Fear, loss of control and cognitive neuroscience

Authors: Dr Lisa Claydon and Caroline Rödiger [1]

Abstract: This article considers the introduction of the partial defence to murder of loss of control in England and Wales. It examines whether the structuring of one of the triggers of the defence around the need for fear of serious violence will be helpful to jurors. The article looks at the case law of three jurisdictions: England and Wales, Germany and Australia. It considers a case from each jurisdiction which required the evaluation by the court as to whether someone who killed his or her abusive partner deserved to be excused criminal responsibility. The article considers how the fear of serious violence might be interpreted in the future by the English courts; and whether the jury will be able to appreciate the circumstances of an abused person when evaluating his or her actions. It will also consider how expert evidence may or may not help and what neuroscience tells us about the emotional states of anger and fear. Finally, it concludes by considering the question are these emotional states separable and, if they are not, will that pose a difficulty for jurors?


This article will examine the difficulty for the courts in considering the responsibility of those who kill, out of desperation, the partner who has abused them. Case reports from around the world show that courts struggle to do justice when someone intentionally kills the person who has violently mistreated them. In addition to examining case law from England, Australia and Germany it will consider the reform to the law of England and Wales [2] which introduced loss of control as a partial defence to murder. It will argue that there are flaws in the present loss of control defence.

The difficulty for the law in this area is that of distinguishing between types of violent killing. Most systems of criminal law view intentional killing as a most heinous crime. Whereas an intentional, violent killing which occurs because someone reacts after years of abuse is seen, in many societies, as less blameworthy. This is the case even where there is evidence of premeditation. The problem is how to identify matters of legal relevance that may partially excuse one killing, where the victim is the abuser; whilst not allowing an excuse to another killer who has committed the same act, but in very different circumstances. Often such determination will require an investigation of the emotional and other circumstances experienced by the person who kills. The relevant circumstances are not easy to identify and may stretch over the whole time period of the abuse. Additionally, the experience of violent abuse and its effect on human behaviour is not something of which most courts or jurors will be cognisant.

In the case of violent killing it is arguable that some emotional states trigger behaviour which is excusable and other emotional states do not. This article will consider whether a greater understanding of the work of cognitive neuroscience could provide a stronger basis from which to understand the present English partial defence to murder of loss of control. Part of the reason for the introduction of the English loss of control defence was that it aimed to provide women who killed their abusive partners with an arguable partial defence to murder. The additional aim of the reform of the English law, which should not be forgotten, was to prevent people who killed their partners from claiming the loss of control defence in circumstances where they acted out of jealousy or possessiveness. This requires that the defence provides a means of distinguishing between angry jealous reactions and the reactions of those who kill to escape their abuser.

In England loss of control is a partial defence to a murder charge. It replaced the previous partial defence of provocation, which according to one commentator was, “bound to encourage a view of human behaviour which is sexist, homophobic and racist.” [3] Since its introduction the courts have on occasions struggled with interpreting the legal meaning of the structured elements that must be proved before the defence can be established. This is particularly true of a series of cases where men were charged with murder and on the evidence the killing of their partner was, in part, motivated by possessiveness or jealously. [4] Partial defences, if accepted by a jury, mean that the killer though charged with murder is convicted of the lesser homicide offence of manslaughter. If the verdict is manslaughter the sentence is then determined by the trial judge and a wider range of sentencing options are available. If a killer is convicted of murder then, under English law, the judge must impose a mandatory life sentence.

As stated previously the defence of loss of control was, in part, an attempt to deal with considerable criticism of the failure of pleas of provocation by women charged with murder who had suffered considerable domestic abuse. The cases of Emma Humphreys [5] and Kiranjit Ahluwalia [6] are two examples where the abuse suffered by both women at the hands of their partners was horrific. At the time of their trials for murder the old provocation defence required that if someone wished to argue that a killing was provoked there had to be evidence of a sudden and temporary loss of control, by the killer, to establish the defence. However, many abused women kill their abusers in circumstances where there has been a cooling off period; and their victim is vulnerable.

Ahluwalia killed her abusive husband of many years by throwing petrol over him and setting light to him when he was asleep. It was the fact that she waited until he was asleep which posed a problem to her ability to plead provocation to the charge of murder. This led to much criticism of the partial defence. Ahluwalia could not claim the defence at her trial precisely because she did not react immediately to her husband’s threatening actions. This was despite the fact that she was in fear of further abusive, possibly life threatening, behaviour in the morning when her husband awoke. Much academic argument ensued and it became accepted that the requirement of a sudden loss of control was sexually stereotypical. That is that men were more likely to react to challenging circumstances by sudden violent losses of control than women. It was argued that women’s reactions to provoking events, particularly where they had suffered abuse, were more accurately described as “slow burn” than a sudden loss of control. This led Susan Edwards to comment on the old defence: “[t]he standard human subject [in provocation] is inexorably male.” [7]

In English law, under normal circumstances, waiting to launch a murderous attack when someone is highly vulnerable would be an aggravating circumstance. The problem for women who killed their abuser was that the killings were intentional; and could often be portrayed by the prosecution as premeditated. However, as acknowledged by the Law Commission, the issue of domestic abuse was central to the reform of the English Law. [8] One of the central issues therefore for the reform of the defence became how this driving need to escape the abuser might be established by evidence in court. The loss of control defence was the result of much consultation between interested parties and the Ministry of Justice. What emerged from this consultation, was a partial defence that recognised fear as a triggering condition that might reduce the blameworthiness of intentional killing.

