The relationship between ‘rough sex’ and the criminal law has recently been subject to considerable scrutiny. Much of this debate has focused on male defendants in homicide cases, who have claimed the death of the female victim resulted accidentally from consensual rough sex. As a result, more commonplace occurrences of rough sex that do not result in death have received limited attention. In particular, the role that rough sex plays in abusive relationships, whether and how this is criminalised, has been largely overlooked.
This article addresses this gap by illuminating the role that rough sex can play as both an instrument and a manifestation of coercive control. We problematise the ‘stories’ that are told about rough sex and coercive control in the criminal law. Three constructions of ‘rough sex’ have historically been applied in the case law: ‘Violent sexual assault’, ‘deviant sexuality’ and ‘accidental injury’. The introduction of a new offence of ‘controlling or coercive behaviour’ was an opportunity to uncover a new, more accurate narrative of abusive rough sex. We argue, however, that the courts are still telling the same old stories.
The grisly murder of young British backpacker Grace Millane, who was raped, murdered and then dumped in a forest in New Zealand, reignited the debate over the so-called ‘rough sex’ defence. Grace’s murderer tried to argue that Grace died accidentally after the pair engaged in consensual rough sex that ‘went too far’ (The Guardian 18 December 2020). In November 2019 a jury rejected that argument and found him guilty. The preceding year, in December 2018, there was outrage in England over the inadequate sentence handed down to John Broadhurst for the manslaughter of his girlfriend Natalie Connolly. Broadhurst claimed Natalie died accidentally after consensual ‘rough sex’ (Dearden 2018). The campaign group ‘We Can’t Consent To This’ (www.wecantconsenttothis.uk) was formed in response to this verdict by Fiona Mackenzie, and has since run a successful campaign to include a clause in the Domestic Abuse Act 2021 clarifying that consent to serious harm for the purpose of sexual gratification is not a defence. As part of her work to raise awareness of ‘rough sex’ and the role that it plays in the perpetration of violence against women, Mackenzie and other volunteers collect and publish survivors’ stories on their campaign site. Unfortunately stories of ‘rough sex’ within an ongoing abusive intimate relationship are well represented on the site. These demonstrate that the stories of homicide which generate so much media attention are only the tip of the iceberg. This is not a surprise: research shows that sexual violence within an abusive relationship is a carefully constructed and critical component of an abuser’s strategy of domination, what Evan Stark has labelled ‘coercive control’ (Stark 2007).
This article examines the use of ‘rough sex’ within a coercively controlling abusive relationship and its construction within the criminal law. Part One interrogates the key concepts of ‘rough sex’ and ‘coercive control’. Using case studies from the literature, we illustrate the role that rough sex can play as both an instrument and a manifestation of coercive control. Part Two examines three alternate constructions of ‘rough sex’ historically applied in the criminal law of England and Wales: ‘Violent sexual assault’, ‘deviant sexuality’ and ‘accidental injury’. Finally, in Part Three, we review the legal construct of ‘controlling or coercive behaviour’ that was introduced by section 76 of the Serious Crime Act 2015. We conclude that the new law fails to accurately capture the wrong of rough sex as a dimension of coercive control. The wrong story, in other words, is still being told and the criminal justice system continues to fail survivors of abuse as offenders are not held to account for the specific harm they have caused.
Throughout this article, we use the term ‘rough sex’ to refer literally to the use of rough or forceful physical contact in the course of a sexual encounter. It can result in the infliction of physical pain and/or injury. This would include, for example, spanking, slapping, choking, forcefully grabbing or pinning down a partner in the context of sexual activity. It may involve the use of weapons, sex toys or other implements, such as beating with a whip or penetrating with a dildo or a bottle. Sexual penetration – whether with a penis, other body part or object – does not by itself amount to rough sex. However, it would cross this threshold where a substantial degree of force is used such as where an object is forcefully shoved into the vagina or anus, or where one person thrusts their penis into another’s mouth and throat. For our purposes, ‘rough sex’ includes, but is not limited to, sadomasochistic sex. In other words, it includes sexual activity where forcefulness, pain or injury are deliberately employed for the sexual gratification of any of the parties, but we also recognise that physical roughness during sex is not always used with the goal of heightening sexual pleasure. It may be used, for example, to frighten, hurt, humiliate or dominate and this would also come within our definition of rough sex. To be clear, we do not view these purposes as mutually exclusive. A sadistic rapist may, for example, use physically aggressive sex to hurt and dominate his partner and because he is sexually aroused by doing so (Groth 1979, Finkelhor and Yllö 1987).
We have defined rough sex in deliberately broad and neutral terms to accommodate sexual activity involving enthusiastic and willing participants as well as that which is abusive, as our purpose is not to position rough sex as inherently ‘good’ or ‘bad’. Rather we seek to expose the uses to which rough sex can be put - its function and meaning - in the specific context of coercive control. We then problematise socio-legal constructions of rough sex which have served to obfuscate its insidious role in abusive relationships.
Recognition of sexual violence as a critical component of a coercively controlling strategy is not new. Work began on modelling coercive control in the 1980s with the now well-known “power and control” wheel developed by Ellen Pence and Michael Paymar (1993) in Duluth, Minnesota. This work has since been developed (Dobash and Dobash 2004, Dutton and Goodman 2005, Dempsey 2006, Johnson 2008, Hanna 2009, Hester 2013, Stark and Hester 2019). Sexual and physical violence have a prominent and distinct role as the outer circle of the wheel, with all of the non-physical aspects of control (such as threats, intimidation, isolation) labelled as constituent spokes. What is critical about this representation of control (and possibly one of the reasons it is still widely used in clinical practice (Wiener, forthcoming), is the way in which the physical and sexual violence are represented as only one part of the model, but an ‘enabling’ part, the part which holds the behavioural spokes in place. The spokes, in other words, need the rim of the wheel to give them shape and structure: the physical and sexual violence are what reveal the behavioural spokes’ significance as part of a coercively controlling strategy of domination.
