Matthew Kramer recently dismissed the existence of ‘small-scale evil’ which has been argued for by Garrard and Morton; this is the view that evil actions do not necessarily ‘involve killing, major suffering, or life-destroying humiliation’Footnote 1. Kramer, in contrast, contends that all evil actions are ‘underlain by sadistic malice or heartlessness or extreme recklessness that is connected to severe harm in the absence of any significant extenuating circumstances’Footnote 2. He insists that the infliction of severe harm is a ‘necessary constituent of wickedness’Footnote 3. Given his account of evil conduct, Kramer dismisses purported examples of small-scale evil actions offered by Garrard and Morton as either inappropriately described or as conduct that does indeed inflict serious harm. Even the sadistic voyeur, someone who maliciously enjoys watching the sadistic acts of others but takes no action himself, engages in evil conduct due the relevant way it is connected to the infliction of severe harm. For Kramer, this is the case despite the fact that the voyeur has not increased his victims’ pain or anguish. By subordinating the victims ‘in this fashion’ means that unbeknownst to them ‘they have been harmed by the contravention of their dignity’Footnote 4. Kramer goes further to claim that even if the sadistic voyeur is enjoying the pain and suffering of those who are deceased, posthumous harm still results. In short, Kramer claims that the voyeur’s actions are evil and this is so because his conduct is relevantly connected to the severe harm inflicted by another person or groups of persons on the victim.

In setting out his case against the existence of small-scale evil, Kramer throws out a challenge to those who continue to advocate for it in the face of his response. To support the existence of small-scale evil, one needs to ‘supply some persuasive examples of wicked actions that do not involve any very serious harmsFootnote 5. Responding to this challenge is the main task of this brief paper. I concur with Kramer that the examples of small-scale evil offered by both Garrard and Morton are problematicFootnote 6. However, it is one thing to successfully rebut the examples offered by Garrard and Morton, but quite another to confidently claim that small-scale evils are impossible.

While most evil actions do indeed involve severe harms connected to an evil mode of culpability, there is nevertheless the possibility of small-scale evils which cause moderate or no harm to their victims, and which Kramer is wrong to summarily dismiss. While my short paper has a very specific and narrow aim, if I am correct its conclusion will have wide-ranging implications for many extant theories of evil. Yet this paper will not offer a detailed discussion of the nature of evil, nor the role of motivations, moral landscape or character in understanding and defining evil actions. I have done this elsewhere.Footnote 7 Rather it focuses on a specific issue in Kramer’s account of evil, one that is widely supported by other theorists offering definitions of evil actions; namely, the supposed non-existence of small-scale evil.Footnote 8 My hope is that my paper will raise serious doubts about Kramer’s explicit rejection of this possibility thereby raising doubts about the implicit rejection found in many theories of evil that require the criterion of serious harm as a component of any evil action.

My response to Kramer takes two paths. The first challenges how he understands the notions of ‘connected with’ and ‘severe harm’, specifically in those situations where a repugnant immoral action does not increase the victim’s suffering in any way. I argue that his use of ‘connectedness’ and ‘severe harm’ are not sustainable in certain cases that are nonetheless plausibly labelled as evil. The second path is to respond to his direct challenge and offer two hypothetical examples that demonstrate cases of small-scale evil; actions that are rightly labelled evil yet the harms caused either are small, moderate or non-existent.

Kramer’s Notions of ‘Connectedness’ and ‘Severe Harm’

I begin with Kramer’s notion of connectedness and its importance for defining evil actions. Kramer explains that connectedness does not mean merely causal connections that bring about severe, grave, or catastrophic harms. His view is far more subtle and sophisticated. As he states, ‘not every wicked misdeed actually results in the infliction of grave detriments’ yet they all involve some connection to grave detriments.Footnote 9 Failed attempts to commit evil, cases where harm is overdetermined, attenuated links where the organisers of evil deeds are not those who carry out the actual act all persist as evil because of their relevant connection to severe harms. So what always remains relevant and necessary for evil actions, in Kramer’s view, is that there is this relevant connection to severe harms ‘in some repugnant manner’.Footnote 10

In many, perhaps most, cases of evil actions I agree with Kramer that there is this connection with severe harm. However, I differ in that evil actions occur on a spectrum of connectedness (and harmfulness) such that some actions are better classified as small-scale evil; cases where the connectedness with harms is tenuous but the acts are nonetheless evil due to their motivation, intention and the perverse satisfaction they give to the agent. The claim, therefore, that a necessary condition for defining evil actions is the appropriate connectedness to a severe harm is unsustainable and misleading in a small but important set of cases. To insist on this need for connectedness is to set the bar too high for identifying some actions which we rightly judge as evil. It is important to point out here that arguing for the existence of small-scale evils is not a merely scholastic enterprise. The possibility of small-scale evil has important implications for how we ought to respond to them. For example, an agent’s small-scale evil actions might not require punishment, even if we think very badly of him for acting in that way. The commission of large-scale evils, in contrast, always merit severe punishment and provoke social policies to prevent or discourage their occurrence in the future.

