1.1. Virtue and Absolute Prohibitions

Peter Geach worried that Aristotelian virtue ethics cannot generate absolute prohibitions. He claims the following:

Geach's thought is that virtuous agents would sacrifice their ‘hard-won integrity of soul’ in the face of some great (consequential) good. If virtuous agents can prevent harm to others, then they should not be concerned with getting their hands dirty, for that harms only them. What should a secular virtue ethicist say to Geach? Hursthouse has given a preliminary response in claiming that there are some acts that ‘connote depravity’, but because Vogler questions how such a view could be grounded (Reference Vogler, Hoffman, Müller and Perkams2013: 255), I develop Hursthouse's idea below.

Absolute prohibitions in secular virtue ethics are grounded in the fact that there are some actions that are impossible to do virtuously or that directly manifest vice. Foot suggests such an approach when she says that ‘if the frequently unchallengeable description “torture” applies to an action, then, whatever the circumstances, it is in my firm opinion morally “out”’ (2001: 78). She claims that some actions necessarily manifest injustice or cruelty, and that one ‘necessarily acts badly’ if one does them, but that it may be hard to know whether some action is in fact cruel or unjust (2001: 78).

However, both Foot and Hursthouse discuss absolute prohibitions only in passing (and somewhat gingerly). Foot's discussion occupies a few lines in a discussion of Mill, and Hursthouse speaks only of ‘the (possible) existence of a few absolute prohibitions’ (Reference Hursthouse, Jost and Wuerth2011: 39–40). One can see why Vogler is skeptical of the extent to which Anscombe's own colleagues and students have incorporated absolute prohibitions into their theories. But there are historical precedents from which contemporary virtue ethicists could draw inspiration.

Plato thinks that a just person will never embezzle money, take bribes, rob temples, steal, betray friends, betray his country, be untrustworthy in keeping oaths and agreements, commit adultery, disrespect his parents, or neglect the gods (Republic 442e–443a). Notoriously, Plato's just person refrains from these actions because they would disrupt the internal harmony of his soul—but notice that they have this effect because those actions are unjust. They bring injustice into the soul. At any rate, Plato has no qualms in listing actions that a just person would never do.

Aristotle is less forthcoming with examples, but he provides some. ‘Not every action or feeling admits of a mean’, he writes, ‘for some have names immediately connected with depravity, such as spite, shamelessness, envy, and, among actions, adultery, theft, homicide’ (Nicomachean Ethics 1107a8–12). Some actions are by their nature unvirtuous and so cannot be performed virtuously (Irwin Reference Irwin, Hooker and Little2003: 111, 121). One cannot do adultery well.

It is Cicero, however, who says it best: ‘There are some acts either so repulsive or so wicked that a wise man would not commit them, even to save his country. Posidonius has made a large collection of them; but some of them are so shocking, so indecent, that it seems immoral even to mention them. The wise man, therefore, will not think of doing any such thing for the sake of his country; no more will his country consent to have it done for her’ (On Duties I.159). Cicero thinks that saving one's country is one of our most important moral concerns, so his position is that there are some consequences that cannot outweigh the wrongness of the action.

This is not a bad start, but we do not need Posidonius's list to find other actions that cannot be done virtuously (see also Tsu Reference Tsu, Birondo and Braunt2017). Some of these actions are mundane and whose performance is shocking precisely because they violate basic standards of decency: cutting in line; trying to get onto the bus or subway while people are still coming out; singing along at the opera; posting someone else's sex tapes or nude photos without consent; randomly shoving strangers on the street; putting graffiti on artworks or archaeological sites. In addition, there are actions that occur more regularly but that are nonetheless unvirtuous: bullying; cheating; fearmongering; corruption; whining; taking credit for others’ work; plagiarizing; knowingly misquoting texts; fabricating evidence; slander; discriminating on the basis of race, gender, sexuality, and using related epithets and slurs; driving while drunk or chemically impaired. Then there are those actions that are horrifying and extreme paradigms of vice: rape; defiling sacred spaces; genocide; abusing children; abduction; engaging in sexual activity with minors. I could continue with these lists (Anscombe would add lying, knowingly punishing the innocent, and murdering, i.e., intentionally killing the innocent), but the point is to draw attention to the many action prohibitions that most of us have so thoroughly internalized that they rarely figure in ethical discussion or are considered in moral deliberation, precisely because they are not hard cases.