The reason the physical and sexual violence is ‘enabling’ is that it gives credibility to the threats. It is when a threat is credible that a demand becomes coercive. Physical and/or sexual violence as an enabler often occurs early on in an abusive relationship and can function as an important moment of transition. The perpetrator, in other words, introduces threat-credibility via an act of violence that reveals him as an abuser for the first time. This is illustrated by the following example from Cassandra Wiener’s study of coercive control, in which Jessica, an Independent Domestic Violence Advisor, tells the disturbing story of a survivor client who was badly frightened on her honeymoon:
Her story was that everything was groovy, no issues, they got married they went on their honeymoon, and he strangled her with the bathroom towel. Really, really badly. There was a horrific, traumatic incident when he strangled her almost to death with the bathroom towel ... So then after that for that six years of their relationship -… he never ever again used physical violence on her but whenever there was a moment of tension he would go to the bathroom and he would bring out a towel, and he would put it on the table. And that was the sign; and then she would just be, like, “and then I would just give in - I would just do whatever it is he was trying to get me to do” (Wiener, forthcoming).
The towel incident is a good example of the enabling nature of the physical violence, its role as the shape-giver (rim) to the behavioural spokes. The perpetrator in this case nearly killed his victim in a violent attack. This shifted the balance of power in the relationship, permanently, by giving shape and credibility to future threats. The victim in this case knew from this moment on that the perpetrator was capable of a serious, possibly fatal, attack – he had carried one out before. The presence of the bathroom towel, a non-violent threat, is enough in these circumstances to facilitate coercion.
While it can be a ‘moment’ of realisation that introduces the abusive phase of a controlling relationship, the fear and entrapment that result are ongoing. Survivors describe a ‘state of siege’ that is continuous as appeasing the abuser becomes the driving force at the core of every lived day and pervades every aspect of the survivor’s life (Dutton 1992). He dictates, in other words, not only how she has sex, but also how she eats, dresses, parents her children, how and when she goes out, what money she is allowed access to: all these basic aspects of living are controlled by the perpetrator.
Any injuries resulting from physical violence fade into insignificance in the context of the fear, which is expressed as a generalised fear of an innominate event, a terror of something which might happen at any moment if the abuser is not successfully appeased. Translating this experience of all-consuming control into law is challenging, as the criminal law traditionally focuses on discrete incidents. Thus historically, domestic abuse has only attracted legal attention in the form of one-off events, abstracted from their context of ongoing control. Moreover, the seriousness of those events has been assessed in terms of the severity of any physical injuries caused, a measure which does not reflect the victims’ understanding of the harm done to them (Stark 2007).
This mismatch, between events that are specific and abuse that is ongoing, has implications that are important for the analysis of the relationship between rough sex in the context of coercive control and the criminal law. In the specific context of the sexual offences, Tanya Palmer’s work highlights the law’s focus on ‘acute’ incidents of non-consensual sex and its consequent marginalisation of ‘chronic sexual violation’. Palmer coined the term ‘chronic sexual violation’ to refer to a ‘constellation of behaviours designed to control and denigrate [the victims]’s sexuality’ including, for example, rough sex, unwanted touching, denials of privacy, reproductive coercion and rape (Palmer 2020). This conceptualisation ‘draws attention to the multiple interconnected ways that abusers undermine victims’ sexual autonomy, over and above discrete incidents of non-consensual sexual contact’ (Palmer 2020, p 595). As with all aspects of coercive controlling behaviour, rough sex in this context is chronic not acute, pervasive and on-going rather than incident specific.
Inevitably the ongoing nature of the abuse affects the way that survivors learn to understand themselves and the world around them. Even the debilitating anxiety is not the worst effect of the abuse. Mary Ann Dutton talks about the way in which the psychological impact of abuse goes beyond symptom focused conditions such as anxiety to include ‘the ways in which battered women have come to think about the violence, themselves, and others as a result of their experiences’ (Dutton 1992). Survivors explain that worst of all was how they learn to blame themselves for the position in which they find themselves and lose confidence in their ability to make decisions about their own and their children’s lives. As Evan Stark puts it: ‘he changes who and what she is’ (Stark 2007, p 262).
In this context of ‘the state of siege’, it is possible to sketch out the role that rough sex can and does play within a controlling relationship. Rough sex exists as both the rim and the spokes of the Duluth wheel: firstly, as aggressive physical force it gives credence to sexual threats, much in the same way as the towel on the kitchen table in Jessica’s example, above. Like non-sexual physical violence, rough sex gives credibility to the relationship between the demand and the threat by demonstrating to the victim what can happen if she does not comply. This is what Evan Stark refers to as the ‘or else’ proviso that colours every demand, especially sexual demands. Rough sex thus operates as a form of coercion. Secondly though, it can operate as an act of sexual domination and therefore as controlling behaviour in and of itself. It allows a perpetrator access (sex when he wants it) and sexual control (sex how he wants it).
Louise McOrmond-Plummer, an Australian researcher and survivor, writes of her harrowing experience of living with a sexual abuser. She explains in gruelling detail how her recognition of what her abusive partner, Richard, was capable of gave credibility to his threats, and resulted in sex whenever and however he wanted it. She explains that Richard: ‘did indeed use sexual violence in a systematic way, with certain outcomes in mind…This is crucial to know because many people are inclined to confuse IPSV [intimate partner sexual violence] with “sex”, rather than the act of violence, control, and degradation that it is’ (McOrmond-Plummer 2016). She elaborates:
Richard seemed to believe that in order to keep me, he needed to rule me. Early in our relationship, we argued and he called me degrading names. I angrily expressed regret for becoming involved with him, and said, “You will never touch me again. Now, get out.” Richard sneered, “I can fuck you whenever I want to.” I raised my voice, reiterating that he should leave. Richard pushed me to the floor and sat on me, delivering repeated hard slaps across my face, and then raped me, taunting me with the fact that he could and would do what he liked, when he liked. I actually did feel like conquered property: worthless. The rape ended – at least for the time being – further talk of leaving (McOrmond-Plummer 2016, p 99).
In this scenario, Richard makes a threat ‘I can fuck you whenever I want to’ which he carries out with a violent attack. He slaps her, pushes her to the floor and then rapes her. The rough sex culminating in a rape puts the rim of the wheel in place. She feels ‘conquered’ and is coerced into staying with him. For the rest of their time together Richard only needed to warn, ‘don’t make me come and get you’ and she would submit to sex. His threat had credibility. She knew what he would do if she did not submit. The regular (unwanted) sex that they had was Richard exercising sexual control: having sex when and how he wanted.
