Journal of Practical Ethics

The Death Penalty Debate: Four Problems and New Philosophical Perspectives

Masaki Ichinose

The University of Tokyo

Abstract

This paper aims at bringing a new philosophical perspective to the current debate on the death penalty through a discussion of peculiar kinds of uncertainties that surround the death penalty. I focus on laying out the philosophical argument, with the aim of stimulating and restructuring the death penalty debate.

I will begin by describing views about punishment that argue in favour of either retaining the death penalty (‘retentionism’) or abolishing it (‘abolitionism’). I will then argue that we should not ignore the so-called “whom-question”, i.e. “To whom should we justify the system of punishment?” I identify three distinct chronological stages to address this problem, namely, “the Harm Stage”, “the Blame Stage”, and “the Danger Stage”.

I will also identify four problems arising from specific kinds of uncertainties present in current death penalty debates: (1) uncertainty in harm, (2) uncertainty in blame, (3) uncertainty in rights, and (4) uncertainty in causal consequences. In the course of examining these four problems, I will propose an ‘impossibilist’ position towards the death penalty, according to which the notion of the death penalty is inherently contradictory.

Finally, I will suggest that it may be possible to apply this philosophical perspective to the justice system more broadly, in particular to the maximalist approach to restorative justice.

1. To whom should punishment be justified?

What, exactly, are we doing when we justify a system of punishment? The process of justifying something is intrinsically connected with the process of persuading someone to accept it. When we justify a certain belief, our aim is to demonstrate reasonable grounds for people to believe it. Likewise, when we justify a system of taxation, we intend to demonstrate the necessity and fairness of the system to taxpayers.

What, then, are we justifying when we justify a system of punishment? To whom should we provide legitimate reasons for the system? It is easy to understand to whom we justify punishment when that punishment is administered by, for example, charging a fine. In this case, we persuade violators to pay the fine by bringing to their attention the harm that they have caused, harm which needs to be compensated. (Please note that I am only mentioning the primitive basis of the process of justification.) While we often generalise this process to include people in general or society as a whole, the process of justification would not work without convincing the people who are directly concerned (in this case, violators), at least theoretically, that this is a justified punishment, despite their subjective objections or psychological opposition. We could paraphrase this point per Scanlon’s ‘idea of a justification which it would be unreasonable to reject’ (1982, p.117). That is to say, in justifying the application of the system of punishment, we should satisfy the condition that each person concerned (especially the violator) is aware of having no grounds to reasonably reject the application of the system, even if they do in fact reject it from their personal, self-interested point of view.

In fact, if the violator is not theoretically persuaded at all in any sense—that is, if they cannot understand the justification as a justification—we must consider the possibility that they suffer some disorder or disability that affects their criminal responsibility.

We should also take into account the case of some extreme and fanatical terrorists. They might not understand the physical treatment inflicted on them in the name of punishment as a punishment at all. Rather, they might interpret their being physically harmed as an admirable result of their heroic behaviour. The notion of punishment is not easily applied to these cases, where the use of physical restraint is more like that applied to wild animals. Punishment can be successful only if those who are punished understand the event as punishment.

This line of argument entirely conforms to the traditional context in philosophy concerning the concept of a “person”, who is regarded as the moral and legal agent responsible for his or her actions, including crimes. John Locke, a 17th-century English philosopher, introduced and established this concept, basing it on ‘consciousness’. According to Locke, a person ‘is a thinking intelligent Being, that has reason and reflection, and can consider it self as it self, the same thinking thing in different times and places; which it does only by that consciousness’ (1975, Book 2, Chapter 27, Section 9). This suggests that moral or legal punishments for the person should be accompanied by consciousnesses (in a Lockean sense) of the agent. In other words, when punishment is legally imposed on someone, the person to be punished must be conscious of the punishment as a punishment; that is, the person should understand the event as a justified imposition of some harm.1

However, there is a problem here, which arises in particular for the death penalty but not for other kinds of punishment. The question that I raise here is ‘to whom do we justify the death penalty?’ People might say it should be justified to society, as the death penalty is one of the social institutions to which we consent, whether explicitly or tacitly. This is true. However, if my claims above about justification are correct, the justification of the death penalty must involve the condemned convict coming to understand the justification at least at a theoretical level. Otherwise, to be executed would not be considered a punishment but rather something akin to the extermination of a dangerous animal. The question I want to focus on in particular is this: should this justification be provided before administering capital punishment or whilst administering capital punishment?

2. ‘Impossibilism’

Generally, in order for the justification of punishment to work, it is necessary for convicts to understand that this is a punishment before it is carried out and that they cannot reasonably reject the justification, regardless of any personal objection they may have. However, that is not sufficient, because if they do not understand at the moment of execution that something harmful being inflicted is a punishment, then its being inflicted would simply result in mere physical harm rather than an institutional response based on theoretical justification. The justification for punishment must be, at least theoretically, accepted both before and during its application.2 This requirement can be achieved with regard to many types of punishment, such as fines or imprisonment. However, the situation is radically different in the case of the death penalty, for in this case, when it is carried out, the convict, by definition, disappears. During and (in the absence of an afterlife) after the punishment, the convict cannot understand the nature and justification of the punishment. Can we say then that this is a punishment? This is a question which deserves further thought.

On the one hand, the death penalty, once executed, logically implies the nonexistence of the person punished; therefore, by definition, that person will not be conscious of being punished at the moment of execution. However, punishment must be accompanied by the convict’s consciousness or understanding of the significance of the punishment, as far as we accept the traditional concept of the person as a moral and legal agent upon whom punishment could be imposed. It may be suggested that everything leading up to the execution—being on death row, entering the execution chamber, being strapped down—is a kind of punishment that the convict is conscious of and is qualitatively different from mere incarceration. However, those phases are factors merely concomitant with the death penalty. The core essence of being executed lies in being killed or dying. Therefore, if the phases of anticipation were to occur but finally the convict were not killed, the death penalty would not have been carried out. The death penalty logically results in the convict’s not being conscious of being executed, and yet, for it to be a punishment, the death penalty requires the convict to be conscious of being executed. We could notate this in the form of conjunction in the following way in order to make my point as clear as possible:

~ PCE & PCE

(PCE: ‘the person is conscious of being executed under the name of punishment’)

If this is correct, then we must conclude that the concept of the death penalty is a manifest contradiction in terms. In other words, the death penalty should be regarded as conceptually impossible, even before we take part in longstanding debates between retentionism and abolitionism. This purely philosophical view of the death penalty could be called ‘impossibilism’ (i.e. the death penalty is conceptually impossible), and could be classified as a third possible view on the death penalty, distinct from retentionism and abolitionism. A naïve objection against this impossibilist view might counter that the death penalty is actually carried out in some countries so that it is not impossible but obviously possible. The impossibilist answer to this objection is that, based on a coherent sense of what it means for a punishment to be justified, that execution in such countries is not the death penalty but rather unjustified lethal physical violence.

