Briefing paper prepared for use during virtual expert

roundtable on litigating self-defence in the context of

domestic violence, organised by EHRAC on 29
September 2020
Constantin Cojocariu

Women may kill in reaction to domestic violence in a variety of scenarios, not

limited to the generic case of a woman responding in kind to a direct attack
from her partner. Killings may occur in immediate response to a flare-up of
violence, or they may be opportunistic, with evidence of premeditation or with
the defendant taking advantage of the attacker being asleep or in a situation of
vulnerability. The defendant’s response may be intended to avert a risk to self
or to other family members. In turn, the defendant may be the partner/wife, or
the children.

Each scenario presents distinct challenges in terms of the ability of the

defendant to rely on defences or mitigations available under domestic law.
While violence is a generic feature of such relationships, there have been
efforts in many jurisdictions to capture other types of behaviours that may
amount to domestic abuse, and which may result in a woman killing her
abusive partner. For example, in England and Wales, in December 2015, a
new offence of coercive behaviour was legislated with the aim of improving the
likelihood of successful prosecutions in cases where physical violence was not
present or was not the only form of domestic abuse experienced by the victim.

This briefing paper addresses some of the issues arising when litigating cases
on behalf of women charged with harming their partners in self-defense.

Claiming self-defence

Restrictive or inflexible definitions often prevent women from taking advantage

of self-defence following domestic altercations that result in killing or seriously
harming the attacker.

Several generic elements of self-defence have given rise to particular problems

in this context, being taken as proof of inherent sexism in the law. The
requirement that the attack justifying defence must be imminent has often
been interpreted as not permitting any time gap to elapse between the two
moments. The requirement of proportionate response implicates questions
around reaction that is perceived as unjustifiably violent. The requirement of
retreat often misses the point that battered women may lack any good options
for leaving an abusive relationship.

Lawyers in some jurisdictions have been able to argue that the requirements of
self-defence should be interpreted flexibly to take due account of

1
circumstances pertaining in abusive relationships. Some countries adopted
legislative amendments that modified the definition of self-defence, guided by
similar considerations. In the Australian state of Victoria, legal amendments
introduced in 2014 specified the range of evidence that can be adduced about
the history of the relationship, and the nature of violence in the relationship, to
prove both the subjective (a belief in the necessity of using force) and the
objective (the existence of reasonable grounds for the belief) elements of the
test for self-defence. These provisions also allow for the introduction of “social
framework evidence” that permits evidence of the nature and dynamics of
domestic violence to be introduced with a view to dispelling myths about
domestic violence that exist in the community.

Beyond self-defence

The facts in each case and/or the constraints of each national legal system
may require exploring other options for arguing that a history of domestic
abuse should be taken into account during criminal trial, besides self-defence.

Different countries may allow alternative defences that may be relevant in the
context, such as duress, provocation or temporary insanity; or partial defences
such as loss of control or excessive self-defence; or yet, a history of domestic
abuse may constitute a mitigating circumstance during sentencing. Some
countries/territories introduced standalone defences designed to address the
situation of women who kill in reaction to domestic violence. For example, in
2010, the Australian state of Queensland introduced a partial defence to
murder or killing for preservation in the context of an abusive relationship.

These alternatives are not free from practical or principled problems,

depending on local circumstances. However, there is no reason why self-
defence should be available in each and every case of killings in response to
domestic abuse. Self-defence may objectively provide a limited framework to
address the variety of scenarios in which killing in response to domestic
violence occurs. With that being said, compared to other defences or
mitigating factors, self-defence has the advantage of usually resulting in full
acquittal without any blame assigned to the defendant.

The Sally Challen1 case in England involved psychiatric evidence

demonstrating the effects of prolonged coercive behavior, which justified a
plea of diminished responsibility. Challen killed her husband of 31 years in
2010 with multiple hammer blows. Her husband allegedly bullied and
belittled her, controlled their money and who she was a friend with, and did
not allow her to socialise without him. After convoluted legal proceedings, the
Crown Prosecution Services accepted Challen’s guilty plea to manslaughter on
the grounds of diminished responsibility and sentenced her to nine years in
prison. This was based on new evidence that surfaced during trial proving that

1
This website: https://www.justiceforwomen.org.uk, contains more information about the Challen
case, as well as other cases of women who kill in self-defense.

