1 Introduction

The concept of evil undoubtedly plays a role in everyday or “folk” understandings of morality. However, important questions remain over how best to understand this concept.Footnote 1 In this paper, I focus on accounts that define moral evil either by direct reference to the concept of moral vice or in direct contrast to the concept of moral virtue. We can refer to these as “vice-based” accounts of evil. My aim is to highlight three worries for such accounts, and to argue that it is possible to overcome these worries by developing an account of evil that draws on insights from work in virtue (and vice) epistemology. In this way, I aim both to defend the general strategy of understanding evil in terms of vice, and to provide guidance on how to successfully develop such an account.

I will begin (in Sect. 2) by explaining the basic idea behind vice-based accounts of evil, and by providing examples from the contemporary literature. My focus here, and throughout, is on evil persons rather than other possible bearers of evil, such as evil actions or evil institutions. I will then (in Sects. 3, 4, 5) set out three objections to vice-based accounts. My claim is that these objections can be overcome by reflecting on the decisions we make at the level of virtue theory. By committing to a virtue theory with certain features, it becomes possible for vice-based accounts to avoid the objections. Before concluding, I will consider (in Sect. 6) whether the proposals argued for in this paper could be accommodated by amending existing vice-based accounts. I will also briefly discuss what vice-based accounts of moral evil might imply about evil in other normative domains where talk of virtue and vice is common, including the possibilities of epistemic evil and aesthetic evil.

2 Vice-based accounts of moral evil

Vice-based accounts of moral evil define evil persons either by direct reference to the concept of moral vice or in direct contrast to the concept of moral virtue. Two contemporary examples of this approach can be found in the works of Peter Brian Barry (2010: 26) and Daniel Haybron (2002: 269):

Barry’s account: A person is evil in virtue of possessing extremely morally vicious states of character.

Haybron’s account: To be evil is, on my view, to be consistently vicious in the following sense: one is not aligned with the good to a morally significant extent. Evil persons… have no good side, but are consistently vicious.

These accounts differ in important ways. For example, Barry’s account focuses on extreme vice, while Haybron’s account focuses on consistent vice. As such, the accounts generate different verdicts concerning who ought to be considered morally evil. What these accounts have in common, however, is that they both seek to explain the concept of evil by directly appealing to the concept of moral vice. For this reason, both are examples of a vice-based approach to evil.

Another way of being included among vice-based accounts is by defining evil in direct contrast to the concept of moral virtue. For example, Eve Garrard approaches this topic by first committing to a claim about the nature of virtue, and then inverting that claim in order to account for the nature of evil.Footnote 2 Garrard’s chosen account of virtue, inspired by the work of John McDowell, tells us that a virtuous agent is one who “silences” any reasons that tell against performing a morally required action (Garrard, 2002: 329; and see Garrard, 1998). While a merely continent agent needs to fight against temptation, a genuinely virtuous agent is no longer moved by considerations that tell against doing the right thing. As Garrard (2002: 329) explains, having thus silenced any competing considerations, there is nothing for a virtuous agent to overcome, and so there is no struggle in doing what is right.

Garrard (2002: 330) inverts this understanding of moral virtue to account for moral evil. Whereas a virtuous agent silences any reasons that tell against doing what is right, an evil agent silences reasons against doing what is wrong:

Garrard’s account: It is not just that he allows less important considerations, such as his own power or pleasure, to outweigh these more forceful considerations, e.g., the suffering and loss of life of others; rather he is completely insensitive to these features’ reason-giving force. For him, there is nothing to be outweighed; he has (psychologically) silenced such considerations, and is unable to see that they are reasons for acting or refraining from action.

Importantly, Garrard (1998: 51–55) distinguishes between two types of silencing. “Psychological” silencing refers to when an agent does not see some feature of the situation as giving them a reason for (or against) acting. For example, an inconsiderate housemate might not see the inconvenience caused to others as giving him any reason to avoid leaving unwashed dishes around the house. “Metaphysical” silencing refers to cases where considerations that might otherwise be reason for acting genuinely cease to be reasons in the face of contrary considerations. In Garrard’s example (1998: 53), a football fan’s preference to get to the match on time ceases to be a reason for continuing his journey once he is faced with the need to save the life of a drowning child on the way.

According to Garrard, a virtuous agent has a pattern of psychological silencing that tracks metaphysical silencing. That is, a virtuous person (psychologically) silences those reasons for acting that are genuinely (metaphysically) silenced by competing considerations. This is why virtuous agents do not struggle with temptation when doing what is morally required. In cases of moral evil, this pattern of silencing is inverted. The evil agent psychologically silences reasons to avoid acting (such as the pain caused to others) that are in fact metaphysical silencers for the very reasons that motivate the evil agent (such as the evil agent’s own pleasure).

The accounts from Barry and Haybron explain evil by direct reference to the concept of vice. Garrard’s account explains evil in direct contrast to the concept of virtue. All three are contemporary examples of a vice-based approach to moral evil.

The vice-based approach to evil can be compared with alternative approaches. For example, an act-based approach to evil assigns primacy to the concept of evil action. Evil persons are defined as those who perform (or who are disposed to perform) evil acts, and supporters of this approach must then say something about the nature of evil acts. Positions available in the debate about what makes an action evil are many and varied. Prominent suggestions include the claim that evil acts are those that (culpably) cause significant or intolerable harm, and the claim that evil acts constitute a significant assault on another’s dignity.Footnote 3 The main contrast with vice-based accounts is that act-based accounts define evil persons in terms of what they do, whereas vice-based accounts focus on the evil person’s character (or on who they are).