Rough sex within a controlling relationship can therefore function to cement an understanding of ownership – of the total control that a perpetrator has over his victim. In sexual terms this means that a perpetrator such as Richard can have what they want, when they want it. Louise explains how Richard ‘needed to rule me’, and how, after the first rape, she felt like ‘conquered property: worthless’. Once ownership has been asserted, small rough gestures can serve as insidious reminders. Not all of the violence that Louise experienced in the following years of her abusive relationship with Richard was as extreme, but the expression of ownership was ever present in rough gestures such as forced touching/grabbing. Even low level rough sex can – when imbued with menace because of the credibility-threat nexus – be a visible manifestation of the worst kind of coercive demand. In this way, a behaviour such as Richard’s forced touching and grabbing can function as ‘an assertion of ownership – a frequent reminder that her body is not her own’ (Palmer 2020, p 579).
Other survivor stories in the literature can be used to illustrate the point. Wiener describes an interview with a survivor, Sarah, who, like Louise, was raped early on in her relationship with her abuser. After that, she realised that ‘he would not take no for an answer for sex’ (Wiener, forthcoming). She felt that her body was not her own, and that her abusive partner could, and did, have sex as and when he wanted it. Sarah gave a particularly poignant example of the repercussions of his ownership of her body. She explained how, when she was well into her pregnancy with her first child, she tried to say ‘no’ to sex. He raped her, and the physical trauma of this non-consensual rough sex caused a placental abruption. She was rushed into hospital for an emergency caesarean. Her abuser forced her to discharge herself from hospital four days early, and then on her return home, raped her again. She explained: ‘sitting there having nursed my baby, massive lactating breasts and still had this metal suture, but he again forced sex – I hadn’t even – and basically his view was that because I’d had a caesarean that area wasn’t affected’ (Wiener, forthcoming).
Critical to all of these stories of rough sex is the impact the control backdrop has on both the ‘why’ and the ‘how’ of the aggressive physical force that takes place. It is at the same time an instrument/enabler (the rim of the wheel) and a behavioural manifestation (a spoke) of coercive control. Sarah and Louise submitted to their abusers because they were terrified of what might happen to them if they did not. In this way aggressive physical force applied in the context of a sexual encounter enables sex for the abuser. It allows him to erode his victim’s sexual autonomy. Secondly, physically aggressive sex within an abusive relationship is an example of control – it can happen whenever and however an abuser wants it to, and communicates to the victim that she is merely an object for him to use. Small, insidious gestures can be terrifying reminders of this fact. We will now turn to Part Two of this paper with a review of how the law of England and Wales conceptualises rough sex.
Our analysis of relevant statute and case law reveals three alternate ways in which rough sex has historically been framed within the criminal law: As a violent sexual assault, as an expression of sexual deviance, or as the accidental causing of injury via legitimate sexual activity. The potential for a fourth framing – rough sex as control – was introduced by section 76 of the Serious Crime Act 2015. In this section we set out the first three of these constructions, highlighting their inadequacies in the context of relationships characterised by coercive control. In Part Three we consider whether and to what extent section 76 allows for a reframing of rough sex as coercive control.
One way in which rough sex can be framed under existing law is as a violent sexual assault which closely resembles the ‘real rape’ stereotype (Estrich 1988). (The term ‘sexual assault’ is used throughout this section as an umbrella term to include all forms of non-consensual sexual activity contained in ss1-3 of the Sexual Offences Act 2003: rape, assault by penetration and sexual assault). Here, the sexual activity itself is understood as non-consensual and the physical force used and/or injuries caused function both as evidence of non-consent and as an exacerbating feature that heightens the severity of the sexual assault. Where prosecutors understand a case as fitting within this narrative, non-consensual sex becomes the central wrong and the defendant will accordingly be charged with an offence of rape, assault by penetration or sexual assault pursuant to sections 1-3 of the Sexual Offences Act 2003 (SOA) respectively. Evidence of physical violence (beyond the violence inherent in non-consensual sexual contact) can then be used to bolster the prosecution case. The use of violence against the complainant during or immediately before the relevant sexual act raises a presumption that the complainant did not consent to the sexual act. It also raises a presumption that the defendant did not reasonably believe the complainant was consenting (Sexual Offence Act 2003, s 75(2)(a)). These are, however, rebuttable presumptions. They thus leave the door open for the defendant to counter that this was in fact an encounter involving consensual rough sex.
If the jury accepts the narrative of violent, non-consensual sex and the defendant is convicted of an offence under the SOA, the defendant’s use of violence will be taken into account at the sentencing stage (Sentencing Council, ‘Sexual Offences Definitive Guideline’). Where a sexual assault is accompanied by extreme physical violence and/or injury, this can also be charged as a separate additional offence under the Offences Against the Person Act 1861 (OAPA). This happened, for example, in the landmark marital rape case of R v R (1992). In practice this is only likely to occur in cases involving the highest levels of physical violence that could not be sufficiently reflected in sentencing for a sexual offence (Crown Prescution Service, ‘The Code for Crown Prosecutors’). In the most extreme cases, where a sexual assault culminates in the death of the victim, the narrative of violent sexual assault becomes one of sexually motivated homicide, with the killing as the central wrong and sexual assault as an aggravating factor. Here, the defendant will be charged with murder or manslaughter, potentially accompanied by a charge under the SOA to reflect this additional dimension to the offence. This was the approach taken in Broadhurst, where the defendant was initially charged with murder and assault by penetration. (Broadhurst was ultimately convicted of a single count of gross negligence manslaughter, as explored further below.)
There are both practical challenges and conceptual problems with applying a narrative of violent sexual assault to coercive controlling rough sex. In terms of practical challenges, the key issue here is proving a lack of consent. However, even where this is proven, constructing rough sex in a coercive controlling relationship as sexual assault is conceptually flawed because it frames the encounter as an acute incident of sexual violation as opposed to an expression of ongoing control.