The aforementioned actions directly connote vice. Some of these actions are morally reprehensible and manifest an especially vicious character, whereas other actions are more trivial but nonetheless show a defect in character. While it is jarring to consider these cases side by side, we should remember the scope of virtuous deliberation and action. Prohibitions show up all over the place—not just in serious moral dilemmas—and many of them are so obvious we do not often think about them. Moreover, we feel confident in determining that people who do these things are not good people without needing to make reference to the consequences of the action. What explains why we feel justified in making such judgments, and what is it that makes those actions impermissible? There is no single thing because of which all of those actions are impermissible—nor should we expect to find one. What makes singing along at the opera wrong depends on the inconsiderateness of making noise in conjunction with the conventions of concert-going, where the understanding is that people have paid money to hear the people on stage and not those sitting next to them. This is very different from what makes rape wrong, or why a lack of academic integrity undermines the project of scholarship. Thus we should not expect there to be a single ‘prohibition-making’ feature of all actions that directly connote vice.

But what makes such actions manifestly vicious in general? We can say something more about what these actions have in common that makes them indefeasibly wrong as opposed to merely wrong in some respect by considering the structure of virtuous and vicious action. Virtuous actions have noble ends and demonstrate deliberative excellence, so what these prohibited actions have in common is that either (a) there is no way to give a virtuous justification of how they would achieve some good, or (b) their ends and natural consequences are things that a virtuous person avoids or prevents. Thus the error consists in choosing actions that are either reliably bad ways of achieving an end or require taking an end that is unvirtuous.

For example, cutting in line at the airport is an unvirtuous means because it does not make the plane leave any faster and it is unfair and disrespectful to the other passengers; thus, it reliably involves vice and reliably fails to achieve any good. By contrast, torture manifests vice because it necessarily involves intending intense pain and distress to another sentient being; thus, the action is prohibited because the end of the action is vicious. A virtuous agent would respond to someone who performs these kinds of actions by saying either, ‘But that's not how you go about getting your end!’ or, ‘But you shouldn't want to achieve that!’ and would then proceed to give a further explanation based on the particular wrong-making features of the case, highlighting either the deliberative failure in instrumental reasoning or the teleological failing of wanting to achieve the end in the first place. So, in general, actions manifest vice when some part of the structure of the action is corrupt, but what explains the corruption and the degree of moral seriousness will depend on the particulars of the action.

Of course, we could fabricate examples in the philosophy laboratories where it is claimed a virtuous person would get their hands dirty and take a means that would not normally bring about a virtuous end: only by posting nude photos of your ex while singing along to Die Walküre will we prevent nuclear war. But taking such thought experiments seriously and letting them dictate the shape of one's normative theory displaces obvious moral truths and distorts ethical theorizing (more on this in §3.3). Ethics should not be about what could be right in some grotesque possible world where rape prevents genocide—it should be about what is right in our ethical world, and it is ethically important that, in our experience, rape never prevents genocide. This is why rape is never the means a virtuous agent would take to achieve some good, for rape involves intending evils, and no good would in fact be achieved. These actions are thus impermissible, not by logical necessity but by facts about how our world works.

It might be thought mysterious that there could be actions that directly manifest vice, but there is an epistemic litmus test for intuitively identifying such actions. An action directly manifests vice if, upon hearing that somebody performed it, we do not need to ask why they did it before we are justified in condemning the action. If someone confesses to abusing children, we do not wait to find out if this was part of some greater plan for good—we know that the person has done something vicious. (In normal circumstances, we have good reason for making judgments about the character of the person who performs such actions too—only a vicious person would do such an action. There are harder cases where the person may be hallucinating or acting under hypnosis, such that we should not infer anything about character from the action performed. But these are unusually exculpating circumstances because the normal connection between character and action on the basis of which we would reasonably make inferences about someone's character is severed.) Consider, by contrast, our reaction when we hear that forty people have been killed in an Iraqi village. We are upset, but it makes a great deal of difference whether this was avoidable, or the result of recklessness or negligence or political grandstanding, or part of an ingenious strategy to minimize civilian casualties in war. Killing does not directly manifest vice; genocide does. Thus actions that are characteristically vicious but sometimes required differ from actions that are characteristically vicious and categorically prohibited. The former permit of justifications that can take account of consequences; the latter do not permit such justifications. Determining act-types that directly manifest vice is the main way that virtue ethicists can account for absolute prohibitions.

1.2. Virtue and Roles

The second source of rules and absolute prohibitions is generated from our institutional roles and practical identities. These rules will not be universal, as not everyone occupies the same roles, but they are binding in the same way as absolute prohibitions: there are some things people cannot do nor should they deliberate about doing if they occupy certain roles, regardless of the good consequences that might be produced. To be sure, some roles are incompatible with virtue (guards at concentration camps) and some roles likely cultivate values antithetical to virtuous agency (state executioner). Nevertheless, people (even virtuous people) often have jobs and occupy roles within good institutions. While role duties are conditional on your occupying a position, everybody at some point occupies a role and will be bound by the rules of that role (Annas Reference Annas2015: 8–9, 12). Some roles, like soldiers, psychiatrists, police officers, doctors, lawyers, and judges, have well defined institutional rules and regulations, while others, like chefs and tradespeople, have strong institutional norms without having a formal code of conduct. Just because kitchen workers do not have a lawyer's rulebook does not mean that chefs are not breaking an institutional rule when they spit in your soup.