Thus the new law attempted to recognise the real problem of abusive relationships. However, in offering a partial defence to those who kill their abuser it was argued that the law must avoid offering an excuse to those who kill their partner out of anger through possessiveness or jealousy. The question posed in this article is: is it really possible to separate out emotional triggers to violent action in this manner? Before moving on to look at case law from England, Germany and Australia it would be a good idea to analyse how the partial defence of loss of control operates at present in England.

1. Loss of Control

From October 2010, following the enactment of the Coroners and Justice Act 2009, the partial defence of provocation was repealed. In its place the new partial defence of loss of control was introduced. The new defence is highly structured and requires that where someone charged with murder asserts a defence of loss of control they meet detailed requirements. [9] These requirements are that the killing results from D’s loss of control. [10] The loss of control has to be linked to the following designated qualifying triggers: fear of serious violence against D or an “identified person”, or alternatively the loss of control results from a thing said or done which “constitutes circumstances of an extremely grave character” and causes D to have “a justifiable sense of being seriously wronged.” [11] Where the victim’s reaction in doing or saying something or in creating fear in D results from D’s own actions in inciting that behaviour then the defence is denied. [12] A further restraint is imposed on the defence, the jury must determine whether “a person of D’s sex and age, with a normal degree of tolerance and self-restraint and in the circumstances of D might have reacted in the same or a similar way to D.” [13] Finally, sexual jealousy is specifically excluded as a qualifying trigger. [14] The new partial defence removes the requirement for the loss of control to be sudden. [15] However, it does require that the killing was not a considered act of revenge. [16]

This creates profound difficulties in terms of evidencing loss of control but was seen as a more equitable means of providing a defence when women reacted to abuse. Women’s “slow burn” was recognised by the Law Commission as a major problem with the provocation defence. [17] The defence was also seen as remedying the possibility of men, who through jealousy or possessiveness killed an unfaithful spouse, benefitting from the reformed defence.

2. The English Case Law

Since the new partial defence was introduced there has not been an appeal case reported in the law reports where a woman, who was subject to abuse from a partner, has pleaded loss of control to murder. This simply reflects the fact that no appeal reports exist in the legal databases. It does not mean that there have not been successful pleas of loss of control by women. Indeed, if the plea was successful at first instance, there would be no need for the accused to appeal. There have been appeals by men who sought to clarify the application of the partial defence. The meaning of loss of control was further considered by the Court of Appeal in R v Jewell. Lady Justice Rafferty stated that loss of control entailed: “a loss of the ability to act in accordance with considered judgment or a loss of normal powers of reasoning.” [18]

However, from searching the legal databases and the news media, [19] it is clear that certain of the issues underlying the reform of the law are still causing problems. The most problematic issue being that men who kill their partners because their partner is unfaithful continue to receive sympathetic treatment from the courts. This has to be a matter of concern because the statistics demonstrate that in 2012/13 53% of women, aged 16 or over, who were victims of homicide, were killed by their partner or ex-partner. This contrasts starkly with the figures for male homicide where only 4% of men over 16 were killed by their partner or ex-partner. [20] Professor Jeremy Horder, who was the Criminal Law Commissioner who oversaw the discussion of the reform of the provocation defence, reviewed the new law in an interview with the BBC. He suggested that some of the problems with the provocation defence have transferred to the loss of control defence. The evidence is that these men continue to receive sympathetic treatment from the courts with regard to sentencing. Horder commented that he did not consider: “that the judges are really following through in the way they should on what parliament was saying in 2009”. [21]

But nonetheless Horder has argued that the problem is one of taking a balanced approach to interpreting the new law. Three cases [22] heard at the same time by the Court of Criminal Appeal illustrate the problem that the courts have with considering the relevance of male jealousy and possessiveness in the case of murder. Giving the judgment Judge LCJ, accepted that sexual infidelity could never form a qualifying trigger for the defence. However, he stressed that emotions such as jealousy, whilst excluded as a trigger, might be relevant to assessing the gravity of the action done or words spoken by the victim to the killer. Judge LCJ argued:

“to seek to compartmentalise sexual infidelity and exclude it when it is integral to the facts as a whole is not only much more difficult, but is unrealistic and carries with it the potential for injustice. In the examples we have given earlier in this judgment, we do not see how any sensible evaluation of the gravity of the circumstances or their impact on the defendant could be made if the jury, having, in accordance with the legislation, heard the evidence, were then to be directed to excise from their evaluation of the qualifying trigger the matters said to constitute sexual infidelity, and to put them into distinct compartments to be disregarded. In our judgment, where sexual infidelity is integral to and forms an essential part of the context in which to make a just evaluation whether a qualifying trigger properly falls within the ambit of subsections 55(3) and (4), the prohibition in section 55(6)(c) does not operate to exclude it.” [23]

Commenting on the judgment, Horder suggested that in a sense the reasoning in the case was correct: “this arguably strikes the right balance between restricting the availability of the defence and ensuring that the court can take into account all relevant circumstances.” [24]

The really problematic issue for a court is how to separate out reactions to circumstances which engender an emotional response and result in a loss of control. The difficulty for the courts arises from trying to distinguish those who murder when they have lost control as a result of anger or from fear of serious violence. To gain further insight it is worth considering the facts of some of the cases where women kill having suffered considerable abuse from their partners. England is not the only jurisdiction that struggles in this regard as the case law of Germany and Australia demonstrates.

