In this paper, I highlight three objections to vice-based accounts of moral evil: (1) the worry that vice-based accounts of evil are explanatorily inadequate; (2) the worry that even extreme vice is not sufficient for evil; and (3) the worry that not all vices are inversions of virtue (and so vice-based accounts will struggle to explain the “mirror thesis”). I argue that it is possible to respond to these objections by developing a vice-based account of evil that draws on insights (...) from virtue (and vice) epistemology. In this way, I seek to defend the strategy of understanding evil in terms of vice, and to provide guidance on how best to develop such an account. I also briefly consider what vice-based accounts of moral evil might imply about evil in other normative domains where it is common to talk of virtue and vice, including the possibilities of epistemic evil and aesthetic evil. (shrink)

Virtue ethics has many times been suggested as a promising recipe for the construction of artificial moral agents due to its emphasis on moral character and learning. However, given the complex nature of the theory, hardly any work has de facto attempted to implement the core tenets of virtue ethics in moral machines. The main goal of this paper is to demonstrate how virtue ethics can be taken all the way from theory to machine implementation. To achieve this goal, we (...) critically explore the possibilities and challenges for virtue ethics from a computational perspective. Drawing on previous conceptual and technical work, we outline a version of artificial virtue based on moral functionalism, connectionist bottom–up learning, and eudaimonic reward. We then describe how core features of the outlined theory can be interpreted in terms of functionality, which in turn informs the design of components necessary for virtuous cognition. Finally, we present a comprehensive framework for the technical development of artificial virtuous agents and discuss how they can be implemented in moral environments. (shrink)

Although virtue ethics has repeatedly been proposed as a suitable framework for the development of artificial moral agents, it has been proven difficult to approach from a computational perspective. In this work, we present the first technical implementation of artificial virtuous agents in moral simulations. First, we review previous conceptual and technical work in artificial virtue ethics and describe a functionalistic path to AVAs based on dispositional virtues, bottom-up learning, and top-down eudaimonic reward. We then provide the details of a (...) technical implementation in a moral simulation based on a tragedy of the commons scenario. The experimental results show how the AVAs learn to tackle cooperation problems while exhibiting core features of their theoretical counterpart, including moral character, dispositional virtues, learning from experience, and the pursuit of eudaimonia. Ultimately, we argue that virtue ethics provides a compelling path toward morally excellent machines and that our work provides an important starting point for such endeavors. (shrink)

Virtue ethics has many times been suggested as a promising recipe for the construction of artificial moral agents due to its emphasis on moral character and learning. However, given the complex nature of the theory, hardly any work has de facto attempted to implement the core tenets of virtue ethics in moral machines. The main goal of this paper is to demonstrate how virtue ethics can be taken all the way from theory to machine implementation. To achieve this goal, we (...) critically explore the possibilities and challenges for virtue ethics from a computational perspective. Drawing on previous conceptual and technical work, we outline a version of artificial virtue based on moral functionalism, connectionist bottom–up learning, and eudaimonic reward. We then describe how core features of the outlined theory can be interpreted in terms of functionality, which in turn informs the design of components necessary for virtuous cognition. Finally, we present a comprehensive framework for the technical development of artificial virtuous agents and discuss how they can be implemented in moral environments. (shrink)

Adam Morton, Stephen de Wijze, Hillel Steiner, and Eve Garrard have defended the view that evil action is qualitatively distinct from ordinary wrongdoing. By this, they do not that mean that evil actions feel different to ordinary wrongs, but that they have motives or effects that are not possessed to any degree by ordinary wrongs. Despite their professed intentions, Morton and de Wijze both offer accounts of evil action that fail to identify a clear qualitative difference between evil and ordinary (...) wrongdoing. In contrast, both Steiner's and Garrard's accounts of evil do point to qualitative distinctions between kinds of action, but it is implausible that either account correctly characterizes evil. The most plausible accounts maintain that evil actions have a necessary connection to extreme harms, and this suggests that evil is not qualitatively distinct from ordinary wrongdoing. (shrink)

