Self‐manslaughter and the forensic classification of self‐inflicted death

Abstract

By emphasising the intentions underlying suicidal behaviour, suicidal death is distinguished from accidental death in standard philosophical accounts on the nature of suicide. A crucial third class of self‐produced deaths, deaths in which agents act neither intentionally nor accidentally to produce their own deaths, is left out by such accounts. Based on findings from psychiatry, many life‐threatening behaviours, if and when they lead to the agent's death, are suggested to be neither intentional nor accidental, with many apparently suicidal behaviours being of this sort, especially the so‐called “cries for help”. This category may be usefully analogised to the existing legal category of manslaughter.

One of the more difficult challenges of forensic psychiatry is classifying deaths in which the deceased appear to have a role in bringing about their own demise. One complication is that the relationship between a person's dying and that death being a suicide is not straightforwardly causal: there are clear cases where people are causally responsible for their own death, but which are not suicides, such as when a person mistakenly believes a rifle is not loaded and ends up fatally wounding himself or herself. Clear cases of suicide also exist where if people are causally responsible for their death at all, this is so only at one remove—for example, when a person arranges for doctor‐assisted suicide. Consequently, most philosophical analyses of suicide recognise that the crucial factor in ascertaining whether a person's death should count as suicide is not whether the person caused his or her own death but whether it was intended that the actions would cause the death.¹,²,³,⁴,⁵,⁶,⁷ The obvious advantage of the thesis that suicide is intentional self‐killing is that it enables us to distinguish suicide from those deaths in which death is genuinely accidental, in that the agents acted in a situation in which the prospect of their own death played no part. We may say that in the cases of accidental self‐killings, death never enters the mind of the agent who dies.

Yet, this division of self‐produced deaths into suicides or accidents turns out to be too neat. My purpose here is to draw on empirical evidence regarding suicidal ideation to highlight that many suicidal behaviours exhibit features that fall short of intentional self‐killing, but are none the less not genuinely accidental. Such behaviours indicate the need for a third category of self‐killing, analogous to the category of manslaughter. I conclude by exploring the practical and moral implications of acknowledging the existence of such self‐manslaughter.

Ambivalent life‐threatening behaviour

Except in the rarest of cases, suicidal behaviour is a means rather than an end, in that suicidal people seek the state of death not for itself but instead, death has some other justification or aim: the relief of physical pain, the relief of psychological anguish, martyrdom in the service of a moral cause, the fulfilment of perceived societal duties (eg, suttee and seppuku), the avoidance of judicial execution, revenge on others, protection of others' interests or well‐being.⁶ Therefore, suicidal people do not intend death as such, but rather death is perceived, rightly or wrongly, as a means for the fulfilment of any of the agent's aims.

Still, many instances of life‐threatening behaviour are not so clear‐cut with respect to what the agent intends. Firstly, suicidal agents may be ambivalent about death as a means, despite being fully committed to the end it is presumed to achieve. A disgraced soldier can be wholly devoted to the value of acting in accordance with his society's notions of honour, notions that necessitate his taking his life, but be ambivalent about suicide because of his natural fear of death. Similarly, severely addicted drug users may recognise that the dose they are about to take may lead to a fatal overdose, but be willing to take such a risk in pursuit of an ever more intense high. Conversely, agents may believe wholeheartedly that death is the best means to the satisfaction of an end to which they are less than wholeheartedly committed—for example, when a patient with a painful illness acknowledges that death is the best way of bringing the suffering to an end, yet agonises over whether to seek doctor‐assisted suicide because he or she cannot decide whether continued life would in fact be worth living. We should also expect to find both sources of ambivalence combined in some instances. At the least, we cannot easily guage from an agent's engaging in life‐threatening behaviour the depth of commitment either to dying or to the end that dying might serve.

Psychiatrists and psychologists have long recognised that suicidal thought and behaviour is often suffused with precisely these forms of ambivalence.⁸,⁹,¹⁰,¹¹,¹²,¹³,¹⁴,¹⁵,¹⁶,¹⁷,¹⁸,¹⁹ Although such ambivalence is more common in some populations (women,²⁰ children and adolescents²¹) than in others (the elderly),²² uncertainty about whether to commit suicide is widespread. Results from a wide array of psychological screening tests, along with the methods used in life‐threatening behaviour, can be used to measure this ambivalence. Such ambivalent life‐threatening behaviours are sometimes called suicide gestures, although this term has no agreed‐upon meaning. The distinguished suicidologist Edwin Shneidman separates “death seekers”, who clearly intend to end their lives, from “death darers”, whose life‐threatening behaviour occurs against a background of mixed feelings or ambivalence.²³ The most common manifestation of such ambivalence is the parasuicide or cry for help, in which people engage in a behaviour believed to be life threatening, not with the purpose of causing their own death but to express their anguish to others, typically in the hope that those others will provide aid, comfort or rescue.