The relationship between the vice-based approach and two other approaches discussed in the literature is more complicated. An affect-based approach defines evil persons by reference to certain evil-making feelings or emotions, while a motive-based approach defines evil persons as those who are motivated in certain ways, such as desiring to cause significant harm to others (Calder, 2020, Haybron, 2002: 265–271). The complication here is that (on a standard account of moral virtue), virtues are character traits, and character traits include both motivational and emotional dispositions.Footnote 4 Accounts of evil personhood that appeal to the notion of virtuous and vicious character traits are thereby appealing to a notion that is broader than, but encompasses, those features that are appealed to by affect-based and motive-based accounts.Footnote 5

My focus in this paper is on those accounts that can be clearly identified as vice-based accounts of evil due to their direct appeal to the concepts of virtue and vice. This includes at least those accounts proposed by Barry, Haybron and Garrard. I will not attempt to survey the merits and demerits of each of the other approaches. Instead, my aim is to advance the development of the vice-based approach, and to support the conclusion that this approach is a promising contender within the debate.

Whatever else we think about evil persons, it makes sense to expect that we will find them among the morally vicious, perhaps among the worst of the morally vicious. For this reason, vice-based accounts of evil have at least an initial plausibility. However, as I will now show, vice-based accounts also face significant objections. By setting out and responding to these objections, I aim to support the general strategy of understanding evil in terms of vice, and to provide guidance on how a successful vice-based account can be developed.

3 Objection 1—vice-based accounts are explanatorily inadequate

A standard aim among those who develop accounts of evil is the following: we want some explanation for why evil persons are willing and able to perform actions that non-evil persons would avoid. Indeed, providing this explanation has been suggested as a criterion for any successful account of evil.Footnote 6 If vice-based accounts are unable to provide this explanation, then this will, at the very least, give us reason to consider whether alternative approaches could do better.

An inability to explain evil actions is suggested as a possible weakness by Alfred Archer. Focusing specifically on Barry’s account, Archer (2016: 207) argues that this account is unsatisfying because it fails to provide any real explanation for why an evil person might act in morally troubling ways:

If someone were to ask how an evil person was able to perform an act that most people would find too morally repelling to perform they are unlikely to be satisfied with the explanation that the evil person possesses extremely vicious character traits.

Perhaps the thought here is that “X is extremely vicious” provides no more clarity on why X acted in a particular way than simply saying “X is evil”. What we want is an explanation for why being evil makes it possible (or perhaps even likely) for someone to act in problematic ways. If vice-based accounts cannot provide that explanation, we will have reason to be dissatisfied with those accounts.

And yet, rather than a damaging objection, I view Archer’s contribution as helpful in highlighting an important requirement for all vice-based accounts. Vice-based accounts must be supplemented with an accompanying theory of virtue and vice. To be told that someone acted in a particular way because they are vicious, only to then be told nothing about what it means to be vicious, would indeed be unhelpful. But there is no reason to think that a supporter of a vice-based account will stop there. Instead, they can appeal to a specific understanding of what virtues and vices are like, and of how these traits relate to one another. Once that additional level of detail has been provided in the form of a theory of virtue (and of vice), there is no reason to suspect that vice-based accounts will face a particular challenge in terms of their explanatory adequacy.

The worry here is similar to one that Paul Formosa (2013: esp. 240, 245) raises for any proposed theory of evil. One of Formosa’s suggested “theoretical virtues” of a successful theory is that it be “pitched at the right level of generality”, such that it is not problematically vague. And one way of failing to meet this requirement is by failing to provide sufficient explanation for terms that are key components of the theory. In the same way that providing more detail on the underlying theory of virtue (and vice) can help vice-based accounts to avoid concerns about explanatory adequacy, so too will this be necessary to avoid Formosa’s related concern about being theoretically unsatisfying.

The potential for vice-based accounts of evil to avoid concerns about explanatory adequacy by providing detail on their underlying theory of virtue can be illustrated by looking again at existing accounts. Garrard’s account, for example, does not merely say that we should understand evil persons as inversions of morally virtuous persons. Instead, Garrard goes further by committing to a specific theory of moral virtue. It is this commitment to a specific virtue theory that leads Garrard to propose an account of evil on which evil agents psychologically “silence” important considerations that tell against acting wrongly. Whatever else we think of this account, it provides exactly the sort of explanation that is demanded by the present objection. When asked why an evil person can perform actions that others would find morally repelling, Garrard’s account explains this by saying that evil persons psychologically “silence” important moral reasons to refrain from so acting (where those moral reasons are themselves metaphysical silencers for the reasons that motivate the evil person).

It is for this reason that Archer (2016: 205) accepts Garrard’s account as being able to avoid this objection. However, Garrard is not alone among proponents of vice-based accounts in saying more about her underlying view of the nature of virtue and vice. Barry (2013: 57–58) commits to the standard view that moral virtues are character traits and that character traits are “multitrack dispositions”, where this involves more than simply consistently acting in certain ways. It also involves dispositions to be motivated in certain ways, and to experience emotions in certain ways. For example, on this view, the cluster of related dispositions that constitute the character trait of honesty might be thought to include not only acting in non-deceptive ways, but also being motivated to avoid deception, and being disposed to experience certain emotions, such as experiencing shame when reflecting on past deceptive behaviour or experiencing indignation when witnessing the deception of others.