Abused women who feel threatened by their partner often wait to attack their partner at a vulnerable moment, sometimes planning the attack, a long time in advance. Thus the defence of self-defence rarely arises. The killing is not an immediate and reasonable response to violence or abuse, because it occurs sometime after the triggering event. In England the use of force in self-defence to protect yourself or another individual from attack, if successfully pleaded, would render the killing lawful and result in acquittal. However, to succeed there are qualifications on the use of self-defence. These require that the force used be reasonable in the circumstances and the response proportionate in the circumstances as B perceived them to the threat. If the threat is not imminent the defence is unlikely to be arguable. [25]

3. The German Haustyrannenfall

In 2003 the German Federal Court of Justice (Bundesgerichtshof, BGH) heard the case of a woman who had killed her abusive partner whilst he was asleep. German law does not have a concept of partial defences to murder. However, there are provisions which differentiate levels of responsibility for homicide and mitigate sentence. The case became known as the “house tyrant” case (“Haustyrannenfall”). [26] The defendant (A) had been abused and terrorized by her husband (B) on numerous occasions over many years. [27] On the night A killed him, B had beaten her repeatedly in a very brutal manner. He then went to bed. When cleaning up after the attack, A found B’s loaded revolver. A said, when giving evidence, that she believed that she had to kill B. She could see no other alternative to escape from further attacks and to protect her children. A said she was in a completely hopeless situation and that separating from him and seeking shelter would be pointless. B had told her on previous occasions that he would find her if she left him and take his revenge. While B slept, A fired eight shots at B, at close range, two of which hit him.

The Hechingen District Court heard the case. Its verdict was that A had taken advantage of the fact that B was asleep. This was therefore an aggravating feature of the crime. A was convicted of murder. [28] She appealed to the BGH on the grounds that the District Court had erred in its application of the law.

3.1 The Approach of the BGH

The BGH confirmed that the defence of self-defence (§ 32 StGB) was not available to A as there was no imminent danger. [29] However, the BGH stated that the Hechingen District Court had failed to consider whether A could claim the excusatory defence provided by § 35 StGB (duress) [30]. If this defence succeeded A would be excused criminal responsibility in accordance with § 35 para. 1 sec. 1 StGB which states:

“[a] person faced with an imminent danger to life, limb or freedom which cannot otherwise be averted, commits an unlawful act to avert the danger from himself, a relative or a person close to him, acts without guilt.” [31]

If an imminent danger can be identified, the next issue to be considered is whether the danger could have been avoided. The BGH thought it relevant that A did not try to call the national agencies, ask for help from charitable institutions or try to move into a women’s shelter with her children to escape the abuse. The court concluded that this type of threat can usually be averted through other means than killing. For this reason, the BGH referred to § 35 para. 2 StGB. [32]

The focus of this subsection is on A’s perception of danger at the time of the killing. No punishment is imposed, even in cases where the danger could have been otherwise averted, provided one of two circumstances may be said to apply. The defendant, in error, believes in the presence of an imminent danger, or that she believes there was no less harmful means to avoid the danger. [33] The court will then consider whether the error of judgment was unavoidable. If the error is thought to be “avoidable”, then it remains in the court’s power to mitigate sentence according to § 49 para. 1 StGB.

When considering whether the defendant acted in error, the BGH pointed to the fact that A had given evidence that she believed that she was in a completely hopeless situation; and could only protect herself and her children against future attacks by killing B. Therefore, it concluded that A had erroneously believed that killing was the only way out. The further question remained was the mistake avoidable? This was the question which was referred back to the lower court for consideration.

The Hechingen District Court was asked to consider whether A had sought ways to escape the threat. The lower court was asked to bear in mind the seriousness of the offence and the circumstances of the defendant. In particular, the time available for reflection. This should focus upon whether A was able to think calmly before killing B. Finally, the lower court was instructed to consider the reasons why the defendant felt compelled to act as she did.

The BGH in its decision said that if the defendant had a long period in which to reflect before acting, perhaps being able to seek advice, then the assumption should be that the act was avoidable. The BGH said that it was unlikely that A’s physical and mental condition following the long-term abuse and humiliation prevented her from realistically assessing her options. The Hechingen District Court ruled that the threat could have been averted and sentenced A to imprisonment for 4 years and 6 months. [34]

3.2 Legal analysis of the BGH’s decision

Legal commentators have pointed out that at the time of B’s killing it was unlikely that calling the police and moving into a women’s shelter would have been an effective means to escape from the violence. [35] This effectively rebuts the argument made by the Hechingen District Court that the threat could have been averted. A took the action she did in 2002; evidence from that time suggests woman were at their most vulnerable to attack by their violent partner when they were in the process of leaving or they had already left their partner. [36] Arguably taking the decision to leave placed victims of domestic violence in an even more dangerous situation. [37] A’s liability rested on the court’s view of whether she mistakenly assumed that circumstances existed which would have excused her, that is whether her forming a mistaken belief in the need to kill B could have been avoided. The prevailing view in German legal literature – which is not shared by most legal commentators in Anglo American jurisdictions – suggests that victims of long-term abuse have normally a lot of time to think about and evaluate their options. [38] Thus it is unsurprising that A’s actions were not deemed excusable. The Haustyrannenfall remains the leading authority in this area in Germany, even though it is now some years since the arguments were heard.