But do virtuous agents respect these institutional rules and take them as categorical? Surely, it will be argued, someone who truly understands what is morally important will not think of themselves as limited by the rigid general rules of their profession in the face of the particular challenges they will face. While there are certainly cases of good people valiantly breaking the rules of their role—most clearly, soldiers disobeying cruel orders—most sophisticated institutions have provisos built in for when the general rules do not apply. This is why the military and hospitals have ethics boards composed of people from that role who can understand the unique demands of the situation.

It is a mistake, however, to think that virtuous people who occupy roles never take their role duties to be categorical. Firstly, because the successful functioning of institutions often depends on social trust and the expectation that certain rules will not be broken, agents who find themselves occupying roles within these institutions should understand the far-reaching consequences of breaking these rules. You do not want your doctors deliberating about whether your organs would be better used elsewhere. Secondly, the size and compartmentalization of institutions often makes it such that the person who occupies the role is not in an epistemic position to understand the situation adequately. Soldiers and service people are often not given all the intelligence that informs their orders. In such situations, well intentioned people can make serious errors because they did not defer to the rules their role had established for them. Finally, while cases of people valiantly defying the rules make for good movies, the truth of how virtuous people act in difficult roles is more mundane. A good lawyer may perform her role best in getting her (guilty) client acquitted through an elaborate legal loophole, but then that same lawyer is in the best position to advocate a reform of that law to avoid similar results in the future (Hursthouse Reference Hursthouse, Farrelly and Solum2008). Thus, jumping to the conclusion that an institutional rule should be broken if it does not produce the right result in a particular case shows a deficient understanding of how one can act virtuously within the constraints of a role across time.

Moreover, a virtuous agent's roles are not limited by her formal institutional roles. Virtuous agents also have moral personalities and practical identities (Baril Reference Baril2016; Korsgaard Reference Korsgaard and O'Neill1996: 100–105). These practical identities, like institutional roles, bring with them expectations, commitments, and principles. A parent is not just somebody who has children, but somebody who adopts the values and responsibilities of parenthood. It would be a mistake to think of these responsibilities as optional simply because they are chosen. Note that some practical identities we currently think are optional may actually be required by virtue, for example, vegetarianism or philanthropy. In such cases, these identities would not be a separate source of moral rules apart from the demands of virtue.

Cicero, following Panaetius, takes practical roles (personae) as a source of our duties and shows how one's practical identities can generate different standards of conduct (Gill: Reference Gill1988). Marcus Cato the Younger had devoted his life to fighting Caesar and protecting the Roman Republic. Cicero thus argues that it was reasonable for Cato to kill himself rather than bend the knee, even though it was reasonable for others to accept Caesar's clemency, as their personae had not been shaped by Cato's strongly republican commitments (On Duties I.112). Similarly, Bernard Williams has argued for the necessity of particular actions being generated from an agent's practical identities, specifically from the shame that they would incur if they did not do certain things (Reference Williams1994: chapter 4).

These requirements are not universal, but neither are they optional. One cannot simply shed one's identities when they are inconvenient, and so in practice virtuous people will stick to the principles they have adopted. In Cato's case, this meant choosing death rather than living under an emperor, but for an animal rights activist this may mean causing offense to one's hosts rather than eating meat. That an animal rights activist's identity is developed voluntarily does not undermine the categorical force of the principles of that identity. This is not to say that one must never shed a practical identity. But while virtue does not require us to have certain practical identities, the values and principles of people who do have those identities are not disposable or flexible. Therefore, because virtuous agents develop such identities and are committed to them in a nonoptional way, practical roles will prohibit some actions categorically.

This does not rule out identities changing and being shaped from within. Consider the current diversity of Christian attitudes toward homosexuality or contraceptives. In such cases, people are working out what is central to Christian moral identity and what is not. Ethical demands from identity go beyond the demands of virtue, but should demands of identity conflict with demands of virtue, it is up to people who have that identity to decide whether the conflict is merely apparent and they can get around the conflict in a morally creative way, whether the identity can be changed, or whether the identity ought not to be adopted (as with institutional roles that are incompatible with virtue).