In his book The Myth of Evil , Phillip Cole claims that the concept of evil divides normal people from inhuman, demonic and monstrous wrongdoers. Such monsters are found in fiction, Cole maintains, but not in reality. Thus, even if the concept of evil has the requisite form to be explanatorily useful, it will be of no explanatory use in the real world. My aims in this paper are to assess Cole’s arguments for the claim that there are no actual (...) evil persons, and, in so doing, to develop a clearer framework in which to think about evil personhood. While Cole is right to claim that there are no actual evil monsters or supernatural demons, he underestimates the extent to which ascriptions of demonic monstrosity are figurative rather than literal. Hence, a lack of actual monsters does not imply a lack of actual evil persons. More plausibly, Cole suggests that the concept of evil implies an unrealistically dualistic worldview, with purely evil people on one side and ordinary people on the other. Since no one is purely bad, Cole claims, the concept of evil fails to refer to actual persons. Cole is wrong to think that the use of extreme moral concepts is incompatible with fine-grained moral evaluations across a broad spectrum between the extremes. Nor is Cole sufficiently careful in unpacking the various ways in which a person might be considered purely bad. I will argue that some actual persons are extremely bad, that it is very likely that some actual persons are fixedly bad, and that quite possibly no actual persons are thoroughly bad or innately bad. It is plausible that a person is evil only if he is extremely and fixedly bad, but Cole is wrong to suppose that a person is evil only if he is thoroughly and innately bad. Thus, even if we accept Cole’s claim that no actual person is thoroughly or innately bad, it still seems very likely that some actual persons are evil, and hence that evil can be an explanatorily useful concept. (shrink)

It is intuitively plausible that not every evildoer is an evil person. In order to make sense of this intuition we need to construct an account of evil personhood in addition to an account of evil action. Some philosophers have offered aggregative accounts of evil personhood, but these do not fit well with common intuitions about the explanatory power of evil personhood, the possibility of moral reform, and the relationship between evil and luck. In contrast, a dispositional account of evil (...) personhood can allow that evil is explanatory, that an evil person can become good, and that luck might prevent evil persons from doing evil or cause non-evil persons to do evil. Yet the dispositional account of evil personhood implies that some evil persons are blameless, which seems to clash with the intuition that evil persons deserve our strongest moral condemnation. Moreover, since it is likely that a large proportion of us are disposed to perform evil actions in some environments, the dispositional account threatens to label a large proportion of people evil. In this paper I consider a range of possible modifications to the dispositional account that might bring it more closely into alignment with our intuitions about moral condemnation and the rarity of evil persons. According to the most plausible of these theories, S is an evil person if S is strongly disposed to perform evil actions when in conditions that favour S's autonomy. (shrink)

Christine Korsgaard’sSelf-Constitution: Agency, Identity, and Integrityis an impressive endeavour to synthesize the ethics of Plato and Kant in a comprehensive account of action and agency that locates the key to understanding both in self-constitution. A purportedly comprehensive account of action and agency will fail on its own terms if it cannot adequately account for some morally salient phenomenon. Korsgaard’s account fails to adequately account for the possibility of evil actions and evil people. If self-constitution is key to action and agency, (...) then we must abandon the Platonic and Kantian elements that Korsgaard endorses.L’ouvrage de Christine Korsgaard intituléSelf-Constitution: Agency, Identity, and Integrityreprésente un remarquable travail de synthèse des éthiques platonicienne et kantienne qui, de manière exhaustive, rend compte de l’action et de l’agentivité sous le prisme de la constitution de soi. Toute tentative de saisie parfaite de l’action et de l’agentivité en termes propres ne saurait aboutir sans une prise en compte adéquate d’un quelconque phénomène moral saillant. L’approche de Korsgaard ne prend pas suffisamment en compte la possibilité du mal sous forme d’actions ou d’individus. Si la constitution de soi demeure centrale pour l’action et l’agentivité, alors il faudra mettre à l’écart les éléments platoniciens et kantiens que Korsgaard fait siens. (shrink)

Philosophers theorizing about ‘evil’ usually distinguish evil actions from acts of ordinary wrongdoing. They either attempt to isolate some quality or set of qualities shared by all evil actions that is not found in other wrongful actions, or they concede that their account of evil is only distinguished by capturing the very worst acts on the scale of moral wrongness. The idea that evil is qualitatively distinct from wrongdoing has recently been under contention. We explore the grounds for this contention, (...) and argue that there is a third option that might be useful for a variety of philosophical accounts of evil. The alternate form of distinctness we propose is called quality of emphasis distinctness. We illustrate this form of concept distinctness with a modified version of Hillel Steiner’s account of evil. We then explain how QE distinctness could also be applied to more complex theories of evil, such as the theories proposed by Claudia Card, John Kekes, and Todd Calder. (shrink)