Importantly, Barry (2013: 57–58, 78–79) accepts this account not only for virtues but also for vices. That is, according to Barry, vices are also multitrack dispositions that are partly constituted by emotional dispositions and motivations towards certain ends or targets. And, as we shall see in Sect. 4, Barry also offers proposals on how to determine which vices an evil person would possess. In this way, Barry goes beyond merely saying that an evil person is one who is extremely vicious. He also commits to an account of vice that has the potential to explain why evil people can be expected to act in problematic ways (because being vicious involves, among other things, an agent’s motivations and emotional responses).

The lesson here is that, in order to be explanatorily informative, vice-based accounts of evil must be supplemented with an underlying theory of virtue and vice. By engaging with virtue theory, it is possible for supporters of vice-based accounts, including Garrard and Barry, to respond to this first objection.Footnote 7 And yet, the need to commit to a particular virtue theory opens up a range of further questions. Supporters of vice-based accounts need to decide which claims to accept or reject regarding the nature and structure of moral virtues, and the claims they endorse will determine, at least in part, the plausibility of their resulting account of evil. There are, therefore, important decisions to be made at the level of virtue theory.

How might these decisions be made? My strategy in the remainder of this paper will be to consider two further objections to vice-based accounts and to argue that it is possible to respond to those objections by committing to an underlying theory of virtue and vice that has certain features. By drawing on insights from work in virtue and vice epistemology, it is possible to avoid the objections and work towards a successful vice-based account of moral evil.

4 Objection 2—vice is not sufficient for evil

Vice-based accounts of evil define evil by reference to the concept of vice. But not everyone who is vicious ought to be considered evil. An account of evil that included anyone who was even slightly unkind, or slightly dishonest, would be too inclusive to be tenable. For this reason, it is necessary to provide a way of distinguishing those who are merely vicious from those who are truly evil.

An obvious move here is to say that evil requires not mere vice but extreme vice. This would explain why being slightly unkind, for example, is not sufficient to count as being morally evil. But a problem remains. Consider a standard list of candidate moral vices: dishonesty, cruelty, intemperance, immodesty, cowardice, injustice, stinginess, indifference, disloyalty, and so on. It is not obvious that someone who possesses an extreme form of one of these traits will thereby plausibly count as being morally evil. Someone who is extremely immodest, or extremely dishonest, for example, possesses a moral vice to an extreme degree, but this does not seem sufficient for moral evil.

There are two ways of conceptualising the problem that this objection raises for vice-based accounts. One is that the objection highlights that being extremely vicious is not (always) sufficient for moral evil. The person who possesses an extreme form of dishonesty (for example) is extremely vicious, but is not thereby morally evil. On this reading, the challenge for vice-based accounts is to provide an explanation for when (and why) being extremely vicious suffices for evil, and when it does not. A slightly different reading of the objection is that it highlights the possibility that possessing a vice to an extreme degree is not (always) sufficient to make one extremely vicious. On this reading, it could be maintained that being extremely vicious is sufficient for evil, but the challenge is then to explain which vices have the potential to make us extremely vicious and which do not. Whichever reading we prefer, the worry is that vice-based accounts owe an explanation of why the extreme possession of vice is sufficient for evil in some cases but not in others.

This worry is anticipated by Barry, whose own account explains evil in terms of extremely vicious character traits. Barry’s (2010: 29–30) response to the worry involves appealing to a distinction taken from Thomas Hurka, according to which a character trait can be extreme in either its “intensity” or the “value of its object”. Barry builds on this distinction and specifies that being evil is a matter of having character traits that are both extremely intense and have extremely dis-valuable objects.

For example, Barry (2010: 29–30) suggests that the object of malevolence is “some state of affairs in which undeserving persons are harmed”. If someone is intensely motivated to bring about such a state of affairs, then they will count as being extremely malevolent. The intensity of the motivation, combined with the (presumably) extremely dis-valuable object of this character trait, means that extreme malevolence suffices for evil on Barry’s account. In this way, Barry seeks to accommodate the idea that extreme vice is not always sufficient for evil. Instead, Barry (2010: 29–30, and see also 36) proposes that only “the morally worst sort of vices” suffice for evil, and only when they are possessed with an extreme intensity. The morally worst vices are those vices with extremely dis-valuable objects.Footnote 8

There are two problems with this response. First, there is the problem that some vices appear not to have a dis-valuable object, and yet at least some of these vices may be sufficient for evil when possessed to an extreme degree. For example, there are vices that appear to have no obvious object or target at all. Indeed, this seems to be true in some of the cases that Barry discusses as candidate examples of moral evil. When diagnosing the killer Robert Harris as evil, Barry (2010: 31) focuses on Harris’s “indifference” and “callousness”. But it is not obvious that either indifference or callousness have extremely dis-valuable objects. Rather, they seem to involve a lack of any object or target, and it is the implied lack of important positive objects (such as a concern for other people) that makes these traits morally objectionable. It is difficult to see how Barry’s strategy for responding to this objection can allow us to accommodate the possibility that vices with no clear target (such as indifference) can suffice for evil when possessed to an extreme degree.

It is also worth noting examples of vices that do have an obvious target, but where that target does not appear to be dis-valuable. For example, the target of cowardice is plausibly self-protection or the avoidance of risk. This is often a valuable aim, which suggests that the explanation for cowardice’s vice-status is not to be found by focusing on its target or object. Perhaps it is not clear whether even extreme cowardice could be sufficient for evil. Still, cowardice is at least widely accepted as a moral vice, and it is difficult to see how this can be accommodated by a theory of vice like the one endorsed by Barry, according to which vices involve dis-valuable objects or targets.

