Women Who Kill Abusive Men: The Limitations of Loss of Control,Provocation and Self-Defence in England and Wales and Canada

Method of Killing

When women kill male intimate partners the use of a weapon to compensate for their physical disadvantage has presented as the main obstacle to successful manslaughter and self-defence pleadings. Method of killing is a significant factor in the success of a defence to murder and to a manslaughter outcome whether involuntary manslaughter (accident/no intent) or voluntary manslaughter (loss of control s. 54 or s. 55(3) fear of serious violence loss of control) and is an aggravating factor impacting on sentencing length in both murder and manslaughter convictions (see later Martin). The relational context is little considered, and body force had been historically regarded as a mitigatory factor and weapons as an aggravating factor.

When men assault and kill women, the primary method used is physical body force in beatings, strangulation, suffocation and drownings. In Edwards’ analysis of cases from the Homicide index when men killed intimate female partners (n = 440) body force (hitting, kicking and punching, pinning down, drowning, strangulation and suffocation etc.) was used in 140 (32%) cases. Strangulation and suffocation and choking alone were used in 102 cases (23%), constituting 73% of all cases involving body force and the method of killing was recorded as ‘not known’ in 21 cases (5%). As strangulation and suffocation do not always leave corroborative signs of injury or signs necessarily commensurate with the degree of force used unlike killings with a weapon or by beating or burning; it is likely that in many of these cases, women died from strangulation or suffocation. A weapon (blunt instrument, sharp instrument or shooting and poison excluding cases where a ligature was used to strangle) was used in 268 (61%) of cases. It is from these forms of non-fatal and potentially fatal assault women try to take self-defensive action.

When women killed male intimate partners (n = 62) during the period 2011–2019, in 58 cases (94%), a weapon was used. Of the remaining four cases, in one the cause of death was unknown, in another death was due to being a passenger in a vehicle and in two cases death was due to suffocation or asphyxiation. Women strangled and/or suffocated male intimate partners in a total of four out of 62 cases, and in two using a ligature. In one of these four cases, there was clear evidence of mercy killing. In three of the four cases, the victims were drunk or otherwise incapacitated which enabled the women defendants who were otherwise physically disadvantaged to overpower male body force and resistance.

Of the 62 male victims killed by intimate female partners and defined visually by police, three were Asian, two were from other ethnicities, two were Black and one was not recorded, totalling eight (11% ethnic minority of all male victims). Of the female suspects two were Asian, one Black and one other ethnicity, accounting for 7% ethnic minority suspect representation. This profile is proportionate to the population data. Drawing on the 2011 census data, ⁵ approximately 13% of the population of England and Wales are from ethnic minority communities with a population of 86% white, 7.5% Asian, 3.3% Black, 2.2% mixed and other groups.

Of the 440 female victims and their male suspects a very different picture is revealed. Of the victims, in 11 cases ethnic identity was not recorded, 54 were Asian (12%), 14 were from other ethnicities (3%) and 37 were Black (8.4%) totalling 105 ethnic minority victims (24% of all female victims). Ethnic minority women's victimisation in intimate partner homicide is disproportionate to their presence in the general population. Of the suspects in three cases ethnic identities were not recorded, 54 were Asian (12%), 13 were other ethnicities (3%) and 59 were Black (13%) totalling 126 ethnic minority suspects (29%).

A factor in these intimate partner homicides which also arises in the Canadian context (below) is the disproportionate representation of ethnic minorities in both suspect and victim categories. Any detailed analysis of why this is the case is beyond the scope of this particular article. However, what is required and has been called for is better data including collecting disaggregated data on incidents of domestic violence by (a) the ethnicity of the complainant, (b) the immigration status of the complainant and (c) police recorded outcome. ⁶

Method of Killing and Legal Outcome Prognosis

The use of a weapon to kill is taken into account in assessing and building a case of murder/intention to kill. The use of a weapon is regarded as a disproportionate use of force when used on a man by a physically weaker woman. The presumption is that the defendant and victim are on a level playing field and thus a woman or a smaller and weaker man who uses a weapon to address the physical disadvantage when encountering a male aggressor is regarded not as levelling the uneven field but as using disproportionate force.