I am not entirely certain whether the ‘impossibilist’ view would truly make sense in the light of the contemporary debates on the death penalty. These debates take place between two camps as I referred to above:

Retentionism (the death penalty should be retained): generally argued with reference to victims’ feelings and the deterrence effects expected by execution.

Abolitionism (the death penalty should be abolished): generally argued through appeals to the cruelty of execution, the possibility of misjudgements in the trial etc.

The grounds mentioned by both camps are, theoretically speaking, applicable to punishment in general in addition to the death penalty specifically. I will mention those two camps later again in a more detailed way in order to make a contrast between standard debates and my own view. However, my argument above for ‘impossibilism’, does suggest that there is an uncertainty specific to the death penalty as opposed to other types of punishment. I believe that this uncertainty must be considered when we discuss the death penalty, at least from a philosophical perspective. Otherwise we may lose sight of what we are attempting to achieve.

A related idea to the ‘impossibilism’ of the death penalty may emerge, if we accept the fact that the death penalty is mainly imposed on those convicted of homicide. This idea is related to the understanding of death proposed by Epicurus, who provides the following argument (Diogenes Laertius 1925, p. 650-1):

Death, therefore, the most awful of evils, is nothing to us, seeing that, when we are, death is not come, and, when death is come, we are not. It is nothing, then, either to the living or to the dead, for with the living it is not and the dead exist no longer.

We can call this Epicurean view ‘the harmlessness theory of death’ (HTD). If we accept HTD, it follows, quite surprisingly, that there is no direct victim in the case of homicide insofar as we define ‘victim’ to be a person who suffers harm as a result of a crime. For according to HTD, people who have been killed and are now dead suffer nothing—neither benefits nor harms—because, as they do not exist, they cannot be victims. If this is true, there is no victim in the case of homicide, and it must be unreasonable to impose what is supposed to be the ultimate punishment3—that is, the death penalty—on those offenders who have killed others.

This argument might sound utterly absurd, particularly if it is extended beyond offenders and victims to people in general, as one merit of the death penalty seems to lie in reducing people’s fear of death by homicide. However, although this argument from HTD might sound bizarre and counterintuitive, we should accept it at the theoretical level, to the extent that we find HTD valid.4 Clearly, this argument, which is based on the nonexistence of victims, could logically lead to another impossibilist argument concerning the death penalty.

There are many points to be more carefully examined regarding both types of ‘impossibilism’, which I will skip here. However, I must stop to ponder a natural reaction. My question above, ‘To whom do we justify?’, which introduced ‘impossibilism’, might sound eccentric, because, roughly speaking, theoretical arguments of justification are usually deployed in a generalised way and do not need to acknowledge who those arguments are directed at. Yet, I believe that this normal attitude towards justification is not always correct. Instead, our behaviour, when justifying something, focuses primarily on theoretically persuading those who are unwilling to accept the item being justified. If nobody refuses to accept it, then it is completely unnecessary to provide its justification. For instance, to use a common sense example, nobody doubts the existence of the earth. Therefore, nobody takes it to be necessary to justify the existence of the earth. Alternatively, a justification for keeping coal-fired power generation, the continued use of which is not universally accepted due to global warming, is deemed necessary. In other words, justification is not a procedure lacking a particular addressee, but an activity that addresses the particular person in a definite way, at least at first. In fact, it seems to me that the reason that current debates on the death penalty become deadlocked is that crucial distinctions are not appropriately made. I think that such a situation originates from not clearly asking to whom we are addressing our arguments, or whom we are discussing. As far as I know, there have been very few arguments within the death penalty debate that take into account the homicide victim, despite the victim’s unique status in the issue. This is one example where the debate can be accused of ignoring the ‘whom-question’, so I will clarify this issue by adopting a strategy in which this ‘whom-question’ is addressed.

3. Three chronological stages

Following my strategy, I will first introduce a distinction between three chronological stages in the death penalty. In order to make my argument as simple as possible, I will assume that the death penalty is imposed on those who have been convicted of homicide, although I acknowledge there are other crimes which could result in the death penalty. In that sense, the three stages of the death penalty correspond to the three distinct phases arising from homicide.

The first stage takes place at the time of killing; the fact that someone was killed must be highlighted. However, precisely what happened? If we accept the HTD, we should suppose that nothing harmful happened in the case of homicide. Although counterintuitive, let’s see where this argument leads. However, first, I will acknowledge that we cannot cover all contexts concerning the justification of the death penalty by discussing whether or not killing harms the killed victim. Even if we accept for argument’s sake that homicide does not harm the victim, that is only part of the issue. Other people, particularly the bereaved families of those killed, are seriously harmed by homicide. More generally, society as a whole is harmed, as the fear of homicide becomes more widespread in society.

Moreover, our basic premise, HTD, is controversial. Whether HTD is convincing remains an unanswered question. There is still a very real possibility that those who were killed do suffer harm in a straightforward sense, which conforms to most people’s strong intuition. In any event, we can call this first stage, the ‘Harm Stage’, because harm is what is most salient in this phase, either harm to the victims or others in society at large. If a justification for the death penalty is to take this Harm Stage seriously, the overwhelming focus must be on the direct victims themselves, who actually suffer the harm. This is the central core of the issue, as well as the starting point of all further problems.