Difficulties with the claim that vices are identified by their dis-valuable objects (and that moral evil requires vices with extremely dis-valuable objects) is, I believe, the most pressing problem for Barry’s response. A second problem is that Barry’s approach leaves us with no real guidance on how to determine whether the object of any given vice is dis-valuable enough such that extreme possession of that vice suffices for evil.

Suppose we focus for now on only those vices that do plausibly have dis-valuable objects. We still require a method for determining what makes one object worse than another, and for determining when an object is sufficiently dis-valuable for moral evil. Barry’s suggestion of malevolence, with its aim of causing harm to others, is a plausible candidate as a vice with an extremely dis-valuable object. But many other cases are less straightforward. For example, perhaps the object of dishonesty is to deceive others. This does seem like a dis-valuable object, but is it dis-valuable enough for dishonesty to count among the “morally worst” of vices? Of course, it would be unreasonable to demand a comprehensive ranking and discussion of the target of each candidate moral vice. But without at least some guidance on how to determine which vices are among the “morally worst”, we will be unable to develop a vice-based understanding of moral evil along the lines suggested by Barry.Footnote 9

I want to propose an alternative strategy for accommodating the idea that not all vices (even when possessed to an extreme degree) are sufficient for evil. Supporters of a vice-based account of evil would benefit from endorsing a virtue theory that draws upon key insights from virtue epistemology.

In Virtues of the Mind, Linda Zagzebski (1996) proposes what has become one of the most influential contemporary theories of intellectual virtue. The aspect of Zagzebski’s virtue theory that (I will argue) is relevant to the present discussion concerns the distinction between proximal motivations and fundamental motivations.Footnote 10 According to Zagzebski, all intellectual virtues are partly constituted by their own specific proximal motivation. For example, the proximal motivation of intellectual thoroughness might be the motivation to closely examine the evidence. The proximal motivation of inquisitiveness might be the motivation to ask questions. In addition to this, Zagzebski (1996: 166–168) proposes that all intellectual virtues share in the same fundamental motivation. This is the motivation for “cognitive contact with reality” or, to put it another way, the motivation for epistemically valuable states such as knowledge, true belief, and understanding. For every intellectual virtue, the unique proximal motivation is grounded in, and explained by, the underlying motivation for epistemically valuable states such as knowledge (or “cognitive contact with reality”).

Zagzebski (1996: 269) provides several examples of how the proximal motivations of specific intellectual virtues are grounded in this shared fundamental motivation:

So the aim of open-mindedness is to be receptive to new ideas and arguments even when they conflict with one’s own in order to ultimately get knowledge. The aim of intellectual thoroughness is to exhaustively investigate the evidence pertaining to a particular belief or set of questions in order to ultimately get knowledge. The aim of intellectual courage is to defend one’s belief or a line of inquiry when one has good reason to be confident it is on the right track, and to fearlessly answer objections from others in order to ultimately get knowledge.

The possibility of distinguishing between proximal and fundamental virtuous motivations ought to be of interest to those who would develop a vice-based account of moral evil. It provides the resources for responding to the present objection that extreme vice is not (always) sufficient for evil.

The first step in this strategy is to recognise the following: Failures of the fundamental epistemic motivation are plausibly worse, in terms of one’s intellectual character, than are failures in the proximal motivation of any specific intellectual virtue. For example, a significant failure in the drive to be receptive to new ideas tells against someone’s open-mindedness. This suggests a lack of one important intellectual virtue and so can be viewed as a failing in the person’s intellectual character. However, a failure in the fundamental epistemic motivation for “cognitive contact with reality” strikes at the very heart of a person’s intellectual character. According to Zagzebski’s virtue theory, this fundamental motivation is common to all intellectual virtues. For this reason, a significant failure of the fundamental motivation calls into question a person’s intellectual character in a much deeper sense than does the failure of any proximal motivation. It undermines the possession of any intellectual virtue.

Vice-based accounts of moral evil need to explain why some cases of extreme vice are sufficient for evil and some are not. Zagzebski’s distinction between fundamental and proximal motivations provides the resources for meeting this challenge. What is required is to adopt a theory of moral virtue that mirrors Zagzebski’s theory of intellectual virtue, at least in the sense that it accepts a motivational structure that distinguishes between these two types of virtuous motivation. We can then say that vices that consist of extreme failures or inversions of fundamental moral motivations are sufficient for moral evil, while vices that consist of extreme failures or inversions of mere proximal motivations are not sufficient for moral evil.

What might such a theory look like in practice? Importantly, different vice-based accounts of evil can be generated depending on which motivations are proposed as being either fundamental or proximal in the moral sphere. Theorists will have to decide which combination of motivations results in an overall virtue theory that is most plausible. I will not attempt to defend a position on that broader question here. However, it will be helpful to provide a model of one way of developing a theory of moral virtue with this motivational structure, and to highlight the vice-based account of moral evil that would result.

Suppose we accept the following two fundamental motivations for the moral sphere: the motivation of kindness (to protect and promote well-being) and the motivation of justice (to ensure fairness).Footnote 11 Mirroring Zagzebski’s theory of intellectual virtues, we would then say that each moral virtue has its own unique proximal motivation, but that these proximal motivations must be grounded in one of the two fundamental moral motivations. For example, honesty might involve the proximal motivation to avoid deception, but it will only be morally virtuous when that motivation is grounded in either the motivation to protect and promote well-being or the motivation to ensure fairness. Similarly, generosity might involve the proximal motivation to provide time and resources to others, but it will only be morally virtuous when this motivation is grounded in considerations of protecting well-being or ensuring fairness. And so on for other candidate moral virtues.Footnote 12

By endorsing an underlying theory of virtue with this motivational structure, it becomes possible for supporters of a vice-based account of moral evil to respond to the present objection. Each moral virtue shares in at least one of the fundamental moral motivations, while being distinguishable by its own unique proximal motivation. Moral vices involve inversions of these motivations. An extreme inversion of a proximal motivation may be very bad indeed in terms of what it implies about a person’s moral character, but it will not suffice for moral evil. Instead, moral evil requires an extreme inversion of one of the fundamental moral motivations.