2
she was living in a coercive and controlling relationship, and that the long-term
effects of this led her to kill her husband.

Stereotyping

Aside from definitional constraints, harmful gendered stereotypes regularly

interfere with how women defendants are treated during trial and with how
defences are deployed in the context of domestic violence. One such
stereotype is that of the ideal victim, intimating that women should be weak,
passive and helpless. Furthermore, far from acknowledging the implications of
having suffered regular abuse from the hands of their partners, courts in many
countries treat women offenders very harshly and apply disproportionate prison
sentences. Litigants should be awake to the existence of these stereotypes and
evaluate the authorities’ performance though a gendered lens.

The Jaqueline Sauvage2 case in France illustrates well how gender stereotyping
may distort the proceedings and outcome of a trial of a battered woman who
killed her husband.

Sauvage had been married for 40 years, with four children, when, on 10
September 2012, she took a hunting rifle and shot her husband in the back
three times, killing him. This took place on the day after the couples’ 15-year
old son committed suicide and a few hours after another domestic altercation.

During trial, the prosecution argued that Sauvage “executed” her husband and
that her actions were premeditated, relying on the lapse of time between the
earlier altercation and the later killing. The defence, on the other hand,
presented evidence that Sauvage and her children suffered from 40 years of
physical, sexual and psychological abuse, and that this should be considered
as meaning that her reaction was purely defensive for the purposes of
preserving her life and those of her children.

Sauvage was eventually convicted of murder and sentenced to ten years in

prison, an outcome that was upheld on appeal. She was eventually pardoned
by President Hollande, following a petition signed by over 400,000 people.

The institutional actors involved in Sauvage’s trial often deployed gendered

and discriminatory narratives, which sought to undermine her credibility as a
witness. Thus, Sauvage was often portrayed as a woman capable of violence
herself, by recounting that she was an experienced hunter or that she
assaulted her husband’s mistress. She was also represented as a hardworking,
authoritarian person who valued her work success and social status. This in
turn was implicitly contrasted with the idea that longstanding victims of
domestic violence must be passive and helpless. The prosecution also insisted

2
Kate Fitz-Gibbon, Marion Vannier, Domestic Violence and the Gendered Law of Self-Defence in
France: The Case of Jacqueline Sauvage, Feminist Legal Studies, November 2017.

3
on the fact that she was a bad mother and that she did not display emotion
during trial.

A history of domestic abuse: evidentiary challenges

Ultimately, regardless of the procedural means used for mitigating culpability

or the sentence applied, a more coherent advocacy target may be to argue that
courts have an obligation to consider the implications of a history of domestic
abuse during the trial of women charged with killing or harming her partner.
This is consistent with the ECtHR’s oft-stated position that incidents of
domestic violence should be viewed as a continuum of harm and the overall
violence regarded as a chain of connected events.

Courts often rely on the “battered woman’s syndrome” (BWS), with the
assorted notions of “cycles of violence” and “learned helplessness”, which was
developed by Lenore Walker in 1979 to explain the dynamics of domestic
abuse during trial.3 The syndrome consists of a cycle theory and a theory of
learned helplessness. The cycle theory postulates that male violence against
women typically follows a repeated three phase pattern: a period of heightened
tension, a sudden eruption of violence from the man following some small
trigger, and a loving contrite phase during which the male pleads for
forgiveness, is affectionate, and swears off violence.

Learned helplessness is a theory which suggests that the randomness and

apparent unavoidability of a woman’s beatings lead her to accept her fate and
to develop a number of common characteristics, such as low self-esteem, self-
blame for the violence, anxiety, depression, fear, general suspiciousness, and
the belief that only she can change her predicament.