The second stage appears after the killing. After a homicide, it is common to blame and to feel anger towards the perpetrator or perpetrators, and this can be described as a natural, moral, or emotional reaction. However, it is not proven that blaming or feeling angry is indeed natural, as it has not been proven that such feelings would arise irrespective of our cultural understanding of the social significance of killing. The phenomenon of blaming and the prevalence of anger when a homicide is committed could be a culture-laden phenomenon rather than a natural emotion. Nevertheless, many people actually do blame perpetrators or feel anger towards them for killing someone, and this is one of the basic ideas used to justify a system of ‘retributive justice’. The core of retributive justice is that punishment should be imposed on the offenders themselves (rather than other people, such as the offenders’ family). This retributive impulse seems to be the most fundamental basis of the system of punishment, even though we often also rely on some consequentialist justification favor punishment (e.g. preventing someone from repeating an offence). In addition, offenders are the recipients of blame or anger from society, which suggests that blaming or expressing anger has a crucial function in retributive justice. I will call this second phase the ‘Blame Stage’, which extends to the period of the execution. Actually, the act of blaming seems to delineate what needs to be resolved in this phase. Attempting to justify the death penalty by acknowledging this Blame Stage (or retributive justification) in terms of proportionality is the most common strategy. That is to say, lex talionis applies here—‘an eye for an eye’. This is the justification that not only considers people in general, including victims who blame perpetrators, but also attempts to persuade perpetrators that this is retribution resulting from their own harmful behaviours.

The final stage in the process concerning the death penalty appears after the execution; in this stage, what matters most is how beneficial the execution is to society. Any system in our society must be considered in the light of its cost-effectiveness. This extends even to cultural or artistic institutions, although at first glance they seem to be far from producing any practical effects. In this context, benefits are interpreted quite broadly; creating intellectual satisfaction, for example, is counted as a benefit. Clearly, this is a utilitarian standpoint. We can apply this view to the system of punishment, or the death penalty, if it is accepted. That is, the death penalty may be justified if its benefits to society are higher than its costs. What, then, are the costs, and what are the benefits? Obviously, we must consider basic expenses, such as the maintenance and labour costs of the institution keeping the prisoner on death row. However, in the case of the death penalty, there is a special cost to be considered, namely, the emotional reaction of people in society in response to killing humans, even when officially sanctioned as a punishment. Some feel that it is cruel to kill a person, regardless of the reason.

On the other hand, what is the expected benefit of the death penalty? The ‘deterrent effect’ is usually mentioned as a benefit that the death penalty can bring about in the future. In that case, what needs to be shown if we are to draw analogies with the previous two stages? When people try to justify the death penalty by mentioning its deterrent effect, they seem to be comparing a society without the death penalty to one with the death penalty. Then they argue that citizens in a society with the death penalty are at less risk of being killed or seriously victimised than those in a society without the death penalty. In other words, the death penalty could reduce the danger of being killed or seriously victimised in the future. Therefore, we could call this third phase the ‘Danger Stage’. In this stage, we focus on the danger that might affect people in the future, including future generations. This is a radically different circumstance from those of the previous two stages in that the Danger Stage targets people who have nothing to do with a particular homicide.

4. Analogy from natural disasters

The three chronological stages that I have presented in relation to the death penalty are found in other types of punishment as well. Initially, any punishment must stem from some level of harm (including harm to the law), and this is a sine qua non for the issue of punishment to arise. Blaming and its retributive reaction must follow that harm, and subsequently some social deterrent is expected to result. However, we should carefully distinguish between the death penalty and other forms of punishment. With other forms of punishment, direct victims undoubtedly exist, and those convicted of harming such victims are aware they are being punished. In addition, rehabilitating perpetrators in order for them to return to society—one aspect of the deterrent effect—can work in principle. However, this aspect of deterrence cannot apply to the death penalty because executed criminals cannot be aware of being punished by definition, and the notion of rehabilitation does not make sense by definition. Only this quite obvious observation can clarify that there is a crucial, intrinsic difference or distinction between the death penalty and other forms of punishment. Theories about the death penalty must seriously consider this difference; we cannot rely on theories that treat the death penalty on a par with other forms of punishment.

Moreover, the three chronological stages that have been introduced above are fundamentally different from each other. In reality, the subjects or people that we discuss and on whom we focus are different from stage to stage. In this respect, one of my points in this article is to underline the crucial need to discuss the issues of the death penalty by drawing a clear distinction between those stages. I am not claiming that only one of those stages is important. I am aware that each stage has its own significance; therefore, we should consider all three. However, we should be conscious of the distinctions when discussing the death penalty.

To make my point more understandable, I will suggest an analogy with natural disasters. Specifically, I will use as an analogy the biggest earthquake in Japan in the past millennium—the quake of 11 March 2011 (hereafter the 2011 quake). Of course, at first glance, earthquakes are substantially different from homicides. However, there is a close similarity between the 2011 quake and homicides, because although most of the harm that occurred was due to the earthquake and tsunami, in fact people were also harmed and killed during the 2011 quake at least partially due to human errors, such as the failure of the government’s policy on tsunamis and nuclear power plants. Thus, it is quite easy in the case of the 2011 quake to distinguish between three aspects, all of which are different from each other.

(1) We must recognise victims who were killed in the tsunami or suffered hardship at shelters.5 This is the core as well as the starting point of all problems. What matters here is rescuing victims, and expressing our condolences.

(2) Then we will consider victims and people in general who hold the government and the nuclear power company responsible for political and technical mistakes. What usually matters here is the issue of responsibility and compensation.

(3) Finally, we can consider people’s interests in improving preventive measures taken to reduce damages by tsunami and nuclear-plant-related accidents in the future. What matters in this context is the reduction of danger in the future by learning from the 2011 quake.

Nobody will fail to notice that these three aspects are three completely different issues, which can be seen in exactly the same manner in the case of the death penalty. Aspects (1), (2), and (3) correspond respectively to the Harm Stage, the Blame Stage, and the Danger Stage. Undoubtedly, none of these three aspects should be ignored and they actually appear in a mutually intertwined manner: the more successful the preventive measures are, the fewer victims will be produced by tsunami and nuclear-plant accidents in the future. Those aspects affect each other. Likewise, we must consider each of the three stages regarding the death penalty.

5. Initial harm

The arguments thus far provide the basic standpoint that I want to propose concerning the debates on the death penalty. I want to investigate the issue of the death penalty by sharply distinguishing between these three stages and by simultaneously considering them all equally. By following this strategy, I will demonstrate that there are intrinsic uncertainties, and four problems resulting from those uncertainties, in the system of the death penalty. In so doing I will raise a novel objection to the contemporary debate over the death penalty.