It is to be noted however that in the US case of State v Wanrow (1977), the Court recognised the relevance of the relational context in considering the reasonableness of the female appellant's use of a gun against an unarmed intruder. The Court acknowledged that the appellant had reason to believe that the intruder had molested her daughter in the past and was returning to abuse her son. The appellant was a 5′4″ woman with a broken leg. The assailant was 6′2″ and intoxicated. The Court said:

… care must be taken to assure that our self-defense instructions afford women the right to have their conduct judged in light of the individual physical handicaps which are the product of sex discrimination. To fail to do so is to deny the right of the individual woman involved to trial by the same rules which are applicable to male defendants. ⁷

The Supreme Court of Canada in R v Lavallee ⁸ (discussed in detail below) adopted this reasoning and similarly recognised the relational context in which the violence occurs. In England and Wales, the use of a weapon is more likely inferred as indicative of some foresight and planning and therefore an intentional act compared to the use of body force which is more readily argued and accepted as being spontaneous. Where a weapon is used regardless of the relationship of victim to suspect, in both non-fatal and fatal assault, upon conviction lengthier sentences have followed. The use of body force to kill, until relatively recently, has been accepted by the courts as mitigation both at the defence and sentencing stages, to rebut intent or to reduce sentence length in both violent assault and in killing. The legal construction and framing of the use of a weapon as more intentional and more heinous than other methods of killing in loss of control and self-defence pleadings where women are defendants has worked to their great disadvantage.

Method of Killing

Less data is available in Canada on methods used in cases of intimate partner homicide. The Canadian Femicide Observatory for Justice and Accountability (CFOJA) reports that in 2016–2017, of the 927 homicides for which relationship information is available, 101 (16%) of the homicides occurred in intimate partner relationships. ¹⁹ The data confirm the Statistics Canada information cited above by showing that during the relevant period, females were more likely to be killed by an intimate partner than were males (42% vs. 3%, respectively). Female victims were also more likely to be separated from their partners at the time of the killing (32%), whereas all of the male victims were in a current relationship with their partner when they were killed. ²⁰

The CFOJA data also show that women killed by intimate partners during this period were more likely to be beaten, strangled, or suffocated and to be subject to excessive force (or ‘overkilling’). Firearms were more likely to be used when men were killed by their intimate partners, particularly in urban settings. ²¹

Statistics Canada's most recent Homicide Survey reports that in 2020, intimate partner homicides were most likely to be motivated by an argument or quarrel (31%); frustration, despair, or anger (29%); or jealousy or envy (13%). These percentages are not broken down by gender and do not include self-defence or fear of serious violence as possible rationales. ²² An earlier study found that men were more likely than women to kill their partners for reasons of jealousy (25% compared to 8% of cases involving male victims). ²³

Method of Killing and Legal Outcome Prognosis

Although there is no information in the Canadian data sets about self-defence or provocation claims, the data confirm what we might expect about the gender-based differences leading to intimate partner homicides, with women more likely to kill male partners with whom they resided and using weapons to do so in circumstances where they were abused, and men more likely to kill women who had ‘provoked’ them by leaving and using bodily force to do so. Under the Criminal Code, the use of a firearm in the commission of manslaughter results in a minimum punishment of four years imprisonment (s. 236(a)) with no minimum sentence prescribed for manslaughter using bodily force. At the Supreme Court of Canada, there are several decisions involving women who killed their abusive partners in circumstances that could qualify as self-defence, but no cases where the primary issue was provocation in these circumstances. ²⁴ The Court's rulings on provocation have typically involved males accused of intimate partner homicide and not females. ²⁵ These patterns hold true for lower court decisions as well, as sections ‘Canadian Law and Jurisprudence’ and ‘Canada’ will discuss.

The Legal Framework

In England and Wales, the current defences open to a charge of murder include involuntary manslaughter accident /no intent, voluntary manslaughter loss of control s. 54 CJA, manslaughter ‘fear of serious violence’ s. 55(3)(c), manslaughter diminished responsibility s. 52 CJA and gross negligence manslaughter. A conviction for murder requires proof of intention to kill or cause grievous bodily harm.

With regard to sentencing in cases of murder, the statutory framework is found in s. 269–277 CJA 2003 (now Sentencing Act 2020 (SA) s. 321, s. 322 and the Sentencing Code in Schedule 2–10 mirrors schedule 21 CJA 2003 [4–11]). There are three starting points for offenders depending on their age. The starting point for offenders aged 21 years and over is a whole life order, 30, 25 and 15 years. For 18–20 years old, the starting point is 30, 25 and 15 years, and for those under 18 years, a 12-year starting point. Culpability, seriousness and aggravating and mitigating factors determine sentence length. For convictions for manslaughter accident or manslaughter loss of self-control, the manslaughter guidelines apply ²⁶ and the sentence range is 3–20 years’ custody. For gross negligence manslaughter, the range is 1–18 years’ custody ²⁷ and for the diminished responsibility offence, the range is 3–40 years’ custody, a maximum of life imprisonment is however available for all manslaughter. ²⁸

Following the CJA 2009, the defences of manslaughter/provocation and manslaughter diminished responsibility under the common law and the Homicide Act 1957s. 3 and s. 2 were reformed, respectively, whilst self-defence (which will be discussed in the next section) and gross negligence manslaughter remained unaltered.