Using the illustrative model set out above, we can see how adopting such a virtue theory has the potential to generate plausible verdicts when applied to an account of moral evil. For example, it becomes possible to explain why even extreme immodesty and extreme dishonesty do not suffice for evil. Failures or inversions of the proximal motivations specific to these vices do not necessarily imply a failure of one of the fundamental moral motivations of kindness and justice. On the other hand, extreme malevolence and extreme injustice do suffice for evil, because such vices do necessarily involve failures of at least one of the fundamental moral motivations. Malevolence involves a failure of the motivation of kindness (to protect and promote well-being) and injustice involves a failure of the motivation of justice (to ensure fair outcomes).

In this way, it is possible for vice-based accounts of moral evil to provide a principled answer to the question of when extreme vice is (and is not) sufficient for evil. Importantly, it will be open to different theorists to make different decisions regarding which motivations are considered as fundamental or merely proximal in the moral sphere. But by importing this general motivational structure into our virtue theory (a motivational structure first proposed within virtue epistemology), it becomes possible for vice-based accounts of evil to respond to the present objection, and to explain why even extreme vice is not always sufficient for evil.

5 Objection 3—vices are not (all) inversions of virtue

When proposing vice-based accounts of moral evil, contemporary theorists often appeal to one supposed benefit of this approach. It is claimed that vice-based accounts can explain the so-called “mirror thesis” (Barry, 2009). The mirror thesis is the intuitive idea that morally evil people are, in some sense, the mirror image of morally excellent people (or “moral saints”). If we are inclined to accept the mirror thesis, then vice-based accounts offer a ready explanation: (a) moral evil is understood by appeal to the concept of vice; (b) moral excellence is understood by appeal to the concept of virtue; (c) vices are inversions or mirror images of virtues; and so (d) evil people can be understood as inversions, or mirror images, of morally excellent people.

This sort of argument is common among those seeking to provide support for vice-based accounts of evil. For example, Barry (2010: 37–40; see also Barry, 2009 and 2013: 67–71) argues that his account of evil “supports a literal reading of the mirror thesis” and (if we accept that thesis) this provides us with reason to prefer his account. Similarly, Haybron (2002: 275) argues that his account helps us to understand the way in which the idea of an evil person:

belongs to a broader characterology that is itself plausible. The evil character defines one pole of a moral continuum that has the saint at its other pole… The relevant moral space is purely aretaic: from most vicious to most virtuous.

Barry and Haybron both aim to support their accounts by showing that they can explain the supposed “mirroring” of moral evil and moral excellence. Vices are mirrors or inversions of virtue, and so vice-based accounts are well-placed to explain this idea. One way of challenging this positive argument in favour of vice-based accounts is by denying the truth (or perhaps the usefulness) of the mirror thesis itself (Calder, 2015). I want to explore a different worry. When setting out their explanations, supporters of vice-based accounts are making an assumption. They assume that, just as evil is the inversion of excellence, vices are inversions of virtues. If that assumption is false, then vice-based accounts will be less well-placed to explain the mirror thesis than has been supposed. What was thought to be a strength will turn out to be a weakness.

The assumption that (all) vices can be understood as inversions of virtue is challenged by recent work in vice epistemology. Charlie Crerar (2018) argues that epistemologists have wrongly assumed that intellectual vices can be understood as inversions of intellectual virtues. If Crerar is correct, and if we can apply a similar argument to the moral sphere, this will be bad news for vice-based accounts of moral evil. To be clear, the worry here is not simply the loss of one positive reason to endorse vice-based accounts. Rather, the plausibility of the mirror thesis would become an active problem for those accounts, because the elements that vice-based accounts focus on when thinking about morally evil people (the vices) would not necessarily be inversions or mirrors of what they focus on when thinking about morally excellent people (the virtues).

Crerar considers two ways in which vices might be thought to mirror virtues. Assuming virtues involve valuable motivations, vices can mirror this either by actively involving bad motivations or simply by lacking the good motivations. Crerar (2018: 755–761) refers to the idea that vices involve bad motivations as the “presence conception” of vice and the idea that vices involve a lack of good motivations as the “absence conception” of vice. In the epistemic sphere, epistemic malevolence is an example of a vice that actively involves a bad motivation (the motivation to deprive others of knowledge) while epistemic laziness is a vice that involves the absence of a good motivation (the motivation to seek out knowledge). Crerar accepts that vices of these two types can be viewed as “inversions” of virtues but argues that there are examples of vice that fit neither of these conceptions. This is taken to show that not all vices can plausibly be understood as inversions of virtues.