Lack of consent is an actus reus element of each of the sexual assault offences. Consent is defined in section 74 of the SOA as agreement by choice with the freedom and capacity to make a choice. The corresponding mens rea element is the defendant’s lack of reasonable belief in the complainant’s consent. Thus the pivotal facts which must be proven in a sexual assault case relate to the complainant’s mental state and the mental state of the hypothetical reasonable person in the defendant’s position at the time of the sexual activity. The intangible nature of these facts, which must often be inferred from only the conflicting accounts of the complainant and defendant, are thus inherently difficult to prove beyond reasonable doubt (Carline and Easteal 2014). Moreover, the statutory definition of consent has been critiqued for its complexity and ambiguity (Temkin and Ashworth 2004). Thus the ‘nebulous concept’ of consent (Carline and Easteal 2014), and the limitations of the evidence from which it must usually be inferred, allow for assumptions about ‘normal sex’ and ‘real rape’ to fill in the gaps and to shape juror deliberations on these crucial issues (Ellison and Munro 2009).
A hallmark of the ‘real rape’ stereotype is violence (Estrich 1988). This is given legal weight, as noted above, by a statutory presumption of non-consent where the complainant is subject to violence at the time of the sexual activity (Sexual Offences Act 2003, s 75(2)(a)). On the face of it, this would seem to bolster the chances of conviction in a ‘rough sex’ case. However, where rough sex takes place in the context of coercive control, assumptions about violence are undercut by intersecting assumptions about intimate relationships. While the substantive law makes no distinction between sexual assault committed by a partner and that committed by a stranger, in practice it is harder to secure a conviction where the complainant and the accused are, or have been, in an intimate relationship (Carline and Easteal 2014).
This is particularly the case where that relationship continues after an alleged rape or sexual assault takes place, as jurors and other actors in the criminal justice process struggle to fathom why a complainant would remain in a relationship with their attacker. Yet separating from an abusive partner is a complex and often drawn out process. Victims rarely ‘leave’ at a point in time, the transactional moment of leaving as imagined by those not familiar with the dynamics of domestic abuse rarely exists (Tuerkheimer 2013). Instead, victims are forced to constantly assess their options in the context of what they know about the dangers of leaving. They strategically deploy compliance and defiance as they prioritise their own safety and that of their children. Indeed, Plummer’s account of her abuse in Part One demonstrates how rough sex, through its tacit reminder of perpetrator capabilities, can coerce a victim to stay.
A failure to understand these dynamics of coercive control means that abusive rough sex in this context can be all too easily recast by the legal system as a legitimate sexual preference, or ignored altogether. This is especially so where there is evidence of rough sex between the parties on multiple occasions. For example, the case of F v DPP (2013) details a relationship which bears a strong resemblance to that described by Plummer. On one occasion described in the judgment, F and her husband argued, he then became ‘aggressive’ and ‘pulled off her pyjama bottoms, tore her underwear and took her by the throat (ibid, para 12). The DPP’s principal legal adviser had concluded that this event should not be prosecuted as a sexual assault because ‘a jury would be justified in concluding that a “pattern of sexual force or roughness had developed between them, to which there was at least a degree of acceptance on her part, and which he understood that she agreed to, even if reluctantly”’ (ibid, para 13). Troublingly, F’s husband’s repeated use of violence against her, and her ‘acceptance’ of that violence, is interpreted as evidence of F’s consent to rough sex rather than as evidence of her husband’s abuse (Palmer 2020).
Such interpretations are facilitated by the rules on sexual history evidence set out in the Youth Justice and Criminal Evidence Act 1999 section 41(3), which allow for the admission of ‘similar fact evidence’, ie evidence of strikingly similar sexual behaviour between the complainant and the defendant or even between the complainant and other parties. Though designed as a limited exception to the prohibition of sexual history evidence, this provision has been interpreted broadly (McGlynn 2017). It thus enables defendants to introduce evidence that the complainant had previously engaged in rough sex, in order to suggest that she probably consented to it on the occasion in question (McGlynn 2017, Carline and Easteal 2014). In fatal rough sex cases, where the defendant is charged with homicide rather than sexual assault, there is no restriction on sexual history evidence and the deceased victim is unable to dispute the defendant’s narrative that they regularly enjoyed ‘rough sex’ (Edwards 2020). It has been argued that juries are increasingly willing to accept such narratives due to the normalisation of rough sex in heterosexual relationships, often referred to as the ‘Fifty Shades effect’ (Harrison 2020). In support of this, campaign group We Can’t Consent To This point to the increasing use of the so-called ‘rough sex defence’ in homicide cases since 2000 (Yardley 2020). However, the strategy of claiming a complainant ‘liked it rough’ in order to undermine her claim of rape is by no means new (Adler 1987). Coercive control victims who have been repeatedly subject to ‘rough sex’ are thus uniquely disadvantaged in their efforts to have this abuse recognised as violent sexual assault.
Despite these difficulties of proving a lack of consent in the context of coercive control, it is not impossible for rough sex in such relationships to be recognised as violent sexual assault and to be successfully prosecuted under the SOA. The cases most likely to fit within this construct are those involving the highest levels of violence and/or injury, up to and including death, where it is harder to interpret the sexual activity itself as consensual. However, these cases present a conceptual problem. A conviction for rape or sexual assault frames rough sex as an acute incident of brutal sexual violence, rather than conceptualising it as part of a pattern of coercive control or ‘chronic sexual violation’ (Palmer 2020).
The violent sexual assault narrative thus abstracts specific non-consensual sexual acts from their broader context, obscuring the gravity and strategic nature of chronic sexual violation as a pattern of sexually violating behaviour. This silences a central aspect of the victim’s experience of their sexual autonomy being gradually eroded. This experience of constant domination, of being alienated from one’s own body and sexuality in big and small ways every day – the state of siege we referred to in Part One – is distinct from the experience of sexual assault as an isolated event in ways that matter to survivors. In addition, attaching liability only to those incidents that meet the criteria of sexual assault means that the other aspects that make up an experience of chronic sexual violation continue to go without legal recognition. To summarise then, the ‘violent sexual assault’ narrative has limited utility in cases of coercive controlling rough sex, as it seems that all but the most violent incidents are problematically interpreted as consensual sexual encounters or ignored altogether.
A somewhat different rough sex narrative emerges in the offences against the person case law on sadomasochism, and specifically in the leading case of Brown (1994). Brown involved a group of men engaging in acts such as beating, caning and whipping, applying stinging nettles to the genitals, and piercing the scrotum and penis. These acts were performed for the sexual gratification of the participants. There was no scope for sexual assault charges to be brought because all of the defendants’ activities were consented to. This consent, however, did not negative all criminal liability and the defendants were convicted of various counts of assault occasioning actual bodily harm (ABH) and wounding, contrary to sections 47 and 20 of the OAPA. The House of Lords ruled that consent is not a defence to either charge unless the activity in question is performed for a ‘good reason’. Specifically, consent will provide a defence in the context of surgery, regulated sport, some forms of body modification and ‘horseplay’, but not in the context of sadomasochistic sexual activity.