1.3. Virtue and Divinely Wise Commands

The easiest way to get absolute prohibitions is through revealed divine commands and secular virtue ethicists should consider the role that divine commands play in theistic virtue ethics. Divine commands are not brute deliverances from on high; they are enshrinements of divine wisdom and guides to help us live a better life (Adams Reference Adams1999: 255). They impose a structure of action and thought that is beneficial, guiding us to achieve good purposes and aim at good ends. Divine commands thus serve a variety of functions: they are warnings that prevent us from falling into common errors; they highlight areas of moral concern that we too easily neglect; they act as checks on courses of action we might otherwise be tempted to pursue. So we should not follow divine commands just because they are ordered by a cosmic authority—we should follow them because it's demonstrably good for us to do so.

Linda Zagzebski's exemplarist virtue ethics provides resources for filling out this account. On Zagzebski's view, what makes motivations good is their similarity to divine motives, because God is the ultimate moral exemplar (Reference Zagzebski2004: 284). Thus we can say that what makes divine commands good is that they flow from an omniscient, perfectly good moral exemplar, i.e. they are wise directives. For a secular virtue ethics, there will be no perfect and omniscient exemplar, but there are still moral exemplars. Another source of rules in virtue ethics is thus the prescriptions and advice of virtuous exemplars (Nussbaum Reference Nussbaum1986: 299; Reference Nussbaum1990: 73). The force of these prescriptions is not mere authority, but the superior understanding these people have about what virtue requires. Part of what is characteristic of the practically wise is that they have an excellent grasp of the concepts of virtue, vice, and happiness, and how these are to be deployed in various situations (Hursthouse Reference Hursthouse, Jost and Wuerth2011). Unvirtuous people will often think that the pronouncements of the wise are merely guidelines. But some categorical rules are enshrinements of hard-won wisdom that certain acts, despite appearances, are never virtuous. Thus, the advice from exemplars should not be seen as a parent sternly telling a child what to do, but as insightful warnings about how we should act and courses of action that we should not be tempted to take.

2.1. Epistemic Humility and Rules as Correctives

Those who have the developed perceptual capacities of full virtue are skilled in determining when general rules do not apply to particular cases. But most of us do not have this refined capacity. So what is involved in becoming virtuous? A common model is that we first learn rules, then, as we grow in understanding, we learn the exceptions. But moral rules also play a more complex role in the development of virtue, as often we will be in nonideal conditions where our judgment will be fallible. An important place for moral rules is for guiding less than fully virtuous people in difficult conditions. While a fully virtuous agent would know that some well-established moral rule does not apply in a particular situation, it may also take superhuman moral confidence to make such a judgment. Good but fallible agents should defer to the rule than take the risk of being wrong.

Nussbaum best explains the role for rules in complex situations:

While rules are not ‘exhaustive guides to practice’, they are nonetheless helpful in guiding people who are not in a position to understand their situation fully. People may be (a) short on time, (b) short on information, (c) morally imperfect, or (d) have some reason to think that their judgment will be biased or clouded (see also Nussbaum Reference Nussbaum1990: 73).

Consider how often Nussbaum's conditions will obtain. Even though she is clear that the virtuous deliberation goes beyond mere rule-following, the practical constraints of the actual world and our deficient characters will necessitate that often we should defer to established moral rules. We will need to be very confident in our own abilities and of our appraisal of the particular situation to see that an established rule does not apply. This is not to say that we are never justified in departing from a general rule (even if we are not fully virtuous, we are sometimes in an epistemic position good enough to make such judgments), but to emphasize that deferring to a rule is going to be the default for agents who care about acting virtuously. It is important that people progressing toward virtue both know moral rules and have internalized them so that they have motivational force.

Increasingly, people progressing toward virtue will understand the values that ground the moral rules, but this does not displace the moral rule as the default. On the contrary, the greater understanding of how, for example, not lying is a manifestation of honesty will reinforce the for-the-most-part rule as authoritative, while the greater understanding of how race-based discrimination perpetuates injustice will cement the prohibition as categorical. Thus, whether complete understanding of virtue and vice will transcend a rule depends on the rule. We have no reason for thinking that a virtuous agent will be unable to codify any categorical rules. The burden of proof is on the person who thinks that there are counterexamples to the claim that it is impossible to rape someone virtuously. At any rate, in nonideal circumstances, a less than fully virtuous person should think, ‘While this rule might have some exceptions, I'm in no position to decide that now’.