BWS became instrumental for pleading mitigation in many jurisdictions and is

often mentioned in documents issued by international organizations. However,
it has been criticized4 for being scientifically unsound, as it only accounts for a
fraction of abusive relationships. In addition, it has well-documented
pernicious effects in terms of generating a false expectation about how abused
women act, distorting facts by focusing attention on the disabling effects of
violence rather than the women’s survival skills, and by pathologizing abused
women. A Council of Europe manual suggests that “because there is some
controversy over the use of evidence of the battered women syndrome, it is
generally recommended that judges not use the term in court unless it has
been raised by the defense.”5

3
L.E. Walker, The Battered Woman (New York: Harper & Row, 1979).
4
For the U.S. see: Martha R. Mahoney, Misunderstanding Judy Norman: Theory as cause and
Consequence, 51 Conn. L. Rev. 671 (2019); for Canada: Kwong-leung Tang, Battered Woman
Syndrome Testimony in Canada: Its Development and Lingering Issues, International Journal of
Offender Therapy and Comparative Criminology, 47(6), 2003; for the UK: Janet Loveless, R. v GAC:
battered woman "syndromization", Crim. L.R. 2014, 9, 655-667.
5
Council of Europe, Training manual for judges and prosecutors on ensuring women’s access to
justice (2017), p. 92.

4
It has been suggested that instead of BWS, expert evidence should focus more
generically on “intimate partner violence and its effects,” based on more
widely accepted theories from social science such as the “survivor theory”
(based on the idea that “battered women respond to severe abuse with
innovative coping strategies and active help seeking”) or “coercive control”
(which would focus on the person’s need to free themself from circumstances
of coercive control). Litigants should ask for expert evidence to explain the
effect of domestic abuse and the abused woman’s reactions, as well as its
relevance to the legal requirements of the defense in question.

In practice, gathering a record of abuse may prove challenging, due to the

well-documented under-reporting among women victims of domestic violence
in many countries, or the nature of the behavior alleged, such as coercive
control. These difficulties are aggravated by widespread disbelief and
secondary victimization during trial, particularly where physical or medical
evidence is absent or patchy.

The ECtHR perspective

As far as I can tell, the ECtHR has not examined any cases involving women
charged with harming their partners after a history of domestic violence,
including with respect to the quality of legal provisions on defenses and their
application in practice. Kalucza v. Hungary involved the refusal to issue a
restraining order against the background of apparently reciprocal violence with
both partners alleging to have acted in self-defense. On this occasion, the
ECtHR acknowledged that women of domestic abuse may repulse an attack in
“legitimate self-defense”.

One could argue cases on behalf of women charged with killing their partners
from several angles. An Article 6 claim (with Article 14) may be based on the
authorities’ refusal to consider any evidence of a history of domestic abuse, as
well as on treatment during investigation and trial, to the extent that it
revealed any gendered stereotypes influencing the outcome for the defendant.
An Article 2 claim (with Article 14), predicated on the fact that the defendant
was protecting her own life when responding to violence, may be based on the
obligation to have in place a legal framework that is capable of protecting life.
An ancillary claim under Article 2 or 3 of failure to prevent and end domestic
violence is an almost inevitable coda to the main claim on self-defense.

Other legal standards

There is limited support in international law specifically addressing the

situation of battered women charged with harming their partners in self-
defense.

The most comprehensive articulation of standards in the area is a General

Recommendation issued in 2018 by the Committee of Experts charged with

5
evaluating the implementation of the Belém do Pará Convention in the
Americas.6 The recommendation describes the international obligations of
State Parties “to ensure women’s ability to argue self-defense in cases where
they were themselves the victims of violence”, by addressing the elements of
the definition of self-defense and the evidentiary difficulties facing women at
trial. It concludes as follows:

“The application of gender perspective in trials where women victims of

violence are accused of killing or injuring their attacker in legitimate defense
of their or a third party’s rights […] requires a paradigm shift in how the acts
should be evaluated, and the law interpreted, resulting in the elimination of
the gender stereotypes that govern our society and criminal justice system from
all legal reasoning. In other words, incorporate a contextual analysis that
makes it easy to understand that the reactions of victims of gender-based
violence cannot be measured according to the traditional standards used to
meet the bar for self-defense in other types of cases, since the violence they
experience at the hands of their attacker based on their gender has specific
characteristics that should permeate the entire legal reasoning of the
decision.”