4. The Queensland case of R v Falls [39]

The German case law, as with the old English case law prior to the introduction of the loss of control defence, focussed on the circumstances in which the abused woman found herself and the immediacy of her reaction. In English law the old provocation defence required a sudden and temporary loss of control. The consideration of whether the killing might be viewed as avoidable, in German Law was objective. Both jurisdictions approached the evidence in relation to killings by abused women in a way which deprived them of the defences because, arguably, it ignored the psychological effect of long term violent abuse.

The assessment made by the German court was that there were safe ways to escape from the abusive relationship even where the evidence showed otherwise. [40] Another means of responding to the legal issues caused by killings resulting from abuse in domestic relationships is to create a specific defence. A partial defence to murder that only applies when someone commits a homicide in circumstances where they are protecting themselves from further harm. Scenarios covered by this type of defence would include cases not covered by the normal laws of self-defence. Recognising that the abused person may feel trapped by the abusive circumstances, or that possibly the use of force by the abused person would in normal circumstances be viewed as excessive. [41] Such reforms are not gender specific, but evidence from the Australian state of Victoria suggests reform of this type may impact on the decision to prosecute or not prosecute women who kill abusive partners. [42]

Perhaps the easiest manner in which a jury may assess the actions of someone who kills after serious abuse is to evaluate the quality of that abuse and to receive expert advice to help them to understand what effect the experience, of that specific type of abuse, has on the ability of an individual to evaluate the need for their violent actions. The case of Susan Falls who was prosecuted by the state authorities in Queensland, Australia is interesting in this respect. It is a case where the jury heard evidence of the effect of abuse on an individual.

Susan Falls had killed her husband, Rodney, following years of well documented abuse. Michelle Edgely and Elena Marchetti detail the abuse that she had suffered at the hands of her husband, by whom she had four children. [43] At trial evidence was presented that from the start of her marriage Susan was controlled by Rodney, forbidden from visiting her parents, subject to abuse and that Rodney raped her. She feared violence if she spoke to anyone. Rodney used his gun to shoot at her parent’s house and at animals. She left him, but when he promised to reform she returned. However, his behaviour worsened. [44]

Giving testimony Susan described her state as one of helplessness:

“He promised me he was going to change, promised me that he was going to try and make things better. I felt totally helpless then because I, when I went back to him he got to me, I had a DVO on him and I had to cancel or withdraw it, so when the police came around to do that I got the feeling that they looked upon me as some sort of fool … that they wouldn’t take me seriously if I were to call them again”. [45]

At trial Susan said that in the weeks before the killing the violence towards her had worsened and Rodney had threated to kill one of their children. She killed him with a gun she had obtained two weeks prior to the killing after having laced his evening meal with sedatives. The law in Queensland provided a specific partial defence to murder where the killing resulted from an abusive domestic relationship. The defence is set out in s304 B of the Queensland Criminal Code. This specifies three things to be established to prove the defence. In short form these are: firstly, evidence that the deceased had committed acts of serious domestic violence, secondly, the defendant believed that the action was necessary for self-preservation; and finally, that the defendant has reasonable grounds for believing in the need for self-preservation in the circumstances. There are some echoes of the German defence here but it is the presentation of the argument by the defence which is of interest.

The prosecution in Susan Falls’ case tried to prevent the admission of evidence from two experts. This evidence was to assist the jury concerning the issue of reasonable grounds for belief in the need for self-preservation in the circumstances experienced by Susan Falls. Justice Applegarth refused this application relying on the explanatory notes to the revised Queensland Criminal Code which suggested that evidence could be taken from experts “regarding the reasonableness of the belief”. [46]

In this case the expert evidence regarding the new partial defence focussed on Susan Falls’ behaviour in the circumstances in which she found herself. It was her appreciation of the risk that was central. In such circumstances the experience of the person who is horrifically abused is exceptional and beyond the ordinary experience of a jury. Justice Applegarth ruled therefore that an expert could be called to explain to a jury the likely effect on the defendant of the abuse. Justice Applegarth advised the jury on both the new partial defence and on self-defence. The jury acquitted Falls of murder on the grounds of self-defence.

This outcome is unusual. Elizabeth Sheehy et al, argue that it resulted in part from the manner in which the evidence regarding the assessment of risk was presented to the jury. The defence counsel in Falls argued that for her defence to succeed there did not have to be an imminent attack: “It might have been the next day, it might have been the next week, but the risk of death or serious injury to her was ever present.” [47] Commenting on this argument Sheehy writes: “In this instance, the manner in which the legal requirements of self-defence were presented to the jury permitted the threat that the accused was facing to be appraised not just in the immediate circumstances of the killing but in terms of the history of violence.” [48] The defence argued context was all and that the abuse and its effects had taken place over a long timescale. The distinction was made between the reasoning process of an ordinary person and the reasoning process of someone who had suffered the level of abuse experienced by Susan Falls. [49] Additionally in his summing up to the jury, Justice Applegarth explained the circumstance of an abused person stating that those in abused relationships “often believed that they have no alternative means of self preservation other than to kill their abusers”, he also said:

“The fact that the Parliament has enacted this partial defence for victims of abuse in an abusive relationship does not mean that other defences such as self-defence are not available for people who have had an abusive relationship.” [50]

But notably it is the nature of the threat that is being evaluated and the reaction to it, not the emotion of fear of serious violence.