Roughly speaking, as I have previously mentioned, the death penalty debate continues to involve the two opposing views of abolitionism and retentionism (or perhaps, in the case of abolitionist countries, revivalism). It seems that the main arguments to support or justify each of the two traditional views (which I have briefly described in section 2 above) have already been exhausted. What matters in this context is whether the death penalty can be justified, and then whether—if it is justifiable—it should be justified in terms of retributivism or utilitarianism. That is the standard way of the debate on the death penalty. For example, when the retributive standpoint is used to justify the death penalty, the notion of proportionality as an element of fairness or social justice might be relevant, apart from the issue of whether proportionality should be measured cardinally or ordinally (see von Hirsch 1993, pp. 6-19). In other words, if one person has killed another, then that person too ought to be killed—that is, executed—in order to achieve fairness. However, as other scholars such as Tonry (1994) have argued, it is rather problematic to apply the notion of proportionality to the practice of punishment because it seems that there is no objective measure of offence, culpability, or responsibility. Rather, the notion of parsimony6 is often mentioned in these contexts as a more practical and fairer principle than the notion of proportionality.

However, according to my argument above, such debates are inadequate if they are simply applied to the case of the death penalty. Proportionality between which two things is being discussed? Most likely, what is considered here is the proportionality between harm by homicide (where the measured value of offence might be the maximum) and harm by execution. However, I want to reconfirm the essential point. What specifically is the harm of homicide? Whom are we talking about when we discuss the harm of homicide? As I previously argued, citing Epicurus and his HTD, there is a metaphysical doubt about whether we should regard death as harmful. If a person simply disappears when he or she dies and death is completely harmless as HTD claims, then it seems that the retributive justification for the death penalty in terms of proportionality must be nonsense, for nothing at all happens that should trigger the process of crime and punishment. Of course, following HTD, the execution should be similarly regarded as nonsensical. However, if that is the case, the entire institutional procedure, from the perpetrator’s arrest to his or her execution, must be considered a tremendous waste of time, labour, and money.

Some may think that these kinds of arguments are merely empty philosophical abstractions. That may be. However, it is not the case that there is nothing plausible to be considered in these arguments. Consider the issue of euthanasia. Why do people sometimes wish to be euthanised? It is because people can be relieved of a painful situation by dying. That is to say, people wishing to be euthanised take death to be painless, i.e. harmless, in the same manner as HTD. This idea embedded in the case of euthanasia is so understandable that the issue of euthanasia is one of the most popular topics in ethics; however, if so, Epicurus’s HTD should not be taken as nonsensical, for HTD holds in the same way as the idea embedded in the case of euthanasia that when we die, we have neither pain nor any other feeling. What I intend to highlight here is that we must be acutely aware that there is a fundamental problem concerning the notion of harm by homicide, if we want to be philosophically sincere and consistent7.

In other words, I assert that the contemporary debate over the death penalty tends to lack proper consideration for the Harm Stage in which victims themselves essentially matter, although that stage must be the very starting point of all issues. We must understand this pivotal role of the Harm Stage before intelligently discussing the death penalty. Of course, in practice, we can discuss the death penalty in a significant and refined manner without investigating the Harm Stage. For example, according to Goldman, one of the plausible positions regarding the justification for punishment in general is a position that combines both retributivism and utilitarianism. Mentioning John Rawls and H. L. A. Hart, Goldman writes (1995, p. 31):

Some philosophers have thought that objections to these two theories of punishment could be overcome by making both retributive and utilitarian criteria necessary for the justification of punishment. Utilitarian criteria could be used to justify the institution, and retributive to justify specific acts within it.

Goldman argues, however, that this mixed position could result in a paradox regarding how severe the punishment to be imposed on the guilty should be, even though this position avoids punishing the innocent (ibid., p.36):

While the mixed theory can avoid punishment of the innocent, it is doubtful that it can avoid excessive punishment of the guilty if it is to have sufficient effect to make the social cost worthwhile.

This argument is useful in providing a moral and legal warning to society not to punish offenders more severely than they deserve, even if that punishment is more effective in deterring future crimes. I frankly admit that Goldman’s suggestion goes to the essence of the concept of justice. However, I must also say that if his argument is applied to the death penalty, then it has not yet touched the fundamental question that forms the basis of the whole issue: whose harm should we discuss? Is it appropriate not to discuss the Harm Stage? Alternatively, I am raising the following question: who is the victim of homicide? At the very least, I think we should admit that this very question is the crucial one constituting the first problem on the death penalty, the Uncertainty of Harm.

6. Feeling of being victimised

Next, I will examine another kind of uncertainty that is specific to the Blame Stage; the idea of retribution matters here. As far as the Japanese context for the death penalty is concerned, according to statistical surveys of public opinion, people tend to strongly support the death penalty in the case of particularly violent homicides in which they are probably feeling particularly victimised. If the death penalty were abolished, it seems that the abolition would be extremely unfair to victims of homicide, as the rights of victims (i.e. rights of life, liberty, property, and so on) would be denied by being killed, whereas those of perpetrators would be excessively protected. Obviously, the notion of retributive proportionality or equilibrium is the basis for this argument. To put it another way, this logic of retribution aims at justifying the death penalty in terms of its achieving equilibrium between the violated rights of victims and the deprived rights of perpetrators in the name of punishment. Is this logic perfectly acceptable? Emotionally speaking, I want to say yes. We Japanese might even say that perpetrators should gallantly and bravely kill themselves to take responsibility for their actions, as we have a history of the samurai who were expected to conduct hara-kiri when they did something shameful. However, theoretically speaking, we cannot accept this logic immediately, because there are too many doubtful points. Those doubts as a whole constitute the second problem concerning the death penalty.

First, we must ask, as well as in the previous section, on the issue of feeling victimised, whom are we discussing? Whose feelings and whose rights matter? Direct victims in the case of homicide do not exist by definition. Then a question arises: why can substitutes (prosecutors and others) or the bereaved family ask for the death penalty based on their feelings rather than the direct victim’s feeling? How are they qualified to ask for such a stringent punishment when they were not the ones killed? The crucial point to be noted here is that the bereaved family is not identical with the direct victim. Second, even if it is admitted that the notion of the victim’s emotional harm are relevant to sentencing (and at least in the sense of emotional harm the bereaved family’s suffering I would agree that this makes them certainly the principal victims even if not the direct victim), it must be asked: can we justify an institution based on a feeling? This question is a part of the traditional debate concerning the moral sense theory. We have repeatedly asked whether social institutions can be based on moral sense or human feeling, when such sense or feeling cannot help but be arbitrary because those, after all, are subjective. The question is still unanswered. Third, if the feelings of being victimised justify the death penalty, then could an accidental killing or involuntary manslaughter be included in crimes that deserve the death penalty? Actually, the feelings of the bereaved family in the case of accidental killing could be qualitatively the same as in the case of voluntary homicide. However, even countries which adopt the death penalty do not usually prescribe that execution is warranted for accidental killing. Fourth, I wonder whether the bereaved family who feel victimised always desire the execution of the killer. It could be that they consider resuming their daily lives more important than advocating the execution of the murderer who killed their family member. As a matter of practical fact, executions of perpetrators need have nothing to do with supporting bereaved families. Fifth, if we accept the logic in which the death penalty is justified by the bereaved family’s feeling of being victimised, how should we deal with cases where the person who was killed was alone in the world, with no family? If there is no bereaved family, then no one feels victimised. Is the death penalty unwarranted in this case? In any case, as these questions suggest, we should be aware that retributive justification based upon the feeling of being victimised is not as acceptable as we initially expected. Once again, there is uncertainty here. Uncertainty of blame leads to the second problem concerning the death penalty.