Nussbaum's comments highlight the importance of rules as a safeguard for people who might be biased or are tempted to engage in ‘special pleading’. Most people know plenty of moral rules, but many people do not follow those rules. It is possible that such people do not care about the rules—they know it is wrong to litter, but they are unbothered by it. This is a good account of many unvirtuous people. But Nussbaum's claim draws attention to a more subtle and pernicious way in which people can become vicious: they make an exception of themselves or their situation. I might know perfectly well that I am not to eat desserts on my diet but protest that this is a special occasion and I have had a particularly stressful day. More serious examples have the same structure: I know I am not supposed to become involved with students, but this student is especially talented and would get a lot out of private meetings, and is about to graduate anyway…

So Aristotle was right to say that we become vicious by doing vicious things (Nicomachean Ethics III.1, 5). The general rules of virtue are things everybody knows, but people become vicious because they do not follow them; they voluntarily choose to do something they know is unvirtuous, but they (wrongly) do not think that it is unvirtuous for them, now. Because it is psychologically implausible to say that people do evil for evil's sake, the better Aristotelian explanation of how we become vicious is that we are tempted to say too often ‘I know that normally this is wrong, but this is an exception because …’ The danger is thinking that there are more exceptions than there really are. Thus, we must be sure that we are not engaging in special pleading when we violate a generally valid moral rule (Swanton: Reference Swanton1991).

Moral rules are thus the default starting-point for deliberation, and rules act as important correctives for unvirtuous thinking in which we are tempted to engage. Rules will be indispensable in the development of virtue and will partly constitute the virtues, especially on conceptions where the virtues are understood as correctives (Foot Reference Foot1978: 8). Rule-following, then, plays an important role in the agency of people who are unvirtuous, biased, temped to make exceptions of themselves, and for agents who find themselves lacking in time and full information. That is a lot of us, a lot of the time.

2.2. Virtue, Skill, and Habituation

A more mundane way that rules figure in the development of virtue arises from rule-following during habituation. Contemporary virtue theorists have pointed out that mere repetition in insufficient for being well habituated (Burnyeat Reference Burnyeat and Rorty1980); they have, however, underemphasized the ways in which rule-based repetition is necessary for successful habituation. Virtuous agency is also a kind of skilled agency, so we can learn to become virtuous in a way analogous to how we learn skills (Annas Reference Annas2011: chapter 3). If this model is illuminating, then it is worth highlighting that rule-following plays a major role in developing skills and in the competent exercise of those skills.

To become habituated to something, you have to do it a lot. Many things that are initially painful become pleasant over time or the pain eventually does not bother you. The Greeks emphasized physical education in making you resistant to bodily pain. But for leg day to be beneficial and not just a gratuitous display of cruelty, you have to keep going to the gym. You have to perform actions consistently in order to recalibrate your pleasures and pains.

Plato and Aristotle both thought that for habituation to be effective, we needed laws to encourage habits from childhood (Laws 659d, 643b–644d, 752b–c, 797d–798d; Nicomachean Ethics II.3, 1104b3–12, 1105a1–13; X.9, 1179b31–35). Ideally, laws would ensure that people behave in the same way over many years, though smaller-scale improvements can be made by imposing rules on ourselves. While the fully virtuous person may be moderate enough to break a strict diet and still enjoy eating healthily, the person becoming virtuous may not be allowed such indulgences, as that may undermine the progress they have made. Thus concerns about habituation impose on not yet fully virtuous people stricter rule-following requirements than are appropriate for the fully virtuous.

We can make a similar point about the development of skills (Chappell Reference Chappell and van Hooft2014: 84). Learning a skill regularly involves following and internalizing rules (Annas Reference Annas2011: 33). These rules often safeguard against common ways trainees go wrong or enshrine the wisdom of skilled practitioners, thus constituting competent practice: baseball players are taught to keep their back elbow up; pianists are taught to warm up with scales; drivers are taught to look at where they want to go and not what they want to avoid, and to pulse the breaks when driving on ice; athletes are told to stretch after exercise; cooks are told not to open the oven while baking soufflé, and to fold—not beat—muffin batter; people playing Minecraft are taught never to mine directly down. Often what separates a good teacher from a mediocre teacher is the ability to come up with effective rules and regimens for their students. Equally skilled people vary in their ability to express the components of their skill.

Learning rules, then, is an integral part of learning skills. So if the skill analogy is an illuminating model for virtue, we should notice that: following rules does not undermine the learning of a dynamic skill (even when rules are ultimately left behind); rules often correct for common errors that those progressing in the skill regularly make; and rules that may initially seem odious come to be appreciated as important lessons handed down from experienced practitioners and may partly constitute the skill itself.

2.3. Rule-Following and Avoiding Evil

Rules not only play an important role in stopping agents from committing common errors, but also from committing egregious evils. Roy Baumeister's extensive study of evil points out that people rarely commit evil as evil. Evil acts of varying magnitudes are performed in ways that agents can justify to themselves. Baumeister claims that a common justification for evil acts is as instrumental to some end perceived as good; people often think that they are ‘just being practical’ or doing what is ‘necessary’ for their purposes.