5. The need for expert evidence

The focus of the rest of this article will be on the defences of self-defence and loss of control in England. The problems experienced in the German and Queensland courts in dealing with the helplessness felt by abused women have echoes in the cases which led to pressure for the reform of the English Law. In Ahluwalia evidence was advanced on appeal to substantiate the effect of the abuse:

“The state of humiliation and loss of self-esteem to which the deceased's behaviour over the ten years of the marriage had reduced her is evidenced by a letter she wrote him after he left her for three days about April 1989. It is a letter on which Mr Robertson QC strongly relies. In the course of begging him to come back to her and to grant her ten minutes to talk it over, she made a number of self-denying promises of the most abject kind:

‘Deepak, if you come back I promise you--I won't touch black coffee again, I won't go town every week, I won't eat green chilli, I ready to leave Chandikah and all my friends, I won't go near Der Goodie Mohan's house again, Even [sic] I am not going to attend Bully's wedding, I eat too much or all the time so I can get fat, I won't laugh if you don't like, I won't dye my hair even, I don't go to my neighbour's house, I won't ask you for any help.’” [51]

It would be difficult for the average juror to be able to understand the psychological effect that abuse may have on the abused person. This is also likely to be true of extreme fear reactions. In the present partial defence of loss of control in English law one of the triggering conditions required to establish loss of control – fear of serious violence – may raise issues that are outside the experience of the ordinary juror. Fear reactions are not commonly experienced and for a juror distinguishing a fear reaction, in the circumstances of severe domestic abuse from an angry reaction may pose real difficulty. Jurors may never have experienced severe abuse or that level of fear. There is also the issue that most people committing a violent crime will experience some sort of a reaction which could be viewed as fear related. Even if it is simply a fear of being caught or suffering violence in return for the violence perpetrated against the victim. In these circumstances the use of fear of serious violence as a trigger is a fairly nebulous, hard to define term and the jury may well need expert assistance in appreciating the circumstances of the accused.

Cognitive neuroscience provides an understanding of emotional behaviour and of the psychological drivers of behaviour. In the next section of the article we will briefly examine what neuroscience contributes to our understanding of emotional reactions and how it may assist the law in the future. It does seem that the use of expert evidence in the Queensland case assisted the deliberations of the jury. Law and neuroscience view explanations of behaviour from different perspectives. None the less it is useful to look at the basis of legal excusing conditions from a number of perspectives. Whilst expert evidence will not always be required to make the defence argument it will be required in difficult cases where the experience of the perpetrator falls outside the experience of the layperson. Where behavioural evidence is called to explain the circumstances of the accused it is likely to be based on the latest scientific understanding of cognition and human behaviour.

6. The consequences of fear

One of the first things to state about cognitive neuroscience is that the descriptions given to experiences differ from those explanations of fear reactions that would normally be heard in court by a jury. This is because cognitive neuroscience would separate the conscious experience of fear from the subconscious experience of fear. That fear has a basis in the structure of the brain is relatively undisputed. The brain is seen by neuroscientists as a complex series of interconnected networks. Functionally, it is possible to identify regions of the brain as hubs central to certain networks. In the case of fear, the amygdala is identified as a hub through which our experience of fear is mediated: “fear in the brain depends on the amygdala. If this area is removed a person becomes fearless, but it is the amygdala’s connections and extended network that allow us to really feel fear.” [52]

Looking at the loss of control defence, it is not just the experience of fear that is required to establish a defence but rather its effect. In R v Jewell it was said that to establish a loss of control related to a relevant triggering condition the defendant had to establish: “a loss of the ability to act in accordance with considered judgment or a loss of normal powers of reasoning.” This has some correlation with the Queensland requirement that there be an assessment of the reasonableness of Susan Fall’s belief, in the circumstances in which she found herself, that her actions were necessary. But the English law stresses the overpowering nature of the reaction to the circumstances. This is then filtered through the prism of the reaction of a person of D’s age and sex, with a normal degree of tolerance and self-restraint. [53] It seems that the law is looking for a plausible explanation of behaviour in these circumstances. The framework the English law creates is complex and diffuse evaluating the emotion fear against other ideas of normal tolerance. It is clear that before excusing a defendant all three legal systems are seeking to establish that the behaviour stems from some exceptional occurrence, that the circumstances are grave and the defendant’s response in the circumstances is not unreasonable or otherwise inexcusable. This requires the hearing of evidence from the defendant and witnesses about the context in which the violent act took place.