Lord Templeman notoriously referred to the group’s activities as a ‘cult of violence’ (R v Brown (1993), p 237). Continuing that, ‘Pleasure derived from the infliction of pain is an evil thing. Cruelty is uncivilised.’ (ibid). Thus, rather than a narrative of violent sexual assault as above, the judgment in Brown constructs the infliction of pain and/or injury for the purpose of sexual pleasure as sexually deviant and morally repugnant. There is a rich literature critiquing the homophobic attitudes expressed in the majority speeches, its doctrinal inconsistencies, and the normative inconsistencies in terms of what will be classed as a ‘good reason’ for injuring another (Bibbings and Alldridge 1993, Weait and Hunter 2010, Khan 2009, Tolmie 2012, Kell 1994, Bamforth 1994). While much of this discussion is beyond the scope of this article, it is significant that Brown involved a group of men engaged in homosexual activity. In cases involving heterosexual couples, the courts appear to have been more accepting of rough sex and sadomasochism and less willing either to criminalise it or to describe it with such stigmatising language.
For example, in Emmett (1999), the male defendant injured his female partner in the course of consensual sadomasochistic sexual activity. Following Brown, he was convicted of ABH. However, the judgment contains none of the moral disgust that pervades the Brown judgment, going only so far as to describe the acts as ‘outré sexual activity’. There are also examples of Brown being ignored altogether. In the case of Lock (2013), the defendant chained a woman to his bedroom floor and whipped her repeatedly with a rope. The complainant had, in her own words, agreed to this act as part of a Master/slave contract but she had not anticipated the level of pain she would be subjected to (BBC 2013). Lock then had sex with her and left her chained to the floor, at which point she text a friend to call the police (Arkell 2013). In his defence Lock cited the complainant’s consent and asserted that ‘It was supposed to be kinky fun, I didn’t want her to cry’ (BBC 2013). He was acquitted despite this case clearly involving the infliction of pain and injury for the purpose of sexual gratification.
The outcome of Lock likely has much to do with changing attitudes to sadomasochistic sex in the two decades that had passed since Brown. Nevertheless, when looked at in conjunction with related OAPA case law, it does appear to be part of a pattern of normalising heterosexual rough sex in particular. This suggests that assumptions about rough sex are heavily gendered and heteronormative. Notably, the Brown precedent was recently codified in section 71 of the Domestic Abuse Act 2021. We share concerns that this entrenches the unnecessary criminalisation of genuinely consensual sadomasochistic practices (Bibbings and Alldridge 1993) and that it will have limited application in the homicide cases it aims to reform (Reed 2020, Rogers 2021). However, it may be that in non-fatal assault cases such as Lock, this codification of Brown will make it harder to ignore. For the moment, Brown provides a path to criminalising rough sex, even where a lack of consent to that sex is not proven. However, in cases where the narrative of deviance is played down, rough sex is legitimised and normalised. This opens up the possibility of injuries caused by rough sex to be formulated not as deliberate acts of depravity but as unfortunate accidents.
The courts have drawn a distinction between cases such as Brown, in which the defendant consensually and deliberately inflicts pain or injury on another for the purpose of sexual gratification, and cases such as Slingsby (1995) in which injury is caused accidentally in the course of consensual sexual activity and was not reasonably foreseeable. In the latter scenario, the victim’s consent to sex does provide a defence to OAPA charges. Thus in Slingsby no unlawful act was committed where a woman was injured in the course of ‘vigorous’ consensual sex. Specifically, the defendant inserted his hand into her vagina and rectum, and she sustained a number of cuts from the signet ring he was wearing. She later developed septicaemia and died as a result.
Framing injuries as accidental will not always result in the defendant avoiding liability completely (see, for example, Emmett (1999)). In cases where the victim ultimately dies from her injuries, constructing the harm as accidental means that the most serious charge that can be brought is one of gross negligence manslaughter (GNM). This is demonstrated by the case of Broadhurst (2018). As set out in the Sentencing Remarks, John Broadhurst beat his heavily intoxicated girlfriend, Natalie Connolly, with a boot and with his hand He inserted a spray bottle of cleaning fluid into her vagina and then pulled it out, which caused bleeding. He then left her lying on her back at the foot of the stairs where she died. This final omission to get help, rather than the initial infliction of injuries, formed the basis of Broadhurst’s conviction for GNM. GNM is clearly a serious offence via which the defendant is held responsible for death due to his extreme carelessness and disregard for the victim’s life and safety (Adomako (1995), Misra (2004)). Nevertheless, it frames the victim’s death as an accident. GNM convictions therefore tend to result in shorter prison terms than constructive manslaughter or murder (Edwards 2020). Broadhurst himself was sentenced to 3 years 8 months imprisonment, half of which to be served in prison before being released on licence (Dearden 2018).
On the facts of the case, it would have been possible to tell a story of violent sexual assault and/or sexual deviance. Indeed, the prosecution initially brought charges of murder and assault by penetration but these were dropped part way through the trial when Broadhurst plead guilty to GNM. The beating and insertion of the bottle were thus accepted as consensual, despite evidence that Natalie was extremely drunk at the time. The prosecution could have argued that the injuries inflicted were nevertheless unlawful, following Brown, but they did not do so. The judge, however, did acknowledge the criminality of the beating when passing sentence (Broadhurst (2018)). Nevertheless, the overarching narrative presented in the case is one of accidental death attributable to Broadhurst’s lack of care as opposed to a story of deliberate harm.