7. Violation and forfeiture

Of course, the retributive justification for the death penalty does not have to depend upon the feeling of being victimised alone, even if the primitive basis for it might lie in human emotion. The theoretical terminology of human rights themselves (rather than emotional feeling based on the notion of rights) could be used as justification: if a person violates another’s rights (to property, freedom, a healthy life, etc.), then that person must forfeit his or her own rights in proportion to the violated rights. This can be regarded as a formulation of the system of punishment established in the modern era that is theoretically based upon the social contract theory. The next remark of Goldman confirms this point (1995, p.33):

If we are asked which rights are forfeited in violating the rights of others, it is plausible to answer just those rights that one violates (or an equivalent set). One continues to enjoy rights only as long as one respects those rights in others: violation constitutes forfeiture . . . Since deprivation of those particular rights violated is often impracticable, we are justified in depriving a wrongdoer of some equivalent set, or in inflicting harm equivalent to that which would be suffered in losing those same rights.

However, the situation is not so simple, particularly in connection with the death penalty. In order to clarify this point, we have to reflect, albeit briefly, on how the concept of human rights has been historically established. I will trace the origin of the concept of human rights by referring to Fagan’s overall explanation. According to Fagan (2016, Section 2):

Human rights rest upon moral universalism and the belief in the existence of a truly universal moral community comprising all human beings . . . The origins of moral universalism within Europe are typically associated with the writings of Aristotle and the Stoics.

Followed by the remark:

Aristotle unambiguously expounds an argument in support of the existence of a natural moral order. This natural order ought to provide the basis for all truly rational systems of justice . . . The Stoics thereby posited the existence of a universal moral community effected through our shared relationship with god. The belief in the existence of a universal moral community was maintained in Europe by Christianity over the ensuing centuries.

This classical idea was linked during the 17th and 18th centuries to the concept of ‘natural law’ including the notion of ‘natural rights’ that each human being possesses independently of society or policy. ‘The quintessential exponent of this position was John Locke . . . Locke argued that natural rights flowed from natural law. Natural law originated from God’ (ibid.). Fagan continues (ibid.):

Analyses of the historical predecessors of the contemporary theory of human rights typically accord a high degree of importance to Locke’s contribution. Certainly, Locke provided the precedent of establishing legitimate political authority upon a rights foundation. This is an undeniably essential component of human rights.

Although, of course, we should take post-Lockean improvement including Kantian ideas into account to fully understand contemporary concepts of human rights, we cannot deny that Locke’s philosophy ought to be considered first.

As is well known, Locke’s argument focuses on property rights. He put forth the idea that property rights were based on our labour. Thus, his theory is called ‘the labour theory of property rights’. Let me quote the famous passage I have in mind (Locke 1960, Second Treatise, Section 27):

Though the Earth, and all inferior Creatures be common to all Men, yet every Man has a Property in his own Person. This no Body has any Right to but himself. The Labour of his Body, and the Works of his Hands, we may say, are properly his.

This idea could cover any kind of human rights such as those for living a healthy life, liberty, and property, because human rights are supposed to be owned by us. For example, H. L.A. Hart once argued that legal rights are nothing but legal powers to require others to meet correlative obligations, and then pointed out that; ‘we also speak of the person who has the correlative right as possessing it or even owning it’ (Hart 1982, p.185). If this is the case, we can make property rights representative of all human rights.

However, if we follow Locke’s theory (and many countries, including Japan, still do), then it logically follows that what we cannot gain by our labour by definition cannot be objects of human rights. How does Locke’s idea apply to our life itself (rather than simply living a healthy life)? Are we able to acquire our life itself by our labour? No, we cannot. We can realise a healthy life by making an effort to be moderate, but we cannot create our lives. We are creatures or animals; therefore, our lives are not something that we ourselves made by our labour. Locke uses the concept of power (as Hart does) when he discusses various aspects of property rights. Among those, we should pay particular attention to the following (Locke 1960, Second Treatise, Section 23):

For a Man, not having the Power of his own life, cannot, by Compact, or his own Consent, enslave himself to any one, nor put himself under the Absolute, Arbitrary Power of another, to take away his Life, when he pleases.

Locke also writes (1960, Section 24):

No Man can, by agreement, pass over to another that which he hath not in himself, a Power over his own life.

Obviously, Locke assumes that we have no property rights over our own lives or bodies themselves, or more precisely, no property rights in controlling and destroying our own lives as a whole; therefore, we cannot alienate those rights to others. We cannot alienate or forfeit what we do not have. If this is the case and we presuppose the formulation of the system of punishment introduced above in terms of violation and forfeiture, what would result? The answer is clear. Our lives themselves are conceptually beyond the terminology of human rights, and thus, if the death penalty is defined as a punishment requiring the forfeiture of the perpetrator’s right to life, the death penalty should be regarded as conceptually contradictory or impossible. We cannot lose tails, as we do not have tails. Likewise, we cannot own our lives (i.e. we have no property rights in our life itself), so we cannot lose our lives, at least in such a sense as forfeiture of human rights. This is the third route to an ‘impossibilist’ view of the death penalty. This argument depends heavily on Locke’s original theory. Nevertheless, as long as we have to consider Locke’s classical view seriously in order to discuss the relation between punishment and human rights, we must be aware that we could be involved in theoretical uncertainty in justifying the death penalty through the notion of human rights in a retributivist flavour, as the argument thus far suggests. This is the very puzzle that I want to propose as the third problem concerning the death penalty debates.