In 2015, the CEDAW Committee published its views in X. v. East Timor,

involving a woman (X) who fatally stabbed her partner, while trying to defend
herself from his violent attack. Prior to this, X. suffered years of severe
domestic violence, witnessed by the family and reported to the authorities. X
was found guilty of aggravated homicide and sentenced to fifteen years in
prison, subsequently reduced to seven years.

The Committee considered two issues: whether the State party protected X.
from domestic violence prior to the fatal incident, and whether the authorities
ensured that she received a fair trial, “without bias, discrimination or gender
stereotyping”. On this latter point, the Committee found that the authorities:

“…failed to provide [X] with medical care after her arrest, inform her of her
rights, provide counsel at her first interview or collect evidence that would have
aided her defence; kept [her] in detention for a great deal longer than is
provided for by the law, despite her being a breastfeeding mother; failed to
provide [her] with psychosocial support after her arrest appropriate to a person
claiming to have been attacked and to have killed in self-defence; failed to
ensure, when it appointed counsel, that the assistance provided was effective
[…]; and finally that judges, despite a retrial being granted on the basis that
self-defence had not been duly considered in the first trial, allowed gender
stereotypes and bias to affect the weighing of evidence in the second trial, in
particular by lending [her] voice less credence than that of her nephew, who
had not been present at all relevant times.” Furthermore, during the initial
proceedings, X had been told that, “as a wife, you must protect your husband”,

6
Committee of Experts (CEVI) of the Follow-up Mechanism of the Belem do Para Convention
(MESECVI), ‘General Recommendation N. 1 of the Committee of Experts of the MESECVI’ (2018).

6
which “showed a pattern of deeply held bias that continued into the retrial and
has been enormously detrimental to the life of the author and her son.”

In view of these findings, the Committee asked the State Party to grant X a full
pardon and compensation.

CEDAW also published two useful recommendations on access to justice and

on gender-based violence. The OHCHR has published a report on “gender
stereotyping as a human rights violation.”7 Additional references from UN
reports and other sources, which are relevant for our purposes, are listed in the
appendix.

Other sources

In 2016, Penal Reform International published a very useful comparative

review of criminal legislation on women who kill in response to domestic
violence, covering nine jurisdictions: Australia, Hong Kong, India, Japan,
Mexico, Poland, Spain and the United States.8

Appendix: UN references to the situation of women who kill in response to DV

Report of the Special Rapporteur on violence against women, its causes

and consequences, Rashida Manjoo - Addendum Mission to Kyrgyzstan

• Allow for a defence based on a prior history of being a victim of domestic

violence as a mitigating factor in the sentencing of women who have killed
their partners and/or other family members.

Report of the Special Rapporteur on violence against women, its causes

and consequences, Yakin Ertürk -Addendum Mission to Tajikistan

37. In other cases, women have resorted to killing their abusers. Many of
the women detainees I met in the Nurek prison were convicted for murdering
their partners, whether during a fight or in a premeditated manner. They were
condemned to heavy sentences ranging from 7 to 20 years in prison.
Imprisonment of women is particularly devastating for family maintenance and
childcare. The Presidential amnesty which resulted in the release of most
women detainees in 2006 did not include those women who committed
murder. While recognizing the gravity of their crimes, women’s criminality
under situations of extreme abuse and violence needs to be treated with
diligence, and their cases must be assessed in light of mitigating
circumstances.*

7
CEDAW, ‘General Recommendation No 33 On Women’s Access to Justice’ (2015); CEDAW,
‘General Recommendation No 35 On Gender-Based Violence Against Women, Updating General
Recommendation No 19’ (2017); Cusack S, ‘OHCHR-Commissioned Report: Gender Stereotyping as
a Human Rights Violation’ (October 2013).
8
Linklaters LLP, ‘Women who kill in response to domestic violence’ (Penal Reform International,
2016).