Crerar’s primary focus is on intellectual vice, and this is reflected in the examples he provides. Given my focus on moral vice (and, ultimately, on moral evil), I want to translate Crerar’s argument into the moral sphere and consider how best to respond. Here is a moral analogue of an example provided by Crerar to show that some vices cannot be captured by either the “presence” or the “absence” conception of vice:

Nick is a politician who genuinely desires to promote well-being. However, Nick’s privileged background encouraged him to believe that the best way to achieve this is to listen only to people who “like him, had received a high level of formal education, had studied the relevant issues, and could articulate their position through reasoned and dispassionate argumentation”. As a result of this blinkered approach, Nick ends up supporting policies that “disproportionately harm members of a marginalized group”.Footnote 13

Something seems to have gone awry with Nick. His prejudice concerning who is worth listening to causes him to support policies that are harmful. It is plausible to think that Nick is morally vicious in some sense. However, as the example is presented, Nick does not seem to possess a morally bad motivation. He is not actively trying to harm anyone, for example. And he does not completely lack a morally good motivation either. He does genuinely desire to promote well-being. If Nick is morally vicious, and if possessing bad motivations or lacking good motivations are the only ways in which vices can mirror virtues, then we have an example of moral vice that is not a mirror of moral virtue.Footnote 14

To be clear, the claim here is not that Nick is evil. Rather, the example of Nick is supposed to demonstrate that vice cannot (always) be understood as an inversion of virtue. Vice-based accounts of evil have tended to assume that vices are inversions of virtues in order to help them explain the mirror thesis. For this reason, examples like Nick are potentially problematic for vice-based accounts of evil.

In response to this worry, we ought to turn once more to the underlying theory of virtue and vice that we endorse. In particular, we ought to reflect on the structural aspects of virtuous motivations. While moral virtue requires being motivated towards valuable ends, not just any motivation with the correct content is sufficient for virtue. Virtuous motivations also require certain structural features. First, virtuous motivations must be sufficiently persistent, as opposed to being sporadic or fleeting. Second, virtuous motivations must be sufficiently strong, at least to the extent that they are capable of prompting the agent into action. And third, virtuous motivations must be sufficiently robust in the sense that they are not easily overridden by competing considerations.Footnote 15 For example, my motivation to ensure fairness will not be a virtuous motivation if it is problematically sporadic, or if it is too weak to ever influence my actions, or if it is consistently being overridden by my competing motivation to secure power for myself and my political party.

Reflecting on the structural aspects of virtuous motivations reveals an additional way in which vices can mirror or invert moral virtues. If we focus only on the content of virtuous motivations, then we will assume that vices can only invert virtues by involving an opposite vicious motivation (as in Crerar’s “presence conception” of vice) or by completely lacking the virtuous motivation (as in Crerar’s “absence conception” of vice). Instead, we ought to recognise that vices can also invert the structural aspects of virtuous motivations. That is, vices can involve motivations that have the correct content, but which are problematically fleeting rather than persistent, problematically weak rather than strong, or problematically fragile rather than robust. This is an additional sense in which vices can be “inversions” or problematic “mirrors” of virtues.

The possibility of structural inversions helps us to see why examples like Nick can indeed be viewed as instantiating an inversion of virtue. Would Nick have maintained his belief that only people like himself are worth listening to if his motivation to promote well-being was sufficiently persistent, strong and robust? Or does Nick’s failure to recognise the negative impact of his policies reveal that he does not, in fact, care sufficiently about well-being? Does it reveal that his motivation to promote well-being is being overridden by his competing desire to secure power for himself, or to feel good about himself? If Nick’s failure is explained in these ways, then Nick is instantiating an inversion of a virtuous motivation. He has the structural opposite of proper, virtuous motivation, which, in addition to having the correct content, must be sufficiently persistent, strong, and robust.

Indeed, it is only when Nick’s failure is explained in this way that it is plausible to make the charge of moral vice. Suppose Nick really is persistently, strongly and robustly motivated to promote well-being. It will still be possible for Nick to fail, but his failure will instead be the result of moral (bad) luck. In such cases, it becomes less clear that we have evidence of moral vice. To the extent that Nick is plausibly (morally) vicious, his failure must be explained by some aspect of his motivational profile. In some cases, the problematic aspect of a vicious motivational profile will be either the presence of a bad motivation or the absence of a good motivation. But reflecting on the structural aspects of virtuous motivations helps us to identify a different way in which a motivational profile can be vicious, and an additional way in which vicious motivations can “invert” virtuous ones.

I have suggested that it is helpful to recognise structural inversions of virtuous motivations, in addition to what can be called content inversions of virtuous motivations. This allows us to see that examples of the sort provided by Crerar do involve motivational profiles that invert aspects of virtuous motivations. This is important for our purposes, because the idea that vices are inversions of virtues has been assumed by supporters of vice-based accounts of evil. This is what allows vice-based accounts to provide an explanation for the intuitive idea that morally evil persons are, in some non-trivial sense, the mirror image of moral saints. By providing the resources to support this idea at the level of virtue theory, the case in favour of vice-based accounts of evil (and in favour of vice-based explanations of the mirror thesis) is strengthened.

In addition to allowing for an explanation of the mirror thesis, the distinction between content and structural inversions also allows us to identify different forms of moral evil. In some cases, evil involves being actively motivated against what is morally valuable. For example, we can imagine someone who is motivated to harm others, or to maintain an unjust state of affairs. Such an individual will possess a content inversion of a fundamental moral motivation and, if the inversion is extreme enough, they ought to be considered morally evil. But it is possible to be evil without being evil in this way. In some cases, lacking a moral motivation with sufficient persistence, strength and robustness can also suffice for evil, even if the individual is not actively motivated towards anything morally problematic. Such an individual will possess a structural inversion of a fundamental moral motivation and, again, if the inversion is extreme enough, they ought to be considered morally evil.