First, the CJA abolished the common law defence of ‘provocation’ (s. 56) ‘(1) The common law defence of provocation is abolished and replaced by s. 54 and s. 55’. Second, it limited what could constitute a ‘qualifying trigger’ under s. 55 by excluding or ‘disregarding’ the sexual infidelity excuse (s. 55(6)(c)). Third, it raised the threshold of admissible ‘triggers’ for loss of control now requiring ‘extremely grave circumstances’ s. 55(4)(a) and ‘a justifiable sense of being seriously wronged’ (s.55(4)(b)). Fourth, it was recognised that women who killed abusive partners killed out of fear in a self-defensive and self-preservatory response resulting in the introduction of a new ground or ‘qualifying trigger’ for loss of self-control by s. 55(3) ‘fear of serious violence’.

Reforming Provocation and the Limitation of the 2009 Act

The CJA 2009 was the product of sustained campaigning to address provocation's gendered inequity. Whilst broadly welcomed it is not without limitations being described at the time as a ‘dog's breakfast’ and criticised in feminist jurisprudence ³⁶ for retaining ‘loss of control’ with its gendered fixity whilst creating a special defence for women placing it out of their reach by requiring a high threshold of ‘serious violence’ before it is engaged. The requirement of ‘fear of serious violence’, originally proposed by the Law Commission, ³⁷ permits a lethal response but only where ‘serious violence’ is feared, establishing a defence similar to self-defence in impliedly requiring proportionality. Comparing this with s. 55(4)(a) ‘loss of control’ defence, a lethal response is permitted here where there are ‘extremely grave circumstances’ and where there is ‘a justifiable sense of being seriously wronged’ (s. 55(4)(b)). However, whilst imposing a restriction on excuses, ‘justifiable’ still allows a considerable margin of appreciation for jury determination. It is to be remembered that Lord Hoffmann, in R v Smith (2000) had criticised the imprecision of the earlier ‘sufficiently excusable’ formulation (under provocation) when he said: ‘I cannot agree with that formulation, for “sufficiently excusable” is too vague and prone to opening up the floodgates for whatever the jury considers exonerable’. ³⁸ It seems that despite criticism the insertion of the chameleon ‘justifiable’ continues to grant levity by sanctioning moral and legal judgements.

Further as has already been emphasised the overarching concept of ‘loss of control’ remains largely unaltered and its construction continues to be contiguous with anger which is further compounded by the statutory wording of ‘trigger’. ³⁹ At the bill stage, Lord Thomas of Gresford acceded that ‘loss of self-control’ aligns with male anger. ⁴⁰ The Law Commission in 2004, proposed the abolition of ‘loss of self-control’, which it said was: ‘a judicially invented concept, that lacked sharpness or a clear foundation in psychology’. ⁴¹ Sullivan, Crombag and Child ⁴² agree and consider that too much latitude has been given to sudden violence arising from anger and an over reliance on folk loric understandings of loss of self-control. They argue that failings in the early law especially an absence of a definition of loss of self-control and a clarity concerning whether we are looking for a partial loss of voluntariness or volitional capacity have persisted undermining the effectiveness of the defence. Douglas and Reed in a tour de force review and analysis of loss of self-control in England and Wales and Australia note how many have observed that ‘defence counsel still unfortunately transmogrify back to the utilisation of descriptors often heard in cases of common law provocation such as “snapped,” “went berserk,” “lost the plot,” and “the straw that broke the camel's back,” which are not necessarily helpful when deciding whether the partial defence has been made out’. ⁴³

Provocation may be said to be abolished but it has a continuing legacy in legal reasoning and lay understanding. Baroness Scotland of Asthal expressed her misgivings with the drafting: ‘there is, in reality, no obvious place for killings in fear of serious violence in a defence designed for angry reactions’. ⁴⁴ Fear of ‘serious violence’ sets a very high qualitatively and conceptually different bar when compared with the Act's provisions on the loss of self-control.