There is considerable philosophical dispute about what it takes for an action to be evil. The methodological assumption underlying this dispute is that there is a single, shared folk conception of evil action deployed amongst culturally similar people. Empirical research we undertook suggests that this assumption is false. There exist, amongst the folk, numerous conceptions of evil action. Hence, we argue, philosophical research is most profitably spent in two endeavours. First, in determining which (if any) conception of evil action we (...) have prudential or moral (or both) reason to deploy, and second, in determining whether we could feasibly come to adopt that conception as the single shared conception given our psychological make-up and the content of the conceptions currently deployed. (shrink)

Intuitively, one who counts a morally bad person as a friend has gone wrong somewhere. But it is far from obvious where exactly they have gone astray. Perhaps in cultivating a friendship with a bad person, one extends to them certain goods that they do not deserve. Or perhaps the failure lies elsewhere; one may be an abettor to moral transgressions. Yet another option is to identify the mistake as a species of imprudence—one may take on great personal risk in (...) counting a bad person as a friend. In this paper, I argue that none of these intuitive explanations are entirely convincing; for many such proposals run contrary to widely accepted features of friendship. However, they do point us in the direction of a more satisfying explanation—one which concerns a person’s moral priorities. An individual who counts a morally bad person as a friend is, I propose, one who betrays a distinct kind of defect in her values. (shrink)

In this article, I assess three contemporary criticisms levelled at Kant’s theory of evil in order to evaluate whether his theory can be saved. Critics argue that Kant does not adequately distinguish between evil and mundane wrongdoing, making his use of the term ‘evil’ emotional hyperbole; by defining evil as the subordination of the moral law to self-love his analysis is seemingly overly simplistic and empirically false; and by focusing solely on the moral character of the perpetrator of evil, Kant’s (...) theory apparently ignores the most salient aspect of evil – the suffering of victims. While I will not claim that Kant provides us with a fully adequate theory of evil, I respond to each of these criticisms and conclude that Kant’s theory can still provide significant insight into both the nature of evil and the moral psychology of perpetrators of evil. (shrink)

Evil acts strike us, by their very nature, as not only horrifying and reprehensible, but also as deeply puzzling. No doubt for reasons like this, evil has often been seen as mysterious, demonic and beyond our human powers of understanding. The question I examine in this paper is whether or not we can (or would want to) overcome this puzzlement in the face of evil acts. I shall argue that we ought want to (in all cases) and can (in at (...) least most cases) come to understand why people perpetrate evil acts. This is an appealing conclusion as it allows us to take practical steps to both minimise future occurrences of evil and come to terms with its past abominations. (shrink)

The concept of evil has been an unpopular one in many recent Western political and ethical discourses. One way to justify this neglect is by pointing to the many problemswiththe concept of evil. The standard grievances brought against the very concept of evil include: that it has no proper place in secular political and ethical discourses; that it is a demonizing term of hatred that leads to violence; that it is necessarily linked with outdated notions of body and sexuality; and (...) that it only hinders rather than aids our ability to understand. I shall seek to argue in defence of the concept of evil against these charges. The upshot of this argument is that the language and concept of evil has a justified and important role to play in political and ethical discourses. (shrink)

There are a number of different senses of the term “evil.” We examine in this paper the term “evil” when it is used to say things such as: “what Hitler did was not merely wrong, it was evil”, and “Hitler was not merely a bad person, he was an evil person”. Failing to keep a promise or telling a white lie may be morally wrong, but unlike genocide or sadistic torture, it is not evil in this sense. In this paper (...) we analyze the specific moral difference between “evil” and “mere wrongdoing”. In so doing we shall defend a specific conception of what acts and which persons should count as evil. On the view defended in this paper it is a necessary feature of an evil act that the victim of that act suffer what would at least normally be a life-wrecking or life-ending harm. (shrink)

Is evil a distinct moral concept? Or are evil actions just very wrong actions? Some philosophers have argued that evil is a distinct moral concept. These philosophers argue that evil is qualitatively distinct from ordinary wrongdoing. Other philosophers have suggested that evil is only quantitatively distinct from ordinary wrongdoing. On this view, evil is just very wrong. In this paper I argue that evil is qualitatively distinct from ordinary wrongdoing. The first part of the paper is critical. I argue that (...) Luke Russell’s attempt to show that evil is only quantitatively distinct from ordinary wrongdoing fails. Russell’s argument fails because it is based on an implausible criterion for determining whether two concepts are qualitatively distinct. I offer a more plausible criterion and argue that based on this criterion evil and wrongdoing are qualitatively distinct. To help make my case, I sketch a theory of evil which makes a genuinely qualitative distinction between evil and wrongdoing. I argue that we cannot characterize evil as just very wrong on plausible conceptions of evil and wrongdoing. I focus on act-consequentialist, Kantian, and contractarian conceptions of wrongdoing. (shrink)

Luke Russell's book, Evil: A Philosophical Investigation, is a marvelously clear and rigorous addition to the philosophical literature on the secular moral concept of evil. The term “evil,” in this...