Consider again the possibility that (extreme forms of) vices such as indifference or callousness can be sufficient for evil. As noted above, Barry makes the plausible claim that these vices can suffice for evil, but this claim is difficult to accommodate when using an approach that focuses on the disvalue of the objects or targets of moral vices. The distinction between content and structural inversions allows us to better understand such examples. Even in cases where someone possesses a minimal level of concern for (some) others, it will still be possible for their indifference to suffice for moral evil. If their concern for others is extremely weak, fleeting or fragile, to the extent that they are willing to allow extreme suffering in order to advance their own interests, then it may still be possible for them to be rightly considered morally evil.

In some cases, evil consists in the active pursuit of bad ends, but it is also sometimes a matter of being problematically uncommitted to the pursuit of good ends. This explains why evil can be revealed in the turning of a blind eye to grave injustice, or in the willingness to allow great suffering in order to secure one’s own self-interest or self-advancement. The idea of extreme structural inversions of virtue allows vice-based accounts of evil to accommodate this possibility, and so further strengthens the plausibility of the approach.Footnote 16

6 Implications

I have argued that vice-based accounts of moral evil can respond to three key objections. The first objection (about explanatory inadequacy) reveals the need for vice-based accounts to engage with virtue theory and make decisions about their preferred understanding of virtue and vice. Reflection on the second and third objections reveals the benefits of endorsing a virtue theory with certain features. A virtue theory that distinguishes between fundamental and proximal motivations allows vice-based accounts to explain when extreme vice is (and is not) sufficient for evil. A virtue theory on which vices can consist of either content or structural inversions of virtuous motivations allows vice-based accounts to accommodate the “mirror thesis” and to distinguish between active and passive forms of evil.

The above arguments recommend a vice-based account with the following features:

  1. (1)

    To be morally evil is to possess an extreme inversion of the fundamental motivations involved in moral virtues (rather than an extreme inversion of a merely proximal motivation).

  2. (2)

    This includes extreme inversions of the content of fundamental moral motivations (in the case of “active” evil) and extreme inversions of the structural aspects of fundamental moral motivations (in the case of “passive” evil).

As noted above, such an account can be developed in different ways, depending on which motivations are identified as being fundamental or proximal in the moral sphere. When presenting a model of such an account, I suggested the following:

  1. (3)

    Candidates for the fundamental moral motivations include the motivation of kindness (to protect and promote well-being) and the motivation of justice (to ensure fairness).

I have not argued for (3) in this paper, other than suggesting that this selection of fundamental motivations generates initially plausible verdicts about when extreme vice is sufficient for evil. A final decision on which motivations (and so which virtues) are fundamental in the moral domain needs to take into account a wider range of issues in virtue theory, and not only the issue of which selection results in a plausible account of evil. Instead, I have focused on arguing that (1) and (2) allow for a vice-based account that avoids important objections.

An account of moral evil based on (1) and (2) would mark a significant change from existing vice-based accounts. However, it is interesting to consider which of the claims argued for in this paper could be accommodated by those accounts.

Of the existing vice-based accounts, it is Barry’s account, with its focus on extreme vice, that is closest to the proposal developed in this paper. Both Barry’s account and point (1), above, require extreme vice in order for someone to be morally evil. As Barry (2013: 65) has argued, one of the benefits of this focus on extremity is that it allows us to mark a clear difference between those who are morally evil and those who merely very bad—the morally evil are those who possess “the morally worst sort of vices”.

However, the proposals in this paper require diverging from Barry’s preferred underlying theory of vice in at least two ways. First, as was noted in Sect. 4, Barry’s explanation of which vices are the “morally worst” relies on appealing to the extreme disvalue of the objects of those vices. I have argued that this way of identifying which vices are sufficient for evil is problematic, and that we should instead identify the morally worst vices as those involving inversions of fundamental moral motivations. This in turn requires a virtue theory that distinguishes between the proximal and fundamental motivations of moral virtues.

A second divergence from Barry’s underlying theory of vice concerns point (2), above. I have argued that extreme structural inversions can be sufficient for evil, even in cases where the evil person is not actively motivated towards a bad end, nor completely lacking in motivation towards some good ends. Instead, extreme failures in the strength, persistence and robustness of positive motivations can also suffice for evil. This is at odds with Barry’s (2013: 57–58) claim that vice (and therefore evil) requires more than lacking the right sort of positive motivation. For Barry, vices involve being actively disposed towards the bad.

It is worth noting one further benefit of endorsing (1) and (2), and so diverging from Barry’s underlying account of vice. Barry (2013: 14–15, 56, 60 and 88) considers Hannah Arendt’s much discussed description of Adolph Eichmann, one of the key architects of the Holocaust. According to Arendt’s diagnosis (quoted in Barry, 2013: 15), Eichmann was not driven by “insane hatred”. Barry (2013: 56) suggests that, if it is true that Eichmann lacked “especially virulent animus, precisely what we expect of an extremely vicious person”, then Eichmann will not be classed as evil:

If Hitler is malicious and hateful while Eichmann is merely selfish (as Arendt sometimes suggests), then Hitler’s vices are extreme in a way that Eichmann’s are not because Eichmann’s vices are not among the morally worst. (Barry, 2013: 60)

Endorsing a view of extreme vice on which extreme vice requires an extremely intense motivation towards an extremely dis-valuable object risks implying that Eichmann was not evil (at least given the description of Eichmann as motivated by extreme selfishness or thoughtlessness rather than extreme hatred). As has been noted elsewhere (Archer, 2016: 207), this implication will strike many as a mark against the account. Diverging from Barry’s underlying account of vice in the ways argued for in this paper provides the resources for avoiding this implication. Whatever else was true of Eichmann, it is plausible that any positive motivations he may have possessed (and which are mentioned by Barry, 2013: 14–15, 56) were problematically lacking in their strength, persistence and robustness. Given this, an account that allows extreme structural inversions of fundamental moral motivations to suffice for evil can generate the verdict that Eichmann was indeed morally evil. This provides further reason to diverge from Barry’s specific account of vice, even while the proposals in this paper are broadly sympathetic to Barry’s (2013: 56) more general “extremity thesis” on which evil consists in extreme vice.