In addition, the overarching construct of ‘reasonableness’ also presides intact and requires the jury to consider how a person with a ‘normal degree of tolerance and self-restraint’ (s. 54(1)(c)) might have reacted. The terrified woman is unlikely to meet the standard of reasonableness, and the ‘serious violence’ threshold undermines any understanding of how a woman in anticipation of abuse, coercion and control might react and fails to appreciate how her reaction does not necessarily follow on from the severity of his threat or last act of violence or abuse. ⁴⁵ A woman's use of a weapon (see above) further undermines her lack of intent pleading. ‘Thus the Coroners and Justice Act 2009 despite some gains for abused women, the alchemy persists, and while certainly holding the abused woman within its contemplation for a moment she who kills out of fear with all its despair, hopelessness, sorrow, helplessness, anguish and trauma, is still required to lose self-control … [B]ut the legal template of loss of self-control … remains soldered to a male angered reaction with its outward demonstration embedded in a legacy of serious wrongs and justifiable hubris’. ⁴⁶

The Legal Framework

In Canada, provocation is a partial defence codified in the Criminal Code, which provides that ‘culpable homicide that otherwise would be murder may be reduced to manslaughter if the person who committed it did so in the heat of passion caused by sudden provocation’ (s. 232(1)). The significance of the reduction of the offence from murder to manslaughter is primarily related to the sentence for each offence, ⁶⁵ with a murder conviction resulting in a minimum sentence of life imprisonment with no parole for at least 10 years (for second-degree murder) or no parole for 25 years (for first-degree murder) (s. 235), and a manslaughter conviction resulting in judicial discretion to a maximum of life imprisonment, although as noted above, there is a minimum of four years imprisonment for manslaughter with a firearm (s. 236). As in England and Wales, a conviction for murder in Canada requires proof of intent to cause death or bodily harm (s. 229) and manslaughter is a culpable homicide that is not murder or infanticide (s. 234).

Prior to 2015, provocation was legislatively defined as ‘a wrongful act or an insult that is of such a nature as to be sufficient to deprive an ordinary person of the power of self-control … if the accused acted on it on the sudden and before there was time for his passion to cool’ (s. 232(2)). The Criminal Code was amended in 2015 by changing the definition of the triggering act of provocation to ‘conduct of the victim that would constitute an indictable offence under this Act that is punishable by five or more years of imprisonment’, still requiring that this conduct would be sufficient to deprive an ordinary person of self-control (objective element) and that the accused acted on this conduct suddenly and in the heat of passion (subjective element). ⁶⁶

The 2015 amendment was driven by a Conservative government's concerns about men who kill their partners or other family members and claim they were provoked into doing so to defend their family's honour. It was accompanied by other amendments with a similar focus on issues such as forced marriage and polygamy. The amendments were enacted quickly and with little consultation with legal experts, with their placement in legislation called the Zero Tolerance for Barbaric Cultural Practices Act highly controversial and critiqued for being racist and xenophobic in its ‘culturalisation’ of family violence. ⁶⁷

Although the 2015 amendment was motivated by so-called ‘honour killings’, it also has some potential to alleviate problematic invocations of the provocation defence by men who killed their partners when faced with adultery, separation, sexual insults or perhaps most problematically, ‘nagging’. ⁶⁸ Even before the amendment, in a 2010 case, R v Tran, the Supreme Court of Canada directed that the defence must be interpreted with equality principles in mind, such that the ‘ordinary person’ might, for example, be reasonably provoked by a racist slur, but not by adultery or ‘inappropriate conceptualizations of “honour”’. ⁶⁹ Tran also put this standard in terms of a ‘justifiable sense of being wronged’, mirroring the law in England and Wales. ⁷⁰

While this reasoning could be extended to apply to women who kill abusive partners when provoked by their violence, there was little attention paid to such cases during the Parliamentary hearings on the 2015 reforms. ⁷¹ Furthermore, coercive controlling violence is not recognised as a criminal offence in Canada that could trigger the 2015 version of the provocation defence – indeed there are currently no specific criminal offences related to domestic violence under the Criminal Code. ⁷² Triggering conduct by an abusive partner that amounted to assault (s. 266), sexual assault (s. 271), criminal harassment (s. 264) or uttering threats (s. 264(1)) would be included. However, coercive control that did not include these elements would not qualify as triggering conduct that allowed an abused woman to rely on the provocation defence if she killed her partner, even though coercive control may produce great danger to women when men's control is challenged. ⁷³ It is ironic that given the government's purported focus on vulnerable women in 2015, the pre-reform provision was wider in scope in its reference to a wrongful act or insult sufficient to deprive an ordinary person of self-control, which could have included coercive control.