Baumeister distinguishes two kinds of instrumental evils: those toward common ends, like wealth, power, or (sexual) pleasure, and those toward lofty, idealistic ends (Reference Baumeister1997: chapters 4, 6). The first category is interesting in that people who steal, murder, and rape often do these things for minor benefits to themselves, and they are normally ineffective means (especially in the long run). Thus the evil is heightened by the triviality of the end and the severity of the means. But we cannot explain the evil of these people in terms of their ends, for most people have these ends; the difference is that some people are prepared to do anything to achieve them (Reference Baumeister1997: 102). A prominent feature of vice, then, is a willingness to take any means available.

Baumeister highlights a similar tendency in well intentioned people with idealistic ends. If we looked only at their ends, we might think that they are good people, for reducing gross income inequality, removing corrupt governments, and doing the will of God are noble ends. But what makes these people evil is that they justify horrific actions as practically necessary or morally right in light of the end (Reference Baumeister1997: 178).

Moral rules play an important role in ensuring that people with mundane ends and people with noble ends do not turn vicious through doing evil actions. If a common way in which good people become bad is by doing characteristically unvirtuous things in pursuit of some good end, then we need safeguards against our doing such actions. A fully virtuous agent may know that this otherwise evil action is justified in these extreme circumstances, but it surely takes a huge amount of confidence to make such a judgment. Because full virtue is rare, we are likely wrong about our characteristically unvirtuous action being justified by virtue and we risk performing evil actions thinking they are good. Thus for those developing in virtue, the more pressing evil to be avoided is instrumental evil. Rule-following may not guarantee that we always get it right, but it is likely to prevent us from doing egregious wrongs.

3.1 The Weak View: Rules without Absolute Prohibitions

I now offer a modest way of thinking about virtuous agency that more explicitly gives a place to moral rules.

Virtuous agents make extensive use of for-the-most-part rules and normally act in accordance with them, but they are not committed to there being any absolute prohibitions or categorical rules. These for-the-most-part rules have been internalized by the agent and she employs them spontaneously, as part of her second nature. In many circumstances, conscious deliberation will not be required because a general rule will dictate what is to be done. These for-the-most-part rules thus create an internalized framework, though the rules themselves may fall away in developed moral skill, having been surpassed by a more comprehensive and nuanced understanding (Fridland Reference Fridland2017: 145; Sherman Reference Sherman1997: 273–274).

Moral rules nonetheless play a large role in shaping the affective responses of a virtuous agent. Virtuous agents do not gladly perform characteristically unvirtuous things simply because they know that it is right in the situation (Hursthouse Reference Hursthouse1999: 44–48; 77–78). Because characteristically unvirtuous things are often codified as moral rules, virtuous agents are pained by having to break these rules. Hursthouse argues that even if a virtuous agent could bring themselves to do something characteristically unvirtuous, they would still experience ‘moral remainder’—feelings of guilt, shame, a lower view of themselves, and a regret for having been in that situation (Hursthouse Reference Hursthouse1999: 44).

These considerations suggest that virtuous agents do not think of moral rules as ‘rules of thumb’ (Pace Sherman [Reference Sherman1997: 269–73] and Nussbaum [Reference Nussbaum1990: 68]). Rules of thumb are employed because they are handy and often work, but nothing is lost by noticing an exception case and acting contrary to the general rule. The rules of musical composition are rules of thumb. Composers do not feel bad violating such rules and it would be pathological to do so; such a person has not understood what the rules are for. Moral rules are not like this and the pathology is the opposite: thinking that these rules can be ignored when they do not produce the desired result and not having the slightest qualm about breaking them.

While such a view is compatible with absolute prohibitions, it could be claimed that we do not learn anything important about virtuous agency by thinking about the categorical impermissibility of some act-types. On this account, what is important is moral understanding, and a complete understanding of the virtues need not categorically prohibit anything. Some acts are characteristically unvirtuous, but the world is a messy place and virtue ethicists should not commit themselves to absolute prohibitions when there are situations where the virtuous agent would think it right to do something that she otherwise would never have considered doing.

This is a ‘weak view’ of moral rules in virtue because it would not satisfy Anscombe or Vogler, but it still corrects the tendency to downplay moral rules in the deliberation, action, and emotional responses of the virtuous agent. But ultimately the weak view claims that thinking about ethics in a rule-governed way misses what is distinctive about the virtue ethical perspective and distorts our characterization of the virtuous person.

While this approach is plausible and many contemporary virtue ethicists would accept it readily, it understates the importance of moral rules in virtuous agency. I defend a stronger position: deliberating and acting in accordance with absolute prohibitions is necessary for virtuous agency, and accurately reflects the virtuous person's ethical understanding.