Moreover, we must acknowledge that retributive ideas in the Blame Stage usually include a kind of evaluation of the psychological state of the agent’s behaviour at the time of the crime as a matter of legal fact. In other words, rationality, freedom, or mens rea are usually needed for agents to be judged guilty. However, from a strictly philosophical perspective, we should say that it is far from easy in principle to confirm those states in the past. Indeed, this psychological trend seems to cause controversy in court proceedings, as seen, for example, in the American context known as ‘battered-woman syndrome’. If a woman who has been routinely battered by her partner suddenly fights back and kills her partner, American courts often find her not guilty. People wonder whether such an evaluation concerning battered women could be correctly made without arbitrariness. Additionally, philosophical debates on free will and the development of the brain sciences must be considered. Some philosophers assert that we have no free will because our personality and actions are intrinsically governed by external factors, such as our environments or biological conditions, which are definitely beyond our control. This philosophical standpoint is often called ‘hard incompatibilism’ (see Strawson 2008). In this respect, my analogy to a natural disaster could be seen as appropriate, as our actions might be taken to be just natural phenomena at the end of the day.8 Furthermore, brain sciences often provide shocking data to suggest that our will may be controlled by brain phenomena occurring prior to our consciousness, as shown by Benjamin Libet. In view of such contemporary arguments, we have little choice but to say that we cannot be perfectly certain whether a given perpetrator who committed homicide is truly guilty, as long as we adopt the present standard for judging the psychological states of offenders in court. To sum up, the third problem for the death penalty is the difficulty in knowing whether someone has property in their life itself as well as uncertainty about the mental state of the accused, this is the Uncertainty of Rights Violation.

8. The deterrent effect

Finally, I will examine some problems in the Danger Stage. What matters in this context is the utilitarian justification for the death penalty; I will focus on what is called the ‘deterrent effect’. Firstly, I would like to say that the death penalty undoubtedly has some deterrent effect. This is obvious if we imagine a society where violators of any laws, including minor infractions such as a parking ticket or public urination, must be sentenced to death. I believe that the number of all crimes would dramatically reduce in that society, although it would constitute a horrible dystopia. The argument for the deterrent effect of the death penalty probably arises from the same line of ‘common sense’ thinking. For example, Pojman says, ‘there is some non-statistical evidence based on common sense that gives credence to the hypothesis that the threat of the death penalty deters and that it does so better than long prison sentences’ (Pojman 1998, pp. 38-39). Specifically, this deterrent effect presupposes the utility calculus that a human being conducts, whether consciously or unconsciously, in terms of ‘weighing the subjective severity of perceived censure and the subjective probability of perceived censure against the magnitude of the desire to commit the offence and the subjective probability of fulfilling this desire by offending’ (Beyleveld 1979, p. 219). Therefore, if we presuppose the basic similarity of human conditions, it may be plausible to state the following about the deterrent effect of punishment: ‘this can be known a priori on the basis of an analysis of human action’ (ibid., p. 215). However, in fact, the death penalty in many countries is restricted to especially heinous crimes, such as consecutive homicides (although some countries apply the death penalty to a wider range of crimes), which suggests that we must conduct empirical studies, case by case, if we want to confirm the deterrent effect of the death penalty. Therefore, the question to be asked regarding the deterrent effect is not whether the death penalty is actually effective, but rather how effective it is in restricted categories of crimes. What matters is the degree.

There are many statistical surveys concerning this issue. In particular, an economic investigation by Ehrlich is often mentioned as a typical example. After examining detailed statistical data and taking into account various factors, such as race, heredity, education, and cultural patterns, Ehrlich suggests (1975, p. 414):

An additional execution per year over the period in question [i.e., 1935-1969] may have resulted, on average, in 7 or 8 fewer murders.

Of course, this estimate includes too many factors and presumptions to be perfectly correct. Ehrlich himself is aware of this and thus says (ibid.):

It should be emphasized that the expected tradeoffs computed in the preceding illustration mainly serve a methodological purpose since their validity is conditional upon that of the entire set of assumptions underlying the econometric investigation … however … the tradeoffs between executions and murders implied by these elasticities are not negligible, especially when evaluated at relatively low levels of executions and relatively high level[s] of murder.

Ehrlich’s study drew considerable criticism, most of which pointed out deficiencies in his statistical methodology. Therefore, at this moment, we should say that we are able to infer nothing definite from Ehrlich’s study, although we must value the study as pioneering work.

Van den Haag proposes an interesting argument based upon uncertainty specific to the deterrent effect of the death penalty. He assumes two cases, namely, case (1), in which the death penalty exists, and case (2), in which the death penalty does not exist. In each case there is risk or uncertainty. On the one hand, in case (1), if there is no deterrent effect, the life of a murderer is lost in vain, whereas if there is a deterrent effect, the lives of some murderers and innocent victims will be saved in the future. On the other hand, in case (2), if there is no deterrent effect, the life of a convicted murderer is saved, whereas if there is a deterrent effect, the lives of some innocent victims will be lost in the future (Van den Haag 1995, pp. 133-134). Conway and Pojman explain this argument using the following table, ‘The Best Bet Argument’, which I have modified slightly, having DP stand for the death penalty, and DE the deterrent effect:

THE WAGER

 

	DE works	DE does not work
We bet DP works	save: murderers and innocent victims in the future lose: convicted murderer	save: nothing affected lose: convicted murderer
We bet DP does not work	save: convicted murderer lose: innocent victims in the future	save: convicted murderer lose: nothing affected

 

Following this table, Conway assumes (after Van den Haag’s suggestion that the life of a convicted murderer is not valued more highly than that of the unknown victims) numerical values about each case (each numerical number stands for not a number of people but a hypothetical value for a person to be saved or killed) :

a murderer saved +5

a murderer executed -5

an innocent saved +10

an innocent murdered -10

Moreover, he assumes that for each execution, only two innocent lives are spared (i.e. he assumes the deterrent effect to be almost the minimum). Then, consequently, executing convicted murderers turns out to be a good bet (Conway 1995, pp. 265-266; Pojman 1998, pp. 40-41).

9. Negative causation and where to give priority

Van den Haag’s ‘Best Bet Argument’ sounds quite interesting. However, Conway has already proposed a fundamental challenge to this argument: it mistakenly regards the actual death of convicted murderers as being on a par with the possible death of innocent victims in the future (Conway 1995, pp. 269-270). This is confusing or possibly a rhetorical sleight of hand. I think that Conway’s reaction to Van den Haag’s argument is a reasonable one.

