claims the couple had a consensual SM relationship and there was no coercion or control taking place.
The second point is that it may have far less effect than hoped for. Charging standards and guidelines for prosecutors provides specific guidance on charging decisions for murder and manslaughter. Specifically, where a murder charge is being considered, prosecutors are required to assess whether there is sufficient evidence for a charge of murder giving due consideration to any partial defences and whether the mens rea for murder can be established. Where it is uncertain, prosecutors are required to consider whether a charge of constructive or gross negligence manslaughter is appropriate, either as the primary charge or as a separate, alternative charge. 81 Requiring agreement from the Director of Public Prosecutions (DPP) to charge a defendant with manslaughter rather than murder may provide additional checks and balances in the charging decisions of these cases; however, it will not necessarily lead to different charging decisions. Furthermore, DPP scrutiny would only partially address the 'problem' of manslaughter verdicts in rough sex homicide cases. In all murder cases, manslaughter is an alternative verdict which can be returned by the jury on a prosecution for murder. 82 Thus, even where the Crown Prosecution Service (CPS) prosecute the case as murder, there is no guarantee that the jury will return this as their verdict.
Consequently, although this recommendation appears to offer additional checks and balances and an opportunity to scrutinise CPS decision-making, it is not currently clear that the CPS decision-making in these charges is inappropriate. A review of charging decisions by the Inspectorate would offer an opportunity to identify whether alternative, reduced charges are being inappropriately applied in these cases. If that is the case, it may be that this can be addressed by specific guidance to senior prosecutors who are tasked with making charging decisions in homicide cases, particularly in relation to selection of charges and inappropriate reduction of charges. This should form part of a larger, comprehensive review of decision-making and case management of homicides where the defendant raises the consensual rough sex narrative.
As can be seen from these discussions, the proposed amendments to deal with death during rough sex/ SM were limited by the fact they were proposed as amendments to the Domestic Abuse Bill. While domestic abuse is an important context within which to consider unwanted rough sex/SM, the issues raised by the cases which trouble activists are much broader than the understanding of domestic abuse presented in the Bill. While it was understandable that a desire to change the law as quickly as possible meant that the first available opportunity presented in the Bill was seized, a full answer requires a full analysis of all the issues around rough sex/SM. However, others have cautioned that, simply restricting, or preventing, the use of SHE in homicide trials will not fully address the broader issues of victim blaming and character assassination that forms part of some of the defence' case. Instead, the defence will use a myriad of other behaviour traits. Tyson 89 has argued that there is a need to ensure the broader narratives relied on by the defence which discredit the victim based on gendered moral codes are eradicated, rather than simply displaced. Tyson argues 'there is an urgent need to ensure that insidious court narratives that portray the female victims of domestic homicide as to blame for their own deaths are no longer able simply to be deployed in the guise of other defences to homicide or at other stages of the legal process such as sentencing'. 90

Conclusion
This article has considered some of the reform proposals surrounding the troubling cases involved deaths that have arisen during the course of 'rough sex'/SM. It has explained the complex legal position which is riven with uncertainty and complexity. This can be exploited by defendants seeking to argue that the victim consented to the rough sex/SM and so they should not be convicted, or at least not of murder. The defences draw on fictions about female sexuality and assumptions that can be drawn from someone's sexual history. The article has also highlighted the high levels of unwanted force that is used in 'rough sex'/SM.
The primary reforms in the political agenda has been through amendments to the Domestic Abuse Bill. We have highlighted serious limitations and concerns about these. Arguably, the cultural narratives of victim blaming and victim morality, bound up in gender stereotypes and myths, cannot be simply overcome with the proposed substantive legal reform. Although these reforms seek to place the Brown principle on a statutory footing, the rough sex gone wrong narrative in defence cases is not created by virtue of Brown being common law rather than statutory law. As Fitz-Gibbon 91 points out, it is essential to recognise that the battle against narratives of victim blaming in our criminal courts is not confined to the law of homicide.
Furthermore, the proposed reforms do not cover many of the cases that need to be covered and may not have the intended effect. To deal with the legal response to unwanted 'rough sex'/SM, we need a thorough review of the whole area. The key issue, which has been lost sight of in many of the debates, is over what should count as consent in the context of 'rough sex'/SM. As the research shows, there is a widespread assumption among too many men that women like it 'rough' and consent to rough sex/SM can be taken for granted. Law reform should be focussed on challenging that assumption.