3.2 Are Virtuous Agents Consequentialists Deep Down?

A common view is that virtuous agents think that consequences are morally important, and so virtuous agents will violate moral rules to produce good consequences, but they will also feel bad about breaking the otherwise-good rule. Consider how virtue ethics is often applied to cases: the virtuous agent pulls the lever, killing the one person and saving the five, but (unlike the consequentialist?) she feels guilt and sorrow at having to do such a thing. Virtuous agents, then, are really just consequentialists with feelings.

But should the virtue ethicist side with the consequentialist in hard cases or are there principled reasons why a virtuous agent might accept worse consequences and avoid breaking a moral rule? Following Anscombe, we should resist the move from thinking that because the consequences of following the rule are bad in this instance, the rule must be bad or following the rule must be bad (Reference Anscombe1958:10). Such an inference assumes the truth of consequentialism.

Even when the relevant alternatives both involve things that virtuous agents care about, there may be some actions that virtue categorically prohibits, namely, actions that no virtuous agents could bring themselves to do, or could justify to themselves, or to accept as the best or only option. When somebody asks, ‘But isn't it virtuous to torture someone to find the bomb and stop those people from being cruelly killed?’, it is perfectly coherent to reply, ‘No, because torture is itself cruel and vicious’. It's not that virtuous agents do not care about consequences—it is that they do not think consequences settle the question about what to do.

It might be thought that this amounts to the virtuous agent standing on principle, but there are virtue-based reasons to support resoluteness. Sabina Lovibond has recently shown that virtuous agents exercise hope in circumstances where it seems you may have to get your hands dirty in order to prevent dire consequences (Reference Lovibond2015: 154). In the torture case, a virtuous agent may think that while it is guaranteed that her torturing will cause harm, it is not guaranteed that her torturing will uncover reliable information, nor guaranteed that the bomb will go off. She may hope that the dilemma is only apparent, recognizing that outcomes are uncertain and that situations can change (Foot Reference Foot2001: 74n15). Virtuous agents will be under no delusion about the likelihood of some consequences obtaining, but we most need hope for those situations in which the ends are worthwhile and the likelihood of success low. The consequences of threats do not always eventuate, and so the virtuous response to the threatening tyrant or terrorist may well be to look them in the eye and say, with defiant courage and hope, ‘We shall see’.

Without Providence, there is no guarantee that a moral dilemma will resolve and some would blame this agent for not getting her hands dirty in doing what was ‘necessary’ to prevent the dire consequences. But a virtuous agent might rather take responsibility for letting bad outcomes occur than do vicious things. Taking responsibility amounts to more than being the scapegoat when people point fingers. Virtuously taking responsibility for bad consequences involves making it your business to correct the harms that occurred, and to do one's best to ensure that such situations do not occur again. So while there are hard cases, a virtuous agent would not always choose to violate a moral rule when dire consequences threaten.

3.3 Corrupt Minds

Another way to think about the importance of absolute prohibitions in virtue ethics is to better understand what Anscombe meant when she said that a willingness to violate such prohibitions reflects the workings of a corrupt mind (Reference Anscombe1958: 17).

What is it to have a corrupt mind? One way a person's mind is corrupt is when they allow non-virtuous reasons to take priority over virtuous considerations—their nonvirtuous reasons were not ‘silenced’ (McDowell Reference McDowell1979: 345). If I let my hunger at the grocery store act as a reason to eat the cake before paying for it, then I show a kind of corruption. My hunger is not the kind of reason that could outweigh the fact that I am stealing. But this approach does not help us in cases that Anscombe is interested in, namely cases in which it looks like reasons provided by virtue result in conflicting judgments (Anscombe calls ‘corrupt’ someone who thinks that it right to execute an innocent person to prevent a hydrogen bomb war). Thus justice prevents us from executing the innocent, but it also demands that we avoid letting millions of people get murdered by nuclear weapons.

Another way of understanding corruption is in accepting the thought experiment as a dilemma: the deliberator has failed to exercise her moral creativity in understanding what the situation is and what her options are (Hursthouse Reference Hursthouse2006: 292–294; Kamtekar Reference Kamtekar2004: 470–473). This agent takes as given some description of her situation, one that leads her to think that some action is ‘necessary’, whereas if she had described the situation more fully, she could have identified alternatives. Instead of accepting that we must either execute the innocent or suffer nuclear war, Anscombe points out that we do not really know what other alternatives might be available, for example, ‘stalling by a feigned willingness to comply, accompanied by a skilfully arranged ‘escape’ of the victim’ (Reference Anscombe1958: 17n1).

Somebody could protest that there are genuine dilemmas in the world and it is wishful thinking to claim there is always some way out. But Anscombe's point is that people show a corrupt mind when they consider in advance that a vicious action is necessary (Lovibond Reference Lovibond2015: 148, 155; Anscombe Reference Anscombe1958: 17). The corruption is exacerbated through a willingness to allow such thought experiments to dictate the shape of one's moral theory—especially when they do not accurately reflect our ethical world (recall §1.1). The problem with thought experiments (and even with literature or reported real-life examples) is that we do not know all the details of the situation, so it is easy to fabricate a situation in which breaking some moral rule looks like the only course of action.