However, returning to the English law, given that our fear reactions are both conscious and subconscious is this really possible? In a case where there is considerable objective evidence of abuse or the circumstances surrounding the violent act are well documented, by CCTV footage or independent witnesses, then it may be possible to piece together an objective explanation. However, where the fear reaction is more instinctive this may be problematic, and highly problematic where there is no objective evidence of how the event came about. For example where the killer simply asserts that they were in fear of serious violence and there is some circumstantial evidence that this may be the case.

Antonio Damasio examining how emotional responses are triggered writes: “It is almost as if certain stimuli have the right key to open a certain lock, … This is the case of fear-causing stimuli, which often activate the amygdalae and succeed in triggering the fear cascade.” [54] Damasio goes on to describe what happens as “disturbing” and says it “amounts to an upset of the ongoing life state at multiple levels of the organism”. [55] When a strong fear reaction occurs, that is when people perceive a threat then one of two responses are brought about these are: “Depending on the context in which the fear-causing images appear, one may then freeze in place or run away from the source of danger.” [56] For Damasio therefore the emotion is triggered by the stimuli which are activated when we perceive a threat. He also points out that the ability to control that reaction is limited: “The mechanism is so exquisite that yet another structure, the cerebellum, will struggle to moderate the expression of fear.” [57]

Joseph LeDoux is a cognitive neuroscientist who studies reactions to threats at the subconscious and conscious level. [58] As with most functional explanations of behaviour the problem becomes the marrying of a behaviourist explanation with the human subjective experience of the feelings associated with fear. His work suggests that there may be two types of fear response, one of which we are consciously aware and another which exists at a non-conscious threat processing level. LeDoux writes: “Conscious fear can cause us to act in certain ways”. However, in conditioned fear responses of the type described by Damasio consciousness of fear was not necessary to generate the defensive behaviours: “it is not the cause of the expression of defensive behaviors and physiological responses elicited by conditioned or unconditioned threats.” [59] This type of conditioned fear reaction is brought about by events and does not require a conscious thought process. It could be described as an instinctive reaction.

It is when the defendant recounts a sense of fear linked to the killing that the law will need to interpret the relationship between the accused’s fear of serious violence and the loss of control. This is because the English law requires that the fear trigger must be assessed by the jury who have to consider whether “a person of D’s sex and age, with a normal degree of tolerance and self-restraint and in the circumstances of D, might have reacted in the same or a similar way to D.” [60] The nature of the fear is also highly specific being related to a threat of “serious violence against D or an identified person.” [61] This description of the conditions which may partially excuse murder do not limit the excuse to a conditioned fear reaction but will also encompass reactions where the explanation of the feeling of fear is a subjective emotional experience.

LeDoux confirms that the problem for the law will be most acute when the only evidence of how a violent act came about through a fear response is subjective testimony. He writes: “Conscious mental states should not, in the absence of direct evidence, be the first choice explanation of behaviour, even in humans.” [62] Thus in the context of laypeople’s understandings of fear reactions, LeDoux strongly warns against assuming that there is a sharing of assumptions between scientists and laypeople: “Experiences that we label and talk about as fear are not directly tied to the circuits that detect and respond to threats.” [63] Thus according to LeDoux there are two types of experiences that we would describe as fear responses but one of them occurs at the non-conscious level. In humans LeDoux suggests that “[s]ome fears depend upon survival circuits but others do not”. [64] Fear experiences will differ from person to person: “No one ingredient is essential to fear. Variation in the kind and amount of ingredients determine whether you feel fear, as opposed to some other emotion, and also determine the variant of fear that you feel.” [65]

On this view fear is a subjective experience which will vary in intensity together with the nature of that subjective experience; and the underlying presence or absence of non-conscious factors affecting what LeDoux calls the survival circuits. What laypeople describe as fear would not necessarily be coterminous with a scientific description of a fear reaction. The absence of any means of providing objective evidence of fear is concerning. How is the jury to discern the difference between fear and anger? The conclusion from the neuroscience is that what we as human actors describe as fear reactions are subjective descriptions of emotional reactions to events. The Royal College of Psychiatrists in their response to the consultation on how to reform the criminal law partial defence of provocation made the difficulty very clear:

“Physiologically anger and fear are virtually identical, whilst many mental states that accompany killing also incorporate physiologically both anger and fear. Hence, the abused woman who kills in response even to an immediate severe threat will also be driven at least partly by anger at the years of abuse meted out to her, and perhaps her children. Again, the woman who waits until the man is ‘helpless’ to kill him, is likely not to be merely angry but also fearful that eventually he will kill her, and/or her children, and that there is no way of preventing it other than by the death of the man (partly because her cognitions have been so distorted by the years of abuse that she does not perceive the options for escape, for example legal options, at all in the same way as an ordinary person would do). Any legal solution to the current perceived problems with the partial defences to murder which rested upon the assumption that fear and anger can (even usually) be reliably distinguished must, from a medical perspective, therefore fail.” [66]

This response suggests that the evaluation of this type of trigger by the jury will be both outside of ordinary experience and not easily amenable to explanation by an expert. The Law Commission advocated that fear be included in the loss of control defence in English law in a desire to do justice to two types of defendant. Firstly, the householder who kills an intruder using excessive force through fear; and secondly the abused person who kills as a result of feeling both seriously threatened by and unable to escape the abuser. [67] But the opinion of experts is that fear and anger are emotions that are almost indistinguishable.