Clearly, where a defendant has caused injury to another in the course of a sexual encounter, the most appealing way to frame that event from his perspective is as an accident resulting from consensual sex. In non-fatal cases this provides a path to avoiding liability altogether. In homicide cases the ‘accident’ narrative similarly enables defendants to avoid or at least significantly reduce their criminal liability relative to the alternative constructions of ‘violent sexual assault’ and ‘deviant consensual sex’. Notably, the cases discussed above and in the previous section appear to suggest a particular willingness to apply this narrative in cases of men injuring women in the course of sexual activity, and to normalise a degree of rough sex, reframed as ‘vigorous’ sexual activity in heterosexual relationships. This has implications for the framing of rough sex in coercive controlling heterosexual relationships, which are themselves heavily shaped by normative gender roles (Stark 2007). This assertion is tentative, given the small number of case examples. A systematic study of first instance decisions in assault cases and coercive control cases stemming from rough sex would help to establish the extent to which the rules in Brown are being ignored, and any patterns in their application. Further research could also shed light on the construction of rough sex in coercive controlling relationships that fall outside of the male perpetrator – female victim dynamic.
One further implication of these cases with regard to coercive control is that, just as with the narrative of ‘violent sexual assault’, the narrative of ‘accidental injury’ posits an isolated event. Moreover, under the OAPA, the seriousness of an offence is based on the level of injury caused. This does not mean that evidence of ongoing control could not be introduced in an OAPA trial – to demonstrate for example, that injuries were inflicted without consent and/or that they were intended to harm. However, it does mean that if a conviction is secured under this Act, a fuller picture of the chronic violation experienced by the victim – the ‘state of siege’ – is not reflected in a conviction that emphasises one specific act of violence. Further, the sexualised nature of that violation is lost in charges which focus on physical injury. Perhaps most significantly for coercive control victims, rough sex that does not result in injuries simply does not register under the OAPA framework. The application of the SOA here is also limited, as established above. Yet as we argued in Part One, non-injurious violence, including rough sex, is a prominent feature of these relationships and a key way in which control over the victim is expressed and maintained. The problem of aspects of coercive controlling behaviour going under the radar in this way was one of the key motivators for introducing a new offence under s76 of the SCA, to which we now turn.
So far, we have problematised three alternate constructions of ‘rough sex’ in the criminal law of England and Wales: violent sexual assault, deviant sexuality and accidental injury. The criminalisation of coercive control in England and Wales should have been an opportunity to tell a new story. Drafted properly, a coercive control law has the potential to allow the prosecution of rough sex within a controlling relationship for what it is – an insidious part of a perpetrator’s repertoire as he seeks to control every aspect of his victim’s life, including her sexual life.
Unfortunately, at least in England and Wales, a fourth construction of rough sex, one which takes place within the paradigm of coercive control, has not yet emerged. Section 76 Serious Crime Act 2015 (‘section 76’) attempts to articulate coercive control as ‘controlling or coercive behaviour’. Section 76(1) states that an offence is committed if a person repeatedly or continuously engages in behaviour towards another person that is controlling or coercive. ‘Controlling or coercive’ as a construct is not defined in the Act, but the Government’s initial intention, as expressed in the Home Office consultation prior to the introduction of the new clause and throughout the parliamentary debate that followed, was to ‘reinforce’ the law ‘to capture patterns of non-violent behaviour within intimate relationships’ (Home Office 2014) and so to introduce ‘an extra element that closes a loophole’ (Hansard 2015).
With a maximum sentence of five years, controlling or coercive behaviour is thus positioned as a relatively low-level offence, as one fragment of domestic abuse (psychological abuse) that was previously not being captured by the criminal law. In fact coercive control, as we explained in Part One of this paper, is not a ‘fragment’ of domestic abuse, rather it is a complex web of physical and non-physical perpetrator behaviours. Coercive control therefore incorporates psychological abuse, and defies binaries such as ‘physical’ and ‘psychological’, or ‘sexual’ and ‘non-sexual’. It undermines a victim’s physical, sexual, psychological and emotional autonomy to create the state of entrapment that victims tell us is the ‘worst part’ of their ordeal (McGorrery and McMahon 2019). It is unfortunate that, as Wiener has argued elsewhere, with its determination to ‘plug’ a ‘gap’ the UK Government appears to have misunderstood the very nature of the problem it was trying to address (Wiener 2020).
Inevitably the positioning of ‘controlling or coercive behaviour’ as a fragment of domestic abuse has led to considerable uncertainty as to the boundaries between what behaviours can and should be included in a section 76 charge, and what should be charged separately. Early commentators observe that ‘it is perhaps unsurprising that the various police forces in England and Wales have taken what seem to be quite disparate approaches to charging alleged offences’ (Wiener 2020, p 964). Clarification is needed, in other words, as to ‘the circumstances in which an assault or threat should be charged separately, and the circumstances in which those behaviours should be part of the “course of conduct” underlying a charge of controlling or coercive behaviour’ (Wiener 2020, p 964).
How and to what extent rough sex can technically be prosecuted as part of the course of conduct underlying a charge of controlling or coercive behaviour further to section 76 is therefore currently unclear. What is clear, however, is that a consideration of rough sex as coercive control is completely missing from the early section 76 cases. While there is evidence to suggest that the courts are prepared to use evidence of low-level physical violence as part of the course of conduct underlying the section 76 charge (Dalgarno (2020)), any offending which is violent and sexual is still for the most part conceptualised as something separate to controlling or coercive behaviour. Rough sex in the context of coercive control is therefore still not being recognised for what it is.
At the time of writing, there are 29 reported Court of Appeal decisions on the application of section 76 within the criminal law. In six of these cases (Chanaa (2019), Holden (2019), Osinski (2019), Whitehall (2019), Cunningham (2018) and Parkin (2018)) sexual offences were charged as separate counts on the same indictment. In these five cases, as one might expect (and as we set out in Part Two, above) the rough sex is framed as a violent sexual assault, closely resembling the ‘real rape’ stereotype. In Parkin (2018), for example, the defendant Andrew Parkin was convicted of three counts of rape and one of controlling or coercive behaviour. The rapes took place after Parkin and the victim, ‘C’, had separated. This brings the narrative further into line with the ‘real rape’ stereotype as on each occasion Parkin forced his way into C’s home. In the course of those rapes he was physically aggressive and handled her roughly. There is no mention of rough sex in the context of the one count of controlling or coercive behaviour. We are told simply that ‘The offender displayed very controlling behaviour; wanted to know where C was at all times and would telephone her if he wanted to know where she was. C was expected to answer the telephone whenever the offender rang. If she did not answer he would drive around in an attempt to locate her.’ (Parkin (2018), para 7).