But we do not actually live or deliberate in this way. Despite philosophers’ incessant talk of choosing to kill one person to save five, or of torturing someone to get information about the location of a bomb, the vast majority of us would be extremely reluctant to murder or torture and would think that we could not walk away with our life unmarred (Hursthouse Reference Hursthouse1999: 77). How ethics is taught in classrooms makes the rule-follower look naïve, but this is because of the artificiality of the examples. We make it impossible for our students to follow up their protestations of ‘there must be another way!’ with actual solutions, and we stipulate dilemmas ex hypothesi, so of course they get the impression that moral rules eventually fail. But there often is another way—the classroom is just no place to find out what it is.

While preparing oneself in advance to perform vicious actions is a part of what Anscombe meant by ‘corruption’, there is a further element: thinking that virtue could demand that you do something unvirtuous. The problem with the execution case is not just that it artificially restricts options, but that the person giving the dilemma is assuming that justice could require you to murder the innocent to save the innocent. This generates a contradiction as injustice is directly manifested in knowingly executing the innocent, and so justice in this case both requires and forbids us from performing the action. A solution would be to say that justice requires you to do something characteristically unjust (but not unjust in this instance). But notice the corruption that has occurred in our understanding of the virtues—namely, saying that justice makes right the murdering of an innocent person.

Virtuous agents, then, do not think of rules as mere guidelines for actions that they transcend once they attain goodness of character. Rather, they see these rules as enshrining our deepest moral commitments. These are rules that less than fully virtuous people too easily set aside when the going gets tough or in light of putative gain that amounts to ‘special pleading’. When virtuous agents are faced with a dilemma where it looks as though some rule must be broken, their response is to reason creatively and find alternative courses of action. It is because they take the moral rule seriously that they engage in this process so tenaciously. This is not the attitude of someone who thinks that moral rules are guidelines.

3.4. The Strong View

Given these considerations, here is a virtue ethics that gives moral rules a central place and incorporates absolute prohibitions.

Fully virtuous agents understand that certain act-types characteristically manifest vice and some act-types necessarily manifest vice. The former class generate for-the-most-part rules that guide the deliberation and actions of virtuous agents, while the latter class generate absolute prohibitions and categorical rules. These rules shape the affective responses of the virtuous agent, and she would be pained at the thought of violating them. Virtuous agents think of these rules both as enshrinements of tried and tested conclusions of practical wisdom, and as partly constituting the virtues themselves. While their understanding transcends these rules, virtuous agents do not leave these rules behind or think of them as uninformative in deliberation, as often these rules have been internalized and thus have become automatic, or they are deferred to in the frequent circumstances where the agent cannot fully determine all the particulars of some situation. Rule-following thus characterizes a central part of the virtuous agent's deliberation and action.

While rules play an important role in the functioning of the fully virtuous agent, they are even more important for those still advancing in virtue. Rules act as checks on the inclinations of people who may be tempted to act badly, and to prevent well intentioned people from doing unvirtuous things in the name of some good goal. Rules act as important guides for moral development, as they deliver the prescriptions of wiser people, directing the student's attention to areas of moral concern that they are likely to overlook or underappreciate. Rules also help to form the habits that make up a good character, and provide those progressing toward virtue with parts of the skill that will later come to be incorporated into the expert performance of virtuous actions.

I do not think that virtue ethicists must endorse the stronger view. Rather, my position is that such a view is coherent and has strong reasons in its favor, at least for endorsing some categorical rules (no doubt we may fruitfully argue about which rules are categorical; while I like my list from §1.1, I am not a sage). Thus, not only can secular virtue ethicists justify why some moral rules are inviolable, we can learn what fully virtuous agency involves by showing why virtuous agents are committed to following such rules categorically. While rule-following is not sufficient for virtue, it is necessary.

I hope this account is a plausible development of Anscombe's vision in ‘Modern Moral Philosophy’. For on this view we do not see moral rules as having a law-like force that compels us from the outside, as if by some divine command but without the divine commander. Rather, the force of the rules comes from within one's ethical perspective, from one's understanding about what one's values could never allow and what one could never bring oneself to do (Annas Reference Annas2014: 18–22). But to deny the outside-in vector of moral requirement is not to deny that some moral requirements have a law-like structure in categorically prohibiting or requiring particular actions. Anscombe was providing an alternative way to ground these demands by using secular virtue terms. Virtue ethicists, then, can account for categorical rule-following, historically did include such rule-following in their theories, and have good reason now to think that such rules are morally important.