Neither accidental nor suicidal

No matter what the exact source of the ambivalence, such agents, on engaging in life‐threatening self‐directed behaviour despite this ambivalence, seem not to unconditionally desire their own deaths. None the less, death can and does result from acts with such ambivalent origins. Such deaths, however, do not fall neatly into the categories of suicides or accidents. On one hand, ambivalent, self‐produced deaths do not result from an agent's intending to die. Suppose, for instance, that a person engages in life‐threatening behaviour, issuing the aforementioned cry for help, desperate that others take notice of the person's anguish and respond accordingly. The person earnestly hopes to be saved in the process of engaging in the life‐threatening behaviour (although no precautions are taken to prevent death). If death results, was death intended? Not obviously so, for not only did this agent not want to die but he or she actually had no intention of dying. So, are such deaths accidental? They have the mark of accidental death, in that death was not an intended aim. They, however, lack an essential element of accidental deaths—namely, the absence of the prospect of death from the reasoning the agents perform before their death. In an accidental death, people die as a result of an act, either performed by themselves or by another, which they choose to perform without aiming at dying at all. In fact, accidental deaths require that the agent either not be mindful of the possibility of death (as when a person slips on the ice and has a fatal head injury) or be mindful of its possibility while taking active steps to prevent it. In cries for help, when death results, the agents meet neither of these conditions, because they are consciously intending to undertake a life‐threatening risk, despite not intending to die. To hold that people who, in issuing a cry for help, hang themselves and end up dying do so accidentally is as plausible as holding that a person who dies when playing Russian roulette dies accidentally. In general, unless a person takes precautions to prevent death, whenever a person engages in a behaviour known to be life threatening but does so without intending to die, yet they die none the less, their death is not accidental. Deaths such as these are not a mere coincidence or bad luck, nor are they intended or accidental. I shall hereafter refer to such deaths as neither intentional nor accidental (NINA) self‐killings.

Manslaughter and self‐manslaughter

I would suggest that a significant number of self‐produced deaths fall into the third category, NINA suicide. It might be thought sufficient to allow such self‐inflicted deaths to be counted as examples of self‐inflicted harm, as the World Health Organization's International Classification of Diseases allows.²⁴ This description, however, conceals more than it reveals, and it is simply not true of NINA self‐killings that their causes are in general undeterminable. Instead, diagnostic practice is often constrained by the accepted taxonomies that govern it, and in this instance, those responsible for classifying self‐killings are hampered by too simplistic a classificatory scheme, one that (I shall argue in the next section) is harmful in several ways.

Fortunately, NINA self‐killings can be helpfully analogised to an existing legal category for certain kinds of killings of others—namely, manslaughter. Although legal definitions of manslaughter vary, the central distinction between manslaughter and murder is that in manslaughter, two conditions are absent: premeditation and malice (or intent to harm).

Firstly, malice: malice is ill intent, or intent to cause grievous suffering or harm. In NINA self‐killings, the aim and the actual outcome of the agent's act of self‐killing diverge, in that the agent did not mean to die but dies none the less. In the hierarchy of mens rea proposed by the American Legal Institute in its Model Penal Code (1962, section 2.02(2)), NINA self‐killings fall short of purposeful self‐killing, because people who die do not intend to do so, and indeed do not aim at harming themselves at all. From the agents' perspective, the life‐threatening behaviour that causes death is intended to benefit them, whether in fact, from a more objective point of view, this behaviour is, or turns out to be, a benefit to them. NINA self‐killings are therefore not instances of self‐directed malice.

The other feature absent from manslaughter, premeditation, may seem more problematic, for, in at least some instances, NINA self‐killings are premeditated in so far as premeditation includes a deliberate and persistent preparation before the act. In many cases, NINA self‐killings are impulsive, but many such self‐killings are a result of elaborate planning and preparation. None the less, the agents who die as a result of NINA self‐killing may perform their actions with premeditation, but do not kill themselves with premeditation. Again, because NINA self‐killings are not the product of the intent to kill oneself, they lack the specific intent to kill required for premeditated self‐killing.