As I approach my conclusion, I will propose two problems with Van den Haag’s argument. First, I want to acknowledge that any arguments, including Van den Haag’s, supporting the death penalty in terms of its deterrent effect seem to presuppose a causal relationship between the existence of the death penalty and people not killing others. For example, Pojman writes, ‘the repeated announcement and regular exercise of capital punishment may have deep causal influence’ (1998, p. 48). However, epistemologically speaking, that presupposition is extremely hard to confirm, because the effect of this causal relationship is not a positive, but rather a negative event, which is the event of not killing others. This has something to do with the philosophical problem of how to understand negative properties. By negative properties we mean that, for example, my room is not full of seawater; my room does not consist of paper; my room is not melting us, etc. Such descriptions by negative properties can be made almost endlessly. In other words, one identical event described by a positive property (e.g., this room is well lit) can be re-described in infinite ways in terms of negative properties. Take the example that I am now at my computer in Tokyo, writing a paper. This event can also be described as ‘I am not eating’, ‘I am not sleeping’, ‘I am not killing others’ (!), etc. The positive event, ‘I am writing a paper now’, can be understood through a causal relationship. The event was most likely caused by my intention to do so, which was caused by my sense of duty as a professor, etc. How, then, could we understand the negative description of my action, ‘I am not killing others’? Was this caused by the existence of the death penalty in Japan?

Perhaps I was completely unaware of the existence of the death penalty in Japan when I wrote a paper without killing others. Could the death penalty be its cause? Could the negative event ‘I am not killing others’ be an effect of the death penalty? It is hard to say so.

This problem is the same as the problem of ‘causation by absence’ or ‘omission-involving causation’. Generally, causation by absence is usually examined in the form of answering a question about whether nothingness can cause something. For example, David Lewis discusses a question about how a void (understood as being entirely empty or nothing at all, differing from a vacuum) is regarded as a cause of something (Lewis 2004). He says, ‘If you were cast into a void, it would cause you to die in just a few minutes. It would suck the air from your lungs. It would boil your blood. It would drain the warmth from your body. And it would inflate enclosures in your body until they burst’ (ibid., p.277). However, the problem is that the void is nothing. ‘When the void sucks away the air, it does not exert an attractive force on the air’ (ibid.). Furthermore, another, perhaps harder problem would arise. We can say, ‘If I defended you from being cast into a void, you would not die’. Namely, my omission to defend you would cause you to die. However, should only my omission matter? What of your brother’s omission to defend you? Or the Prime Minister of the UK’s omission to defend you? Are not all of those qualified to be the cause of your death, as least as long as we adopt a common-sense counterfactual analysis of causation? As this argument suggests, in the context of the current debate on this problem, the most troublesome phase is that ‘too many’ absences can be supposed to cause a particular effect. I quote Menzies, who says (2004, p.145):

I am writing this essay at my computer. If, however, there were nerve gas in the air, or I were attacked with flamethrowers, or struck by a meteor shower, I would not be writing the essay. But it is counterintuitive to say that the absence of nerve gas, flamethrower attack, and meteor strike are causes of my writing the essay.

This example takes the issue of absence as a cause, but simultaneously his example refers to the case of effect as absence (not writing the essay). As this shows, the current debate on the problem of causation by absence could extend to the case of effect as absence. In any case, what matters is a possibility that ‘too many’ absences can cause something, and something can cause ‘too many’ absences (Menzies calls this problem ‘the problem of profligate causation’ (ibid., pp.142-145). Then the deterrent effect of the death penalty is definitely classified as a case of absence as effect rather than cause. In other words, the absence of homicide (as effect) matters, whereas in this case execution (as cause) is presupposed to exist. It seems that the current debate on causation by absence is highly likely to contribute to discussing the problem of the deterrent effect.

Of course, someone may counter my argument by saying that what matters in this context is a statistical correlation between the number of executions and the number of homicides, which could be confirmed in an empirical way. I admit that the statistical correlation plays a crucial role here, even though we must simultaneously acknowledge that what is called ‘randomized controlled trial’, the most reliable, statistical methodology to confirm causal relations, is unfeasible due to the nature of the problem. Actually, this kind of correlation is too rough to predict the causal relationship between those, although the causation really matters. Causes of a reduction or increase in the number of homicides can be interpreted or estimated in various ways, considering confounding factors, such as education, economic situation, urban planning, and so on. Therefore, in principle, there always remains the possibility that the apparent correlation between the death penalty and the reduction of homicides is merely accidental. For example, there may be another, common cause, that brings about both people’s tendency to support the death penalty and the reduction of homicides9. We should recognise that there is intrinsic uncertainty here. These difficulties concerning causal relations give rise to a fourth problem related to the death penalty debates – the Uncertainty of Causal Consequences.

Incidentally, let me now return to my distinction of the three stages regarding the death penalty. Obviously, the issue of the deterrent effect belongs primarily to the Danger Stage. Yet it is vital to consider the Harm Stage. How can the deterrent effect affect the Harm Stage? I must say that the retentionist’s argument, in terms of the deterrent effect of the death penalty, completely dismisses this essential point. We need only recall the analogy of the 2011 quake in Japan. ‘Retentionism’ based upon the deterrent effect corresponds to aspect (3), where the improvement of the preventive system matters. This is important, of course, but cannot be a priority. Priority lies in the issues of how to deal with the actual harm that the victims have already suffered (specifically referring to the bereaved family or others in the case of homicide and the death penalty). Without consideration of how to cope with the harm, even if the theory seriously considers the innocent victims in the future, the retentionists’ theory can hardly be persuasive.

It is true that the retentionists’ theory based on the deterrent effect appropriately considers the person harmed in the process of punishment. For example, Walker considers such a phase in the process of punishment as one of the possible objections against retentionism based on the deterrent effect by saying: ‘if the benefit excludes the person harmed this too is nowadays regarded by many people as morally unacceptable’ (Walker 1980, p. 65). However, as the context clearly shows, by ‘the person harmed’ he means the person punished. He does not mention the initial harm suffered by victims. This problem is concerned with my previous claim; that is, we have to consider the ‘whom-question’ when we discuss the justification of punishment. Whom are we discussing? Whose benefit do we consider? In the face of victims before our eyes, can we emphasise only the improvement of preventive systems for the future? Evidently, actual victims are the first to be helped, although obviously it is not at all bad to simultaneously consider the preventive system in the future. It is necessary for us to respect basic human rights and the human dignity of perpetrators and innocent people in the future; however, that respect must be in conjunction with our first taking care of actual victims. We ought not to get our priorities wrong.