Uncovering points of compatibility between the proposal in this paper and other vice-based accounts of evil is more difficult. While all vice-based accounts share in their basic appeal to the concepts of virtue and vice, other accounts either appeal to a very different understanding of the nature of virtue (such as Garrard’s appeal to a McDowellian account of virtue) or are committed to claims that are not obviously consistent with the proposal argued for here (such as Haybron’s claim that evil requires not extreme vice but consistent vice). However, I do want mention the possibility of combining aspects of (1) and (2) with Haybron’s account.

As noted above, Haybron argues that evil people “have no good side, but are consistently vicious”. As Haybron acknowledges, this account will appear overly demanding to some, as it requires that an evil person has no degree of any moral virtue. For example, it seems plausible that someone who is extremely cruel or unjust could be morally evil, regardless of whether they happen to also possess some small degree of modesty or honesty, or even generosity. This appears to be ruled out by Haybron’s account. Indeed, Haybron (2002: 270) suggests that it is conceivable that “even Hitler would not qualify as evil”.Footnote 17

Someone who wishes to avoid this implication but who is otherwise sympathetic to the claim that evil requires consistent vice could choose to adopt a virtue theory of the sort proposed in this paper. Instead of demanding that evil persons be consistently lacking in virtue across the board, it would be possible to amend Haybron’s account to demand only that evil persons consistently lack the most fundamental of virtues. On this amended view, the minimal possession of a less fundamental virtue, such as modesty or honesty, would not be enough to rule someone out as being morally evil. And, as we have already seen, adopting a virtue theory that distinguishes between fundamental and proximal motivations provides the resources for making the sort of distinction that this would require.

It is not clear to me that there is reason to prefer such an amended form of Haybron’s consistency account over the account proposed in this paper, and general worries for consistency accounts have been covered elsewhere (Barry, 2010). However, this discussion demonstrates one way in which those who prefer to stay more closely aligned with an existing vice-based account may benefit from adopting a virtue theory with at least some of the aspects proposed here. Of course, I have argued that accepting both (1) and (2) allows us to overcome significant objections, and so offers a promising way of developing a vice-based account.

Before concluding, I want to briefly raise one further issue. One interesting feature of vice-based accounts of moral evil is that they suggest analogous accounts of evil in other normative domains where it is common to talk of virtues and vices. Whatever the account says about the relationship between moral vice and moral evil, we can ask whether the same relationship holds, for example, between epistemic vice and epistemic evil, or between aesthetic vice and aesthetic evil.

For example, suppose we accept my current proposal for a vice-based account and say that moral evil consists in the extreme inversion of one of the fundamental motivations involved in moral virtue. An analogous account of epistemic evil will say that to be epistemically evil is to possess the extreme inversion of a fundamental motivation involved in epistemic (or intellectual) virtue. If we agree with Zagzebski, we will think that this fundamental epistemic motivation is the motivation for cognitive contact with reality. On such an account, even extreme inversions of the proximal motivations of specific epistemic virtues (such as intellectual rigour’s proximal motivation to exhaustively examine the evidence) will not suffice for epistemic evil. However, extreme inversions of the fundamental motivation for cognitive contact with reality will be sufficient to count as being epistemically evil.

An analogous account of aesthetic evil will say that aesthetic evil consists in the extreme inversion of whatever are the fundamental motivations involved in aesthetic virtues. Possible candidates for this role include the motivation for beauty and the motivation of “appreciation”.Footnote 18 While individual aesthetic virtues may all have their own specific proximal motivations, being an aesthetically evil person will require an extreme (content or structural) inversion of the motivations that are fundamental to an aesthetically virtuous life.

My aim in this paper has been to advance the development of a vice-based approach to moral evil. Given this, I will not attempt to examine the plausibility of different ways of developing accounts of non-moral evil, including epistemic evil or aesthetic evil.Footnote 19 Instead, my reason for mentioning these possible forms of non-moral evil is to highlight the potential for further important work at the intersection of theories of evil and theories of virtue and vice. To the extent that we are interested in the nature (or even the very possibility) of people who are epistemically or aesthetically evil, we will have reason to engage with recent developments in virtue (and vice) epistemology and aesthetics. By defending the plausibility of a vice-based approach to moral evil, and arguing for the importance of reflecting on our underlying theories of virtue and vice, I hope to have played a role in encouraging such future work.

7 Conclusion

I have argued that vice-based accounts of moral evil can respond successfully to important objections. This requires thinking carefully about decisions made at the level of virtue theory. In particular, I have argued that vice-based accounts of evil can benefit from accepting a virtue theory that distinguishes between fundamental and proximal virtuous motivations, and which understands vices as involving either content or structural inversions of those motivations. In this way, I hope to have supported the general strategy of understanding evil in terms of vice, and to have provided guidance on how best to develop such an account. In future work, it will be important to consider which motivations ought to be viewed as fundamental in the moral domain, and whether a vice-based approach can be successfully expanded to account for the possibility of evil across other normative domains.