In the 25 cases with no separate sexual offences charges on the indictment there is plenty of description of most of the insidious behaviours that one would expect to find in a coercive control prosecution. However, there is little to no mention of sexual abuse at all. For example, Robert Conlon was convicted of controlling or coercive behaviour and of one count of assault. There is considerable detail in the judgment of the different controlling strategies adopted by Conlon in the course of his oppression of his partner. We are told at the beginning of the judgment that:
From an early stage the appellant became controlling and violent. He limited the complainant’s contact with friends and work colleagues, controlling the clothes she wore and how she styled her hair. He checked her phone and accused her of being unfaithful. (Conlon 2017, para 3)
The judgment is then peppered with references to physical violence, which occurred on multiple occasions. For example, at paragraph : ‘a neighbour heard screams from the complainant and called the police. The appellant had punched the complainant in the face and when she fell to the floor he continued to punch her’. And at paragraph :
‘the appellant had been violent to her, pinning her to the wall and shouting at her. He would require her to print out her work shift, frequently rang her at work to check she was there, told her what clothes to wear to work and forbad her wearing high heeled shoes to work and described her breasts as “belonging to me”.
Of course, we do not know whether or not defendants such as Conlon engaged in rough sex with their victims. It is possible that the descriptions of coercive control do not contain any references to rough sex because it was not taking place. This would be unusual: rough sex, as we explained in the first part of this article, is commonly an instrument and an expression of coercive control. We suspect that the reason that there is no reference to rough sex in the reported cases is because ‘controlling or coercive behaviour’ is being constructed as something separate to sexual offending. In this way, despite the introduction of the section 76 offence, the three alternate constructions of ‘rough sex’ which we have set out above remain remarkably intact, and a fourth construction (rough sex within a controlling relationship) is yet to be introduced.
While the role of rough sex in homicide cases such as the murder of Grace Millane generate the most media attention, rough sex also plays a critical role as an instrument and a manifestation of coercive control, a type of domestic abuse that is as dangerous as it is insidious. Our analysis shows that there were historically three constructions of rough sex in the criminal law: violent sexual assault, deviant sexuality and accidental injury. All three of these narratives break down in the context of the particular harm and wrong of rough sex that is coercively controlling.
Conceptualising rough sex in the context of control as a one off violent sexual assault misses much of what makes it uniquely harmful: its ongoing and pervasive nature as chronic abuse. It also introduces practical problems from a prosecution perspective as juries struggle to reconcile the severity of the attack with what they frame as a survivor’s ‘decision’ to remain in the relationship – that escape is either possible or helpful from a safety perspective are two fallacies often subscribed to by people with no specialist knowledge of domestic abuse.
Furthermore, reluctance to adopt the second narrative – deviant sexuality – in the context of a heterosexual encounter has led to the construction of rough sex as accidental injury. Rough sex is normalised and injuries that ensue are framed as an unintended and thus legitimate by-product.
The introduction of a new offence of ‘controlling or coercive behaviour’ was an opportunity to tell a new story about rough sex, one that captured the wrong and the harm that survivors tell us is amongst the worst of their experiences of abuse. Unfortunately, the offence that came onto the statute books in 2015 is poorly drafted and inconsistently applied. A fourth construction of rough sex, one which reflects its role within coercive control, has not yet emerged. Instead, section 76 is used to prosecute what is positioned primarily as ‘psychological abuse’, with some low level physical violence incorporated in some of the cases. Coercive control, in other words, is being constructed as something that is separate to sexual offending: and the existing narratives of acute violent incidents and accidental injury, with all of their inadequacies, are still the only stories that can be told.
This article was published in Child and Family Law Quarterly, one of the leading forums for the discussion of key issues in child and family law in the UK and internationally. It publishes four issues each year on a broad range of topics, including adoption, child protection, cohabitation, divorce, domestic violence, education, financial orders, marriage, parentage and parental responsibility, property rights, and welfare issues.
Originally published as: Tanya Palmer and Cassandra Wiener, ‘Telling the wrong stories: Rough sex, coercive control and the criminal law’  CFLQ 331
‘Grace Millane Murderer Loses Appeal against Conviction and Sentence’ The Guardian (London, 18 December 2020) www.theguardian.com/world/2020/dec/18/grace-millane-murderer-loses-appeal-against-conviction-and-sentence last accessed 9 February 2021.
‘We Can’t Consent To This’ (We Can’t Consent To This) <wecantconsenttothis.uk> last accessed 3 March 2021.
Adler Z, Rape on Trial (Routledge, 1987).
Arkell A, ‘“I’d have to say ‘Yes Sir, Yes Master’ or face lashing”: Mother beaten in Fifty Shades sex game suffered “excruciating pain” at hands of lover’ (MailOnline, 28 January 2013) www.dailymail.co.uk/news/article-2269434/Fifty-Shades-woman-tells-Steven-Lock-cleared-ABH-Sir-Master-sex-game.html last accessed 5 July 2021.
Bamforth N, ‘Sado-masochism and consent’  Crim LR 661.
BBC ‘”Fifty Shades”' sex-session assault accused cleared’ (BBC, 22 January 2013) www.bbc.co.uk/news/uk-england-suffolk-21145816 last accessed 5 July 2021.
Bibbings L and Alldridge P, ‘Sexual Expression, Body Alteration and the Defence of Consent’ (1993) 20 J Law Soc 356.
Carline A and P Easteal P, Shades of Grey – Domestic and Sexual Violence Against Women (Routledge, 2014).
Crown Prosecution Service, ‘The Code for Crown Prosecutors’ www.cps.gov.uk/publication/code-crown-prosecutors last accessed 5 July 2021.
Dearden L, ‘Outrage at jail sentence for millionaire who claimed girlfriend died during “rough sex”’ (Independent, 18 December 2018) www.independent.co.uk/news/uk/crime/rough-sex-death-millionaire-john-broadhurst-murder-natalie-connolly-case-sentence-prison-jail-term-a8689366.html last accessed 5 July 2021.
Dempsey M, ‘What Counts As Domestic Violence’ (2006) 12 William & Mary Journal of Women and Law 301.
Dobash R and Dobash R, ‘Women’s Violence to Men in Intimate Relationships: Working on a Puzzle’ (2004) 44 British Journal of Criminology 324.