NINA self‐killings are therefore self‐directed instances of manslaughter, and do not satisfy the descriptive criteria of suicide. In terms of mens rea, self‐manslaughter is voluntary self‐killing performed either knowingly or recklessly, but not purposefully.

Theoretical, moral and legal implications

What does it matter if a person undertook a life‐threatening behaviour intending to die? In particular, why ought we care if a person dies accidentally, intentionally or in an act of self‐manslaughter? From a consequentialist perspective, is not death death?

From a theoretical perspective, the ambivalence associated with self‐manslaughter is an important insight, if only to understand the motivations of many of those we now misleadingly term suicidal: by coming to understand these ambivalent motives, we improve our chances of successfully intervening in life‐threatening behaviours. Moreover, if reasons of autonomy ever give us reasons to intervene to prevent suicide, the ambivalence of self‐manslaughter gives us an even stronger reason to intervene: a central concern about suicide is whether death is really in the long‐term interests of the people as understood by them. People who engage in self‐manslaughtering behaviour actually do not unwaveringly believe that death is in their interests. Acting to stop self‐manslaughter is therefore more morally justified than is intervention in clearly suicidal life‐threatening behaviour.

Families of people who suffer self‐inflicted deaths would also stand to benefit were self‐manslaughter recognised as a third alternative for classifying self‐inflicted death. Suicide continues to carry a stigma, especially within some systems of religious belief, and many people hold that suicide is morally wrong in at least some circumstances. For example, loved ones who learn that the person's behaviour is better captured by self‐manslaughter than by suicide may lessen their concern that a loved one died in an act of sin. This is not to say that the emotional effects of introducing self‐manslaughter into our classificatory scheme would be entirely benign. Some self‐inflicted deaths that were previously counted as suicides, accidental deaths or of undetermined cause would be classified as self‐manslaughter under the revised scheme, and this could be a source of anguish for some family members and loved ones, in so far as they may experience a greater sense of shame for not having intervened before their loved one's ambivalent life‐threatening behaviour. Nevertheless, I contend that by expanding the range of classifications for self‐inflicted deaths, we enable families to know the truth about their loved one's demise, and that on the whole, knowing the truth, however unpleasant it may be, would have salutary effects on the family's psychological health. Suicide is a painful and enigmatic phenomenon no matter what, but classificatory schemes that force self‐inflicted deaths into the binary categories of suicide or accidental death (or leave hard cases as “undetermined”) do little to lift the veil on the complex intentions and rationales that lie behind suicidal behaviour. Recognising self‐manslaughter as an independent class of self‐inflicted deaths can help family members and loved ones deal with the shame, anger, reticence and self‐deception in more open and emotionally mature ways.

Furthermore, whether a death is counted as suicide has important legal and public health ramifications. Insurance companies do not generally honour the life insurance policies of people who commit suicide, and although it is not clear whether that policy would change were self‐manslaughter to enter our shared vocabulary of self‐killings, it would at least allow family members to make the argument that a loved one's death was not intentional and was therefore not an attempt at fraud. Legal recognition of self‐manslaughter would also improve the allocation of mental healthcare resources by giving a more accurate statistical picture of the prevalence of suicidal ideation and behaviours within various population groups.

Finally, the more practically minded might raise epistemological concerns, to the effect that we would not be able to distinguish self‐manslaughter from accidental deaths and from suicide. The decision to classify a homicide as accident, murder or manslaughter is made somewhat easier by the fact that, except in cases of murder–suicide, the chief witness to the homicide—namely, the person responsible for another's death, can serve as a key source of evidence regarding the circumstances, intentions, etc, surrounding the homicide. This is not so with respect to self‐inflicted deaths. Still, pronouncing whether a self‐inflicted death is accidental, suicide or self‐manslaughter is far from hopeless. Conducting a psychological autopsy to determine a deceased person's motivations is no easy task, as the evidence in question will in many cases not be decisive about how to classify a self‐killing. None the less, the relevant evidence is abundant, including the lethality of the means of death, the content of any notes or documents, psychiatric records, the existence of previous life‐threatening behaviour, evidence of reconsideration or midstream modification of one's plan, the likelihood that the person would be found and stopped, and clues from the person's behaviour in the days before death. Still, giving those investigating a death the option of calling it self‐manslaughter might in fact simplify their task, as it frees them from having to force ambiguous cases into suicides or accidental deaths (or leave the cause of death as undetermined). At the least, permitting self‐inflicted deaths introduces no diagnostic or forensic challenges that are not already present to a great extent in the classification of homicides as accidental, murder or manslaughter.

Abbreviations

NINA - neither intentional nor accidental