10. Prospects

I have indicated that the debates on the death penalty are inevitably surrounded by four problems over specific kinds of uncertainties: uncertainty concerning the victim of homicide, uncertainty in justifying the death penalty from the feeling of being victimised, uncertainty in justifying the death penalty on the basis of human rights, and uncertainty over negative causation. In the course of examining these problems, I have proposed the option of developing an ‘impossibilist’ position about the death penalty, which I am convinced, deserves further investigation. However, being surrounded by theoretical problems and uncertainties might be more or less true of any social institution. My aim is only to suggest how the death penalty should be understood as involving uncertainties from a philosophical perspective. Most likely, if there is something practical that I can suggest based on my argument, then what we might call a ‘Harm-Centred System’ may be introduced as a relatively promising option instead of, or in tandem with, the death penalty. What I mean by this is a system in which we establish as a priority redressing actual harm with regard to legal justice, where ‘actual harm’ only implies what the bereaved family suffer from, as the direct victims have already disappeared in the case of homicide. In other words, I think that something akin to the maximalist approach to restorative justice10 or some hybrid of the traditional justice system and the restorative justice system should be seriously considered, although we cannot expect perfect solutions exempt from all of the above four problems. It is certainly worth considering whether some element of restorative justice can play a significant role in the best theory of punishment.

In any case, my argument is at most a philosophical attempt to address problems. How to apply it to the practice of the legal system is a question to be tackled in a future project.

REFERENCES

Bazemore, G. and Walgrave, L. 1999 (1). ‘Introduction: Restorative Justice and the International Juvenile Justice Crisis’. In Restorative Juvenile Justice: Repairing the Harm of Youth Crime, eds. G. Bazemore and L. Walgrave, Criminal Justice Press, 1-13.

———. 1999 (2). ‘Restorative Juvenile Justice: In Search of Fundamentals and an Outline for System Reform’. In Restorative Juvenile Justice: Repairing the Harm of Youth Crime, eds. G. Bazemore and L. Walgrave, Criminal Justice Press, 45-74.

Beyleveld, D. 1979. ‘Identifying, Explaining and Predicting Deterrence’. British Journal of Criminology 19:3, 205–224.

Calvert, B. 1993. ‘Locke on Punishment and the Death Penalty’. Philosophy 68:264,, 211–229.

Collins, J., N. Hall, and L. A. Paul. 2004. Causation and Counterfactuals. MIT Press.

Conway, D. A. 1995 (originally 1974). ‘Capital Punishment and Deterrence: Some Considerations in Dialogue Form’. In Punishment: A Philosophy and Public Affairs Reader, eds. J. Simmons, M. Cohen, J. Cohen, and C. R. Beitz. Princeton University Press, 261–273.

Diogenes Laertius. 1925. Lives of Eminent Philosophers. Vol. 2. Trans. R. D. Hicks. Loeb Classical Library. William Heinemann Ltd.

Ehrlich, I. 1975. ‘The Deterrent Effect of Capital Punishment: A Question of Life and Death’. American Economic Review 65:3, 397–417.

Fagan, A. 2016. ‘Human Rights’. In Chase B. Wrenn, The Internet Encyclopedia of Philosophy, ISSN 2161-0002. Available from http://www.iep.utm.edu/hum-rts/#H2 [Accessed 12 June 2017]

Fischer, J. M., ed. 1993. The Metaphysics of Death. Stanford University Press.

Goldman, A. H. 1995 (originally 1979). ‘The Paradox of Punishment’. In Punishment: A Philosophy and Public Affairs Reader, eds. J. Simmons, M. Cohen, J. Cohen, and C. R. Beitz. Princeton University Press, 30–46.

Hart, H. L. A. 1982. Essays on Bentham: Jurisprudence and Political Theory. Oxford University Press.

Ichinose, M. 2013. ‘Hybrid Nature of Causation’. In T. Uehiro, Ethics for the Future of Life: Proceedings of the 2012 Uehiro-Carnegie-Oxford Ethics Conference, the Oxford Uehiro Center for Practical Ethics, University of Oxford, 60-80.

———. 2016. ‘A Philosophical Inquiry into the Confusion over the Radiation Exposure Problem’. Journal of Disaster Research 11: No.sp, 770-779.

Lewis, D. 2004. ‘Void and Object’. In J. Collins, N. Hall, and L. A. Paul, Causation and Counterfactuals. MIT Press, 277–290.

Locke, J. 1960. Two Treatises of Government, ed. P. Laslett, Cambridge University Press.

———. 1975. An Essay concerning Human Understanding, ed. P. H. Nidditch. Oxford University Press.

Menzies, P. 2004. ‘Difference-Making in Context’. In J. Collins, N. Hall, and L. A. Paul, Causation and Counterfactuals. MIT Press, 139–180.

Pojman, L. P., and J. Reiman. 1998. The Death Penalty: For and Against. Rowman &Littlefield Publishers, Inc.

Scanlon, T. M. 1982. ‘Contractualism and utilitarianism’. In A. Sen and B. Williams, Utilitarianism and Beyond. Cambridge University Press, 103-128.,

Simmons, A. J. 1994. ‘Locke on the Death Penalty’. Philosophy 69:270, 471–477.

Strawson, G. 2008. ‘The Impossibility of Ultimate Moral Responsibility’. In Real Materialism. Oxford University Press, 319–331.

Tonry, M. 1994. ‘Proportionality, Parsimony, and Interchangeability of Punishments’. In A Reader on Punishment, eds. A. Duff and D. Garland. Oxford University Press, 133–160.

Van den Faag, E. 1995 (originally 1969). ‘On Deterrence and the Death Penalty’. In Punishment and the Death Penalty: The Current Debate, eds. R. M. Baird and S. E. Rosenbaum. Prometheus Books.

Von Hirsch, A. 1993. Censure and Sanctions. Oxford University Press.

Walen, A, 2015. ‘Retributive Justice’. The Stanford Encyclopedia of Philosophy (Summer 2015 Edition), Edward N. Zalta (ed.). Available from http://plato.stanford.edu/archives/sum2015/entries/justice-retributive/ [Accessed 12 June 2017]

Walker, N. 1980. Punishment, Danger and Stigma: The Morality of Criminal Justice. Barnes & Noble Books.