There are few cases involving provocation arguments by women under either of these versions of s. 232 of the Criminal Code, with most of the case law involving intimate partner homicide and the provocation defence occupied by men who killed their female partners. ⁷⁴ Abused women who kill, whose fear may accumulate over time and who may use weapons, will typically have difficulty arguing that they acted in the heat of passion with a sudden loss of self-control. ⁷⁵ They may also lack corroborative evidence that helps provide an ‘air of reality’ to their claims, especially given women's under-reporting of violence to police and other authorities. ⁷⁶ As noted by commentators, the provocation defence remains ‘profoundly gendered’ and ‘rooted in the experience of heterosexual male rage’ ⁷⁷ as well as ‘masculinist ideals of ‘honour’ and ‘hot blood’. ⁷⁸

In Canada, women who kill abusive men may be more inclined to argue self-defence, a complete defence to murder if successful. ⁷⁹ The Supreme Court of Canada's contextual approach to self-defence by abused women starting with Lavallee (1990) is discussed in the next section. ⁸⁰ Although provocation could be relevant in trials where a self-defence argument fails or is unavailable, ⁸¹ it is difficult to determine the number of cases in which provocation has been argued. A study by Elizabeth Sheehy, Julie Stubbs and Julia Tolmie found 36 cases of abused women who were charged with homicide for killing their partners in Canada between 2000 and 2010. ⁸² Fifteen cases went to trial, and of these, 11 women were acquitted on the basis of self-defence while one was convicted of murder and three were convicted of manslaughter for reasons that were unknown (given the jury trial context) but could have included provocation. The authors found no cases during this period involving an explicit discussion of abused women killing in response to provocation, although there were 19 cases where women pleaded guilty to manslaughter and provocation may have been a factor in some of these. ⁸³

At the Supreme Court level, the only case of relevance is R v Malott, which was primarily a case about self-defence in circumstances where an abused woman shot and killed her estranged common-law partner and attempted to kill his girlfriend. ⁸⁴ Margaret Ann Malott claimed provocation in the alternative, but a jury found her guilty of second-degree murder, though recommending that she receive no more than the minimum sentence because of the severity of abuse she had sustained. On appeal, the Ontario Court of Appeal ruled there was no air of reality to her self-defence claim and also dismissed her provocation argument, stating that ‘even assuming that the deceased's actions could constitute provocation of the ordinary battered woman’, Malott had no basis for arguing that she was actually provoked, noting that she had put a gun in her purse before the encounter and the passage of time between her partner's actions and her reaction. ⁸⁵ The Supreme Court of Canada upheld Malott's conviction for murder, finding that the trial judge had properly instructed the jury on the available defences. Only the concurring judgment of Justices Claire L’Heureux-Dubé and Beverley McLachlin mentioned provocation, noting generally that evidence of domestic violence is relevant not just to self-defence claims, but also to ‘other situations where the reasonableness of a battered woman's actions or perceptions is at issue’, including provocation, duress and necessity. ⁸⁶

Cases analysing abused women's claims of provocation may also be rare because the Crown elects to charge them with manslaughter or agrees to accept a guilty plea to manslaughter in circumstances where provocation (or even self-defence) was arguable. ⁸⁷ A rare case discussing provocation pre-amendment is R v Ferguson, where an abused woman who shot and killed her partner while he was lying on the couch was convicted by a jury of manslaughter, having rejected her claim of self-defence. ⁸⁸ The sentencing judge found that Lisa Ferguson was subject to some provocation, noting he agreed with the implication of the jury's verdict as well as the Crown's expert that the passage of time between the deceased's last assault and the shooting ‘was not so long in the circumstances that provocation could not be considered’. ⁸⁹ Indeed, Ms Ferguson was prepared to plead guilty to manslaughter but the Crown insisted in proceeding to trial on a murder charge, despite its own expert's opinion that manslaughter was appropriate. ⁹⁰