7
· Review sentences against women detainees who murdered their partners
because of domestic violence, taking into account the mitigating
circumstances around their crime

* The “battered women” syndrome/defence may shed light into the

circumstances in which a violent act was committed. Courts in a number of
countries have acknowledged that battered women use force or kill as a way to
defend themselves or respond to provocation. As a result, courts have given out
more lenient sentences. See e.g. Thornton (No. 2) [1996] All ER 1023, R. v.
Kiranjit Ahluwalia [1992], R. v. Lavallee, [1990] 1 S.C.R. 852.

Human Rights Committee - Concluding observations on the fifth periodic

report of the Sudan

21.The Committee notes the case of Noura Hussein, a woman subjected to a

forced marriage at the age of 16, who was sentenced to death after she
stabbed and killed her husband in self-defence, after the latter attempted to
rape her. The Committee welcomes the quashing, in June 2018, of Noura
Hussein’s death sentence, and its replacement with a five-year prison
sentence. The Committee welcomes the oral assurance from the delegation
that despite an ongoing appeal by the State prosecutor, there will be no
reinstatement of the death penalty in her case. The Committee is concerned
nonetheless that the gender-based violence to which she had been subjected
was not taken into account as evidence by the court (arts. 2, 6–7, 14 and 26).

22. The State party should ensure that Noura Hussein is not subjected to the
death penalty and reconsider her five-year jail sentence.

Report of the Special Rapporteur on violence against women, its causes

and consequences on her mission to Georgia

(j) Consider including a provision on battered women syndrome as a mitigating

circumstance in cases where a wife exposed to long-term domestic violence
kills her spouse;

Committee on the Elimination of Discrimination against Women -

Concluding observations on the combined fourth and fifth periodic
reports of Croatia

(c) Practice of dual arrests, whereby women who are

victims of domestic violence are being arrested and
occasionally sanctioned, along with alleged aggressors, for
being verbally insulting or defending themselves;

(c) To abolish the practice of dual arrests in cases of

domestic violence;

Resolution adopted by the General Assembly [on the report of the Third

8
Committee (A/65/457)] 65/228. Strengthening crime prevention and
criminal justice responses to violence against women

III. Criminal procedure

15. Member States are urged to review, evaluate and update their criminal
procedures, as appropriate and taking into account all relevant international
legal instruments, in order to ensure that:

(d) Evidentiary rules are non-discriminatory; all relevant evidence can be

brought before the court; rules and principles of defence do not discriminate
against women; and “honour” or “provocation” cannot be invoked by
perpetrators of violence against women to escape criminal responsibility;

(k) Claims of self-defence by women who have been victims of violence,

particularly in cases of battered woman syndrome,* are taken into account in
investigations, prosecutions and sentences against them;

* Battered woman syndrome is suffered by women who, because of repeated

violent acts by an intimate partner, may suffer depression and are unable to
take any independent action that would allow them to escape the abuse,
including refusing to press charges or to accept offers of support.

United Nations Rules for the Treatment of Women Prisoners and Non-
custodial Measures for Women Offenders (“the Bangkok Rules”)

III. Non-custodial measures

Rule 57 The provisions of the Tokyo Rules shall guide the development and
implementation of appropriate responses to women offenders. Gender-specific
options for diversionary measures and pretrial and sentencing alternatives shall
be developed within Member States’ legal systems, taking account of the
history of victimization of many women offenders and their caretaking
responsibilities.

Rule 61 When sentencing women offenders, courts shall have the power to
consider mitigating factors such as lack of criminal history and relative non-
severity and nature of the criminal conduct, in the light of women’s caretaking
responsibilities and typical backgrounds.

However, once again, women are often discriminated against in the application
of this principle, due to one or a combination of a series of factors. Firstly,
since the same classification instruments are used for women and men in the
vast majority of prisons worldwide, despite women’s different needs and
circumstances, information about a history of domestic violence, sexual abuse,
and parental responsibility are areas in which screening is lacking for women.
As a result, classification and screening procedures do not provide essential
information about the women, which may increase the probability of their

9
placement in a higher security level than appropriate, while reducing
possibilities of providing suitable prisoner programmes matching individual
needs.

10