In this article I propose a theological reflection on the philosophical assumptions behind the idea that intelligent machine can be governed through ethical protocols, which may apply either to the people who develop the machines or to the machines themselves, or both. This idea is particularly relevant in the case of machines’ extreme wrongdoing, a wrongdoing that becomes an existential risk for humankind. I call this extreme wrong-doing, ‘evil.’ Thus, this article is a theological account on the philosophical assumptions behind (...) the possibility of evil machines, machines that can create an existential risk for humankind, and the ethical remedies that limit that risk. After discussing these assumptions, I argue for the ineffectiveness of these ethical remedies to prevent the possibility of evil machines. The article is an application of Romano Guardini’s thought on technology to evil machines. (shrink)

Almost everyone allows that conditions can obtain that exempt agents from moral responsibility—that someone is not a morally responsible agent if certain conditions obtain. In his seminal Freedom and Resentment, Peter Strawson denies that the truth of determinism globally exempts agents from moral responsibility. As has been noted elsewhere, Strawson appears committed to the surprising thesis that being an evil person is an exempting condition. Less often noted is the fact that various Strawsonians—philosophers sympathetic with Strawson’s account of moral responsibility—at (...) least appear to have difficulty incorporating evil persons into their accounts of moral responsibility. In what follows, I argue that Strawson is not committed to supposing that being evil is an exempting condition—at least, that he can allow that evil persons are morally responsible agents. (shrink)

In this journal, Luke Russell defends a sophisticated dispositional account of evil personhood according to which a person is evil just in case she is strongly and highly fixedly disposed to perform evil actions in conditions that favour her autonomy. While I am generally sympathetic with this account, I argue that Russell wrongly dismisses the mirror thesis—roughly, the thesis that evil people are the mirror images of the morally best sort of persons—which I have defended elsewhere. Russell’s rejection of the (...) mirror thesis depends upon an independently implausible account of moral sainthood, one that is implausible for reasons that Russell himself suggests in another context. Indeed, an account of moral sainthood that parallels Russell’s account of evil personhood is plausible for the same reasons that his account of evil personhood is, and that suggests that Russell himself is actually committed to the mirror thesis. (shrink)

It is plausible that being an evil person is a matter of having a particularly morally depraved character. I argue that suffering from extreme moral vices—and not consistently lacking moral vices, for example—suffices for being evil. Alternatively, I defend an extremity account concerning evil personhood against consistency accounts of evil personhood. After clarifying what it is for vices to be extreme, I note that the extremity thesis I defend allows that a person could suffer from both extremely vicious character traits (...) while possessing some modest virtue as well. By contrast, consistency theses rule out this possibility by definition. This result does not suggest that extremity accounts are flawed, however, since, as I argue, the thesis that evil people must lack moral virtue altogether effectively defines evil people out of existence and prematurely privileges skepticism about evil personhood. Ultimately, I contend that an extremity account is most consistent with common intuitions about putative evil persons as well as plausible assumptions about aretaic evaluations of character quite generally. (shrink)

Some jurisdictions acknowledge, as a matter of positive law, the relevance of evil to capital punishment. At one point, the state of Florida counted that the fact that a murderer’s crime was “especially wicked, evil, atrocious or cruel” as an aggravating factor for purposes of capital sentencing. I submit that Florida may be onto something. I consider a thesis about capital punishment that strikes me as plausible on its face: if capital punishment is ever morally permissible, it is permissible as (...) a response to evil. Call this the Punishment as a Response to Evil thesis, or PRE. If capital punishment is not morally permissible as a response to evil, then, according to PRE, it is not morally permissible, period. PRE admits of at least two different readings: on the first, if capital punishment is ever morally justified it is justified as a punishment for evil crimes; on the second, if capital punishment is ever morally justified it is justified as a punishment for evil people. While this first version of PRE has found advocates in both philosophy and forensic psychiatry, I argue against this first reading of PRE and for the second. To secure this conclusion I appeal to an account of evil and evil personhood that I have developed elsewhere. (shrink)

ABSTRACTAlthough previous studies have examined killing as an outcome-oriented measure, few have explored non-killing as a socially organized process. Using letters written by soldiers, police offi...