What the neuroscience and medical knowledge suggests is that there is no scientifically objective measure for the effective separation of anger and fear. Lack of clarity here is a matter of concern. Unfortunately, the emotional experience of fear is not sufficiently separable or distinguishable as an emotion, from anger states. The risk therefore is that the juror trying to put themselves in the circumstances of the accused may actually fail to appreciate precisely how the presence or absence of fear is relevant to the legal argument regarding the partial defence of loss of control. Such lack of clarity is regrettable as one of the aims of the legislative change was to recognise the problem created by domestic violence. If seriously abused people, who kill their abuser are to receive fair treatment, then the partial defence should be capable of easy understanding and application.

The lack of a strong means of distinguishing fear from anger has consequences which are disturbing. If no definitive evidence can be called to establish fear and distinguish it from outbursts of anger, then the use of the term as a triggering condition to partially excuse a violent crime is worrying. Also as the conscious experience of fear is subjective and its effects vary widely then its use as a tool to evaluate excusing conditions is indeed problematic.

Neurocognitive and medical knowledge would suggest that identifying fear as a constituent trigger may pose problems for juries, as they will be unlikely to have experience of the type of circumstances in which defendants find themselves at the time of the killing. LeDoux’s research also suggests that the experience of fear will vary depending on the ingredients which inspire the emotional state. This could cause real difficulty for juries. One of the keys to the creation of successful partial defences to murder is to make clear the component parts of a defence. Identifying clearly what may be the excusing conditions that may be subject to explanation through scientific expert evidence. Clarity in the component parts of the defence are key to identifying what is appropriate in terms of the admission of expert evidence. Arguably the English loss of control defence obscures rather than clarifies the conditions that may partially excuse murder, resting as it does, in part, on a fear of serious violence. This is regrettable.

[1] Dr Lisa Claydon, Senior lecturer in law, The Open University Law School, and Caroline Rödiger, Research Fellow, The University of Manchester, School of Law. Work on this article was supported by funding from the Arts and Humanities Research Council as part of the project: Sense of agency and responsibility: integrating legal and neurocognitive accounts.

[2] In the rest of this article references to English law, or to the law of England should be read as including Welsh law and the law of Wales.

[3] Celia Wells, Provocation the case for abolition in Andrew Ashworth, and Barry Mitchell eds Rethinking English Homicide Law (OUP 2000) p85.

[4] R v Clinton; R v Parker; R v Evans [2012] EWCA Crim 2, [2013] QB 1. See discussion of this issue n22 and surrounding text.

[5] R v Humphreys [1995] 4 All ER 1008 (CA).

[6] R v Ahluwalia [1992] 4 All ER 889 (CA).

[7] Susan Edwards, “Justice Devlin’s legacy: Duffy a battered woman “caught” in time”, [2009] Crim LR 851-869, 853.

[8] Law Commission: Partial Defences to Murder (Law Com No 290, 2004) [3.75].

[9] S54 and 55 of the Coroners and Justice Act 2009.

[10] Ibid s54(a).

[11] Ibid s55(4).

[12] Ibid s55(6).

[13] Ibid s54(c).

[14] Ibid s55(6)(c).

[15] Ibid s54(2).

[16] Ibid s54(4).

[17] Law Commission: Partial Defences to Murder, Law Com No 290 Para 3.20.

[18] [2014] EWCA 414 [24].

[19] Jeremy Horder BBC News Magazine (2014) http://www.bbc.co.uk/news/magazine-29612916 accessed 16/02/2016.

[20] ONS Crime Statistics, Focus on Violent Crime and Sexual Offences, 2012/13. http://www.ons.gov.uk/ons/rel/crime-stats/crime-statistics/focus-on-violent-crime-and-sexual-offences--2012-13/index.html accessed 16/02/2016.

[21] Jeremy Horder BBC News Magazine (2014) http://www.bbc.co.uk/news/magazine-29612916 accessed 16/02/2016.

[22] R v Clinton; R v Parker; R v Evans [2012] EWCA Crim 2.

[23] Ibid [39].

[24] Jeremy Horder, "Loss of Control an acceptable defence?" Keep Calm and Talk Law, (2014) http://www.keepcalmtalklaw.co.uk/loss-of-control-an-acceptable-defence/ accessed 17/02/2016.

[25] Self-defence exists at common law and is also a statutory defence set out in s3 of the Criminal Law Act 1967, which applies where D uses force to prevent a crime. The Criminal Justice and Immigration Act 2008, s76 outlines the conditions for the operation of the defence. For a sophisticated discussion of how the defence operates see David Ormerod and Karl Laird, Smith and Hogan’s Criminal Law, (14th edn, OUP, 2015) 429.

[26] BGH [2003] 1 StR 483/02, BGHSt 48, 255-263.

[27] His abuse included slapping her face, spitting at her, kicking her in her stomach, causing her serious injury. A had sought refuge in a women’s shelter but returned to their shared home after four weeks after B promised to improve his behaviour. However, the acts of violence increased in terms of the number and intensity. B started to attack their children, too.

[28] § 211 (killing by stealth) of the German Criminal Act (Strafgesetzbuch, StGB).