Dutton M, ‘Understanding Women’s Responses to Domestic Violence: A Redefinition of Battered Woman Syndrome’ (1992) 21 Hofstra Law Review 1191.
Dutton M and Goodman L, ‘Coercion in Intimate Partner Violence: Toward a New Conceptualization’ (2005) 52(11/12) Sex Roles 744.
Edwards S, ‘Consent and the “Rough Sex” Defence in Rape, Murder, Manslaughter and Gross Negligence’ (2020) 84 J Crim Law 293.
Ellison E and Munro V, ‘Of “Normal Sex” and “Real Rape”: Exploring the Use of Socio-Sexual Scripts in (Mock) Jury Deliberations’, (2009) 18 Soc Leg Stud 1.
Estrich S, Real Rape (Harvard University Press, 1988).
Finkelhor D and Yllö K, License to Rape: Sexual Abuse of Wives (Holt McDougal 1987).
Groth N, Men Who Rape (Plenum, 1979).
Hanna C, ‘The Paradox of Progress: Translating Evan Stark’s Coercive Control Into Legal Doctrine for Abused Women’ (2009) 15(12) Violence Against Women 1458.
Hansard HC Deb, vol 591, col 172 (20 January 2015).
Harrison E, ‘Harriet Harman says Fifty Shades of Grey provided “grisly opportunity for men to twist women’s empowerment” (Independent, 22 July 2020) www.independent.co.uk/arts-entertainment/books/news/harriet-harman-fifty-shades-grey-rough-sex-defence-ban-bdsm-el-james-a9632486.html last accessed 5 July 2021.
Hester M, ‘Who Does What to Whom? Gender and Domestic Violence Perpetrators in English Police Records’ (2013) 10(5) European Journal of Criminology 1.
Home Office, ‘Strengthening the Law on Domestic Abuse: A Consultation’  8 www.gov.uk/government/uploads/system/uploads/attachment_data/file/344674/Strengthening_the_law_on_Domestic_Abuse_-_A_Consultation_WEB.PDF last accessed 28 March 2018 (emphasis in original).
Kell D, ‘Social Disutility and the law of consent’ (1994) 14 OJLS 121
Khan U, ‘A Woman's Right to Be Spanked: Testing the Limits of Tolerance of S/M in the Socio-Legal Imaginary’ (2009) 18 Tulane Journal of Law and Sexuality 79.
McGlynn C, ‘Rape Trials and Sexual History Evidence: Reforming the Law on Third Party Evidence’ (2017) 81 JCL 367.
McGorrery P and McMahon M, ‘Criminalising the “Worst” Part: Operationalising the Offence of Coercive Control in England and Wales’ (2019) 11 Criminal Law Review 957.
McOrmond-Plummer L, ‘Lucky To Be Alive: A Battering Partner Rapist’ in L McOrmond-Plummer, J Levy-Peck and P Easteal (eds), Perpetrators of Intimate Partner Sexual Violence: A Multidisciplinary Approach To Prevention, Recognition, and Intervention (Routledge, 2016).
Palmer T, ‘Failing to See the Wood for the Trees: Chronic Sexual Violation and Criminal Law’ (2020) 84 The JCL 573.
Pence E, M Paymar M and Ritmeester T, Education Groups for Men Who Batter: The Duluth Model (Springer 1993).
Reed G, ‘New law seeking to ban “rough sex defence” unlikely to change trial outcomes’ (The Justice Gap, 10 July 2020) www.thejusticegap.com/the-so-called-rough-sex-defence-is-unlikely-to-change-the-outcome-of-trials last accessed 5 July 2021
Rogers J, ‘“Abolition” of the “Rough Sex” Defence: Hurried Legislation and Missed Opportunities’ (Criminal Law Reform Now Network, 17 November 2021) http://www.clrnn.co.uk/blog/law-reform-general/ last accessed 5 July 2021.
Sentencing Council, ‘Sexual Offences Definitive Guideline’ www.sentencingcouncil.org.uk/wp-content/uploads/Sexual-offences-definitive-guideline-Web.pdf last accessed 5 July 2021.
Stark E, Coercive Control: How Men Entrap Women In Personal Life (Oxford University Press 2007).
Stark E and Hester M, ‘Coercive Control: Update and Review’ (2019) 25(1) Violence Against Women 81.
Temkin J and Ashworth A, The Sexual Offences Act 2003: Rape, Sexual Assaults and The Problems of Consent’,  Crim LR 328.
Temkin J and Krahé B, Sexual Assault and the Justice Gap: A Question of Attitude (Hart, 2008).
Tolmie J, ‘Consent to harmful assaults: the case for moving away from category based decision making’  Crim LR 656;
Tuerkheimer D, ‘Breakups’ (2013) 25 Yale Journal of Law and Feminism 51.
Weait M and Hunter R, ‘R v Brown Commentary’ in R Hunter, C McGlynn and E Rackley (eds), Feminist Judgments: From Theory to Practice (Hart, 2010).
Wiener C, ‘From Social Construct to Legal Innovation: the Offence of Controlling or Coercive Behaviour in England and Wales’ in McMahon M and McGorrery P (eds) Criminalising Coercive Control (Springer 2020).
Wiener C, Coercive Control and the Criminal Law (Routledge, forthcoming).
Yardley E, ‘The Killing of Women in “Sex Games Gone Wrong”: An Analysis of Femicides in Great Britain 2000–2018’ (2020) Violence Against Women.
Domestic Abuse Act 2021
Sexual Offences Act 2003
Youth Justice and Criminal Evidence Act 1999, s 41(3)(c).
R v Adomako  1 AC 71
R v Broadhurst Sentencing Remarks, 17 December 2018, Birmingham Crown Court.
R v Brown  1 AC 212.
R v Chanaa  EWCA Crim 2335
R v Conlon  EWCA Crim 2450.
R v Cunningham  EWCA 2704
R v Dalgarno  EWCA Crim 290
F v Director of Public Prosecutions  EWHC 945
R v Emmett (1999) EWCA Crim 1710
R v Holden  EWCA Crim 1885
R v Lock (Ipswich Crown Court, 22 January 2013)
R v Misra  EWCA Crim 2375
R v Osinski 
R v Parkin  EWCA Crim 2764.
R v R  1 AC 599 (HL) 599.
R v Slingsby  Crim LR 570
R v Whitehall  EWCA 62