Reforming Provocation and the Limitations of the 2015 Amendment

In 2019, the British Columbia Supreme Court considered the constitutionality of the 2015 amendment to the provocation defence in R v Simard, a case involving a man who shot and killed his ‘on again, off again’ girlfriend and a man he found her with. ⁹¹ Justice Douglas Thompson recognised that the history of the law of provocation ‘was a reflection of a patriarchal society’ that ‘focused on insults to a man's honour’ and found that the 2015 amendment was intended to ‘protect vulnerable women by removing any vestige of “honour” as a basis for invoking the provocation defence’. ⁹² However, he also noted that the provocation defence is not only used in those circumstances. ⁹³ Thompson J accepted a hypothetical argument raised by the defence that the 2015 provocation defence would not be available to a woman who had suffered a history of abuse and killed her partner while being taunted by emotional abuse that did not create fear that her safety was being threatened. ⁹⁴ He ruled that this and other hypothetical scenarios meant the 2015 amendment was unconstitutionally arbitrary and overbroad in precluding reasonable provocation arguments. ⁹⁵ Accordingly, he severed the part of the amendment that required provoking conduct to amount to an indictable offence, resulting in wording even broader than the pre-2015 version of the defence: provocation is simply ‘conduct of the victim that is of such a nature as to be sufficient to deprive an ordinary person of the power of self-control … if the accused acted on it on the sudden and before there was time for their passion to cool’ with no requirement for a wrongful act, insult or indictable offence. ⁹⁶

Although Simard technically only applies in British Columbia, ⁹⁷ it is potentially important for having recognised the problems with the 2015 version of the provocation defence for women who kill in circumstances where they experienced coercive control or other forms of domestic abuse that are not currently criminalised. ⁹⁸ However, there appear to be no reported decisions following Simard where an abused woman has sought to rely on its version of the provocation defence. Commentators call into question whether a broader approach to provocation has any utility in the actual circumstances in which women kill abusive partners (e.g. by using weapons) and also raise concerns that this broadening favours abusive men in circumstances that reinforce male power, control and rage. ⁹⁹

Generally, then, the provocation defence has not provided much assistance to women who kill abusive men in Canada either before or after the 2015 amendment. Some feminist academics and activists have called for the abolition of this defence, similar to the approach undertaken in New Zealand and some Australian states, where abolition has gone hand in hand with eliminating mandatory minimums and sometimes, with strengthening self-defence in the context of family violence. ¹⁰⁰ Reforming the provocation defence may be too difficult in light of its discriminatory underpinnings, and if the only concrete advantage to claiming provocation is to avoid the minimum sentence for murder, the argument is that the sentencing regime for homicide should be reformed instead. ¹⁰¹

The Legal Framework

Turning now to the law on self-defence this has been largely impervious to self-defence pleadings by women who defend themselves from the abuse and/or violence of an intimate partner. Relying on the common law and s. 3(1) Criminal Law Act, 1967, CLA, self-defence requires first an objectively reasonable use of force in the circumstances ‘as he [subjectively] [and genuinely] believes them to be’ ¹⁰² and second an immediate, necessary and proportionate response (a duty to retreat is a relevant factor). ¹⁰³ It is an absolute defence to any crime.

The constructs of ‘reasonableness’ and ‘proportionality’ which operate within the law and outside the law in the minds of jurors limit an abused woman's access to the defence, since in delaying her response and by using a weapon her response is considered disproportionate.

At trial where the judge considers that there is no evidence of self-defence then applying that other chameleon construct of ‘common-sense’ ¹⁰⁴ it is their duty to withdraw it. ¹⁰⁵ ‘The question is one for the trial judge to answer by applying common sense to the evidence in the particular case … self-defence should be left to the jury when there is evidence sufficiently strong to raise a prima facie case’. ¹⁰⁶ Judicial guidance on instructions to jurors in the Crown Court Compendium state that self-defence is ‘just common sense’ ¹⁰⁷ and that jurors should be instructed in applying common sense to consider whether the force used was ‘necessary’ and ‘reasonable’ in the circumstances. ¹⁰⁸ The Crown Prosecution Service (CPS) guidance, for example, assumes a correspondence between the gravity of circumstances with which a person may be presented and the force used: ‘As a general rule, the more extreme the circumstances and the fear felt, the more force you can lawfully use in self-defence’. ¹⁰⁹

Limitations

Legal scholarship in Anglo-American jurisprudence has questioned for over 50 years the declared objectivity of the principles of ‘reasonableness’, ‘proportionality’ and also the notion of ‘common sense’. Joshua Dressler ¹¹⁰ and Elizabeth Schneider ¹¹¹ pointed out that the law on self-defence developed in a male-oriented legal culture where the interpretation of ‘reasonableness’ was constructed with a male way of seeing. ¹¹² Celia Wells noted: ‘Behind it lies an image, a stereotype, of isolated, one-to-one (man to man) violence’. ¹¹³ In England and Wales, self-defence pleadings by abused women, owing to judicial and/or finders of fact constructions of what is a ‘reasonable’ or ‘proportionate’ reaction to abuse or the threat of abuse from an intimate partner, especially where a weapon is used have invariably failed. ¹¹⁴

In very few cases where women kill abusive and violent male partners is a defence of self-defence successful, even less attempted. Elizabeth Hart-Browne, ¹¹⁵ who was hit and grabbed by the throat by the deceased and in defending herself picked up a kitchen knife and stabbed him, provides one of the few cases where fear of a partner's violence resulted in a successful self-defence pleading. Reasonable force remains elided with a notion of male proportionality, such that leaving jurors to determine what is a reasonable response of an abused woman depends on their appreciation and understanding of the situation of a woman who is abused, coerced, and controlled.