[29] See Michael Bohlander, ‘When the Bough Breaks - Defences and Sentencing Options Available in Battered Women and Similar Scenarios under German Criminal Law’ in Alan Reed and Michael Bohlander (eds), Loss of Control and Diminished Responsibility. Domestic, Comparative and International Perspectives (Ashgate 2011).

[30] The meaning of duress in German law is different from the meaning of duress in English law, it has similarities with self-defence in English law.

[31] http://www.gesetze-im-internet.de/englisch_stgb/englisch_stgb.html#p0174 accessed 22/02/2016. The criteria of § 35 para. 1 sec. 1 StGB of particular relevance to the BGH were thus: first, whether there was an “imminent danger” and second, if that danger could have been “otherwise averted”. The meaning of imminent danger includes a permanent state of threat. Thus A because of the evidence of long-term abuse could be deemed to meet the first criteria.

[32] http://www.gesetze-im-internet.de/englisch_stgb/englisch_stgb.html#p0174 accessed 22/02/2016. This states that: “[i]f, at the time of the commission of the act, a person mistakenly assumes that circumstances exist which would excuse him under subsection (1) above, he will only be liable if the mistake was avoidable. The sentence shall then be mitigated pursuant to sec 49(1).”

[33] BGH [2003] 1 StR 483/02, BGHSt 48, 261.

[34] Unfortunately, the judgment is unpublished, further details including how the court reached its decision in respect of sentencing is unknown. For that reason, the following legal analysis relies on the judgment of the BGH.

[35] Edward Schramm, Ehe und Familie im Strafrecht. Eine strafrechtsdogmatische Untersuchung (Mohr Siebeck 2011) 155.

[36] Joachim Burgheim, Psychologische Bedingungen bei Entstehung und Verlauf von Tötungsdelikten in der Situation der Partnertrennung (Hartung Gorre 1993) 52.

[37] Since the introduction of the GewSchG in 2002 Germany has made a great effort to provide enhanced protection for women who are abused by their partners. The act applies, inter alia, in all cases of unlawful physical assault, harassment and stalking (§ 1 para. 2 and 3 GewSchG) and provides protection orders (§ 1 para. 1 GewSchG) as well as measures for the relinquishment of the jointly used dwelling (§ 2 GewSchG). The underlying principle of the GewSchG is “the aggressor goes, the victim stays”. The protection order takes immediate effect and bans the suspected abuser from the victim's home even if it is the abuser’s home.

[38] Edward Schramm, Ehe und Familie im Strafrecht. Eine strafrechtsdogmatische Untersuchung (Mohr Siebeck 2011) 160.

[39] Unreported Supreme Court of Queensland, Applegarth J. 3 June 2010 I am indebted to Justice Applegarth for bringing this case to my attention.

[40] See n35 and accompanying text.

[41] For a lengthier discussion and analysis of the introduction of such partial defences in the State of Victoria see Heather Douglas, “A consideration of the merits of specialised homicide offences and defences for battered women”, Australian and New Zealand Journal of Criminology (2012) 45(3), 367-382.

[42] See Danielle Tyson, “Victoria’s new homicide laws: provocative reforms or more stories of woman ‘asking for it’” Current Issues in Criminal Justice (2011) 23(1) 203-233, 211.

[43] “Women who kill their abusers: How Queensland’s new abusive domestic relationships defence continues to ignore reality” (2011) 13 Flinders L.J. 125-176, 141-5.

[44] Ibid 143-4.

[45] Examination in Chief of Susan Falls, Transcript of Proceedings, R v Susan Falls, Bradley James Coupe, Christopher Anthony Cummings-Creed and Anthony James Hoare (Supreme Court of Queensland, Applegarth J., 26 May 2010 day 9-19,20).

[46] Douglas (2012) 376.

[47] Elizabeth Sheehy, Julie Stubbs and Julie Tolmie, “Securing fair outcomes for battered women charged with homicide: analysing defence lawyering in R v Falls” (2014) 38 Melb U.L. Rev. 666-708, 676.

[48] Sheehy et al (2014) 677.

[49] Ibid 678.

[50] Marchetti et al (2011) 148 both quotations.

[51] [1992] 4 All ER 889, 893.

[52] Senna A. The 30 – Second Brain, (2014, Icon Books: London), 36.

[53] S54(1) (c) Coroners and Justice Act 2009 this is subject to the further explanation in 54(3) “the circumstances of D is a reference to all of D’s circumstances other than those whose only relevance is they bear on D’s general capacity for tolerance and self restraint.” - emphasis in original.

[54] Antonio Damasio, Self Comes to Mind Constructing the Conscious Brain (2010, Heinemann) 112.

[55] Ibid p113.

[56] Ibid p115.

[57] Ibid pp113-4.

[58] Joseph LeDoux, “Coming to Terms with Fear”, Proceedings of the National Academy of Sciences of the United States of America, (2014) vol 111 no 8 2871-78.

[59] LeDoux (2014) 2873 both quotations.

[60] Coroners and Justice Act 2009 s54(c).

[61] See n11 and accompanying text.

[62] LeDoux (2014) 2874.

[63] Ibid 2874.

[64] Ibid 2876 both quotations.

[65] Ibid 2876.

[66] Law Commission: Partial Defences to Murder (Law Com No 290, 2004) [3.99].

[67] Ibid [4.17–4.31].