Research from the Centre for Women's Justice found of the 92 cases of women who killed partners in their study (2008–2018), six were acquitted of self-defence ¹¹⁶ and in a further 14 cases, where a defence of self-defence failed, women were convicted of manslaughter or murder. ¹¹⁷ In a study of cases in Scotland, 11 of 62 women who killed intimate partners who abused them pleaded self-defence. Three women were acquitted of homicide, three were convicted of murder, and five women were convicted of culpable homicide. ¹¹⁸

Whilst abused women are judged by self-defence's high threshold householders who are invariably men are permitted a concession under s. 76[5A] Criminal Justice and Immigration Act, CJIA. This Act was driven by public policy permitting a ‘disproportionate’ but not ‘grossly disproportionate’ use of force when confronted by an intruder/trespasser who had entered the house, thus allowing a person who uses a weapon to kill under these very strictly defined circumstances. The unruly horse of public policy has not made any concession for women who defend themselves in their own homes from an unwelcome domestic violence abuser. During the passage of the Domestic Abuse Act 2021 the amendment to extend the householder concession to abused women who killed abusers in the abused woman's home was rejected by the government who considered existing defences were adequate. ¹¹⁹ The government minister said: ‘it is inappropriate to extend the so-called “householder defence” to victims of domestic abuse who use disproportionate force against their abusers in self-defence’. ¹²⁰

The Legal Framework

Self-defence is codified in Canada's Criminal Code and provides a justification for homicide if its requirements are made out (s. 34). Prior to 2012, self-defence required that the accused (1) had caused death or grievous bodily harm ‘under reasonable apprehension of death or grievous bodily harm from the violence with which the assault was originally made or with which the assailant pursues his purposes’, and (2) ‘believed on reasonable and probable grounds that they could not otherwise preserve themselves from death or grievous bodily harm’. In 2012, this provision was amended to require that the accused (1) believed on reasonable grounds that force or a threat of force was being used against them or another person, (2) the act that constituted the offence was committed for the purpose of defending or protecting themselves or the other person from that use or threat of force and (3) the act committed was reasonable in the circumstances. ¹²¹ The 2012 amendment also deems several factors relevant to the consideration of reasonableness, including the nature and imminence of the force or threat and the nature and proportionality of the accused person's response to it; use or threatened use of a weapon by either party; the size, age, gender and physical capabilities of the parties; and the relationship between the parties, including any prior use or threat of force (s. 34(2)).

Limitations

The 2012 amendment was part of a package of reforms directed in part at protecting property owners who assaulted or killed intruders, ¹²² but it also responded to feminist law reform proposals for the law of self-defence, in particular by including factors to provide context for the jury's reasonableness inquiry. ¹²³ Commentators have expressed concerns that while the amendment may expand the availability of self-defence arguments in some circumstances that would include women who kill abusive partners, the situation of such women may also be worsened by adding explicit considerations of ‘imminence’ and ‘proportionality’ in assessing the reasonableness of the accused's actions. ¹²⁴ Another concern is that the added factors relevant to reasonableness do not include the systemic factors that feminists called for, including how individual actions (e.g. failing to leave or report abuse to the police) may be shaped by sexism, racism, and other inequalities. ¹²⁵ However, the Supreme Court recognised that such systemic considerations are relevant to the interpretation of self-defence in Lavallee in 1990, so they remain arguable. ¹²⁶

Lavallee was the case of an abused woman who shot her partner in the back of the head following an altercation in which he had physically abused her and taunted her that if she did not kill him, he would get her, handing her a gun. Taking a contextual, equality-informed approach to Angelique Lavallee's self-defence claim, Justice Bertha Wilson held that expert testimony was admissible to assist the jury in understanding her ability to perceive danger from her partner and hence whether she ‘reasonably apprehended’ death or grievous bodily harm, as well as the reasonableness of her belief that killing her partner ‘was the only way to save her own life’. ¹²⁷ On the latter point, Justice Wilson recognised that ‘popular mythology about domestic violence’ includes the misconception that women who do not leave their partners were either ‘not as badly beaten as [they claim]’ or ‘must have stayed out of some masochistic enjoyment of it’. ¹²⁸ As mentioned above, Justice Wilson quoted from the US case of Wanrow in noting that self-defence instructions to the jury should ‘afford women the right to have their conduct judged in light of the individual physical handicaps which are the product of sex discrimination’. ¹²⁹ A majority of the Court restored the jury's acquittal of Ms Lavallee without sending the matter back for re-trial.

Lavallee was a positive decision for its recognition of the need to contextually consider self-defence in the case of abused women. However, many feminists critiqued the Court's reliance on ‘battered women's syndrome’. ¹³⁰ This critique was acknowledged in Malott, discussed above, where Justices L’Heureux-Dubé and McLachlin noted the importance of avoiding a ‘new stereotype’ of the battered woman as ‘victimized, passive, helpless, [and] dependent’, which may adversely impact ‘women who have demonstrated too much strength or initiative, women of colour, women who are professionals, or women who might have fought back against their abusers on previous occasions’. ¹³¹ While this broadening of the context relevant to abused women from a range of social identities was welcome, L’Heureux-Dubé and McLachlin JJ agreed with the majority that the trial judge had adequately instructed the jury on the available defences and upheld Malott's conviction for second-degree murder. ¹³²

In another decision from the Supreme Court, R v Gladue, the Court established a new sentencing approach for Indigenous offenders whereby the systemic racism and colonialism affecting Indigenous peoples in Canada must be considered. ¹³³ However, the Court said little about the specific circumstances of the accused in this case, Jamie Gladue, who had stabbed and killed her common-law partner, other than to note that the offence ‘involved domestic violence and a breach of the trust inherent in a spousal relationship’ (referencing her breach of trust, not his). ¹³⁴ Gladue had pleaded guilty to manslaughter and the Court did not discuss whether a self-defence argument might have been available to her. Sheehy's research concerningly shows that Indigenous women who kill their male partners are more likely to plead guilty and forgo self-defence arguments even if they are legitimate on the facts, and also less likely to be successful in raising self-defence claims. ¹³⁵

Consistent with Malott and Gladue, Sheehy's analysis of decisions post-Lavallee indicates that many abused women who kill their partners have not benefitted from the Supreme Court's decision in Lavallee. ¹³⁶ She notes that abused women often accept plea bargains and plead guilty to manslaughter so as to face a reduced sentence, rather than risk conviction for murder if self-defence arguments fail. ¹³⁷

This was the case in Naslund (2022), which involved a woman who shot and killed her sleeping husband after enduring years of domestic violence, including on the day of the killing. ¹³⁸ Helen Naslund was charged with first-degree murder, but the Crown accepted a guilty plea to manslaughter in exchange for a joint sentence submission of 18 years imprisonment. This sentence was overturned by a majority of the Alberta Court of Appeal. Justice Sheila Greckol importantly recognised that Ms. Naslund was in a ‘distinctly vulnerable position’ in her plea negotiations with the Crown ¹³⁹ and acknowledged that there may be ‘compelling reasons why a battered woman might decide not to risk arguing self-defence at a murder trial’. ¹⁴⁰ She noted that abused women may be ‘irresistibly’ driven to plead guilty because of myriad factors, including ‘the desire to protect children; the worry of publicly discussing family domestic violence in court; and the risk a jury will not understand [their] actions as defensive’ as well as women seeing themselves as ‘worthy only of harsh punishment’. ¹⁴¹ She also critiqued the reasons of the sentencing judge, concurred in by a dissenting justice at the Court of Appeal, who maintained that Ms. Naslund had ‘options’ available to her (i.e. to leave her abusive husband) even though Lavallee debunked this myth in 1990. ¹⁴² The abuse Ms. Naslund sustained was found to be a mitigating factor for sentencing purposes and a sentence of nine years was substituted. ¹⁴³

Concerns, therefore, remain that the law of self-defence in Canada does not adequately protect the interests of abused women who kill their partners, even following the 2012 amendment. ¹⁴⁴ This is the case not only at the level of courts and juries but also in terms of lawyer conduct. That Helen Naslund's own lawyer appears not to have counselled against accepting an 18-year sentence submission for manslaughter in circumstances where self-defence was arguable is deeply concerning. It is also noteworthy that courts – such as in Naslund – continue to use the concept of ‘battered women's syndrome’ rather than the more modern understanding of coercive control, perhaps because the latter is not recognised criminally in Canada. ¹⁴⁵