What then is problematic about Kramer’s claim linking all evil acts as necessarily connected to severe harms? While in most cases of evil acts, the connection to severe harm is clear and obvious, there are others where this relevant connection is trivial, irrelevant or highly doubtful. In these cases, we can and should still classify the action as evil without needing to stretch the idea of connectedness beyond credibility.

Consider again the case of Sadistic Voyeur. Here an agent, unbeknown to the victim, watches the suffering and humiliation of another person with gloating delight and satisfaction. He takes no action himself and is not seen enjoying this voyeuristic act by anyone else. The connection between his act and the severe harm done to the victim is simply that he has decided to watch the suffering and humiliation of another person for his own gratification. The victim’s suffering is in this case an opportune and fortuitous event from the sadistic voyeur’s perspective. He does not cause the suffering, nor advocate for it, nor add to it (even if you accept Kramer’s arguments that the voyeur’s actions do indeed harm the victim by reducing her to a mere object of gratification). The connection to severe harm here is very weak and all it does is serve as a convenient background context that enables the voyeur to enjoy his sadistic impulses. It is much the same as other important but non-relevant background conditions (such as the victim is a sentient being that can suffer and be harmed, that the victim is physically in the vicinity, and so on) that are required for the situation to obtain such that the voyeur can revel in his perverse pleasures. In short, if there is a connection to severe harm in the Sadistic Voyeur scenario, it is one of the plethora of existential conditions that obtain to make this situation possible. However, in such cases the issue of harm does not have a relevant bearing on why this particular voyeuristic act ought to be classified as evil.

Perhaps my point here can be is made clearer when we think of the sadist sitting alone imagining horrendous scenarios and deriving perverse pleasure from them. Here there is no possible connection to a specific real event as the scenario is a fantasy. The only possible vague and distant connection with severe harms to persons might be that the sadist’s fantasy ultimately derives some of its content from the stories of past atrocities and horrific acts with which he is familiar. It seems in such a case we would think such imaginings, and the perverse pleasure it gives, as deeply distasteful, grotesque, and repugnant. But we are unlikely to think this is so because of an identifiable connection to some unknown severe harms perpetrated on actual victims. It is enough that the sadist’s act of imagining the torture and humiliation of others gives him such perverse gratification and pleasure. To concentrate on seeking a connection to severe harm rather than focusing on the motives and perverse pleasure of the voyeur, is to focus on the wrong reason for judging such an action as evil. The insistence that deriving perverse pleasure from imagining the pain of others can only be evil if it is also connected to actual severe harm is to over determine what is needed to categorise this action. While it is true that what makes the worst kinds of evil acts so terrible is that direct connection with the severe harm, this need not and cannot cover all and every case of evil conduct. What marks the sadist’s act as evil, whether it is connected to severe harm or not, is the disturbing and repulsive gloating pleasure at imagining others humiliated and tortured. It is a small-scale evil that does not inflict harm (or at least not a great deal of harm) on its victims. Yet we still ought to think of such an action as significantly worse than those we judge to be simply immoral. Small-scale evil properly provokes what Kramer calls a ‘shuddering sense of horror’ and which deserves the particular and special normative responses reserved for the ‘grotesque perversity of evil’Footnote 11.

Kramer might respond by pointing out that if the connection to severe harm is too tenuous, then we should think of such acts as simply immoralFootnote 12. Indignation rather than horror is the appropriate response here. However, Kramer does think the Sadistic Voyeur case is classified correctly as evil albeit within the category of ‘low-grade instances of evil conduct’Footnote 13. It sits at the lower end of the spectrum of evil. As he points out:

Moreover, given that the property of wickedness is vague, the range of low-level instances of wicked conduct will shade into the range of instances of conduct that are neither determinately wicked nor determinately not wicked; the demarcation between the former range and the latter is encompassed by higher order vagueness. As a consequence, we plainly need to be circumspect when judging whether low-grade instances of wickedness are such instances or noFootnote 14.

Kramer recognises the existence of different levels of severity when classifying evil conduct. What is puzzling is why he insists that those that are indeed low-grade necessarily have a connection to severe harms. It is enough for this type of conduct to be motivated by vindictiveness, brutality and cruelty to be judged a small-scale evil. The difficulty, as Kramer himself notes, is how to classify conduct that falls on the vague boundary between immorality and wickedness. Some cruel acts are merely immoral while others, in different circumstances, or with further aggravating conditions, would be evil. One of the aggravating conditions is the amount of harm caused, but this need not be the case. It might be the particular kind of gratification, or the intensity of perverse pleasure, or the unambiguous cruelty of the intention to humiliate (even if the act fails to do so) that makes the difference between immoral acts and those that are small-scale evils

I now turn to Kramer’s claim that a necessary condition for all evil actions is that they involve severe harms. It is worth noting here that Kramer is aware that the property of ‘severe harm’ is vague and difficult to pin down. This leads him to discuss the phenomenology of evil. He suggests that one way of knowing the difference between acts that are immoral misbehaviours and those that are properly labelled evil is found in our emotional responses to them. Immoral actions evoke indignation and anger, while evil actions induce ‘a shuddering sense of horror’Footnote 15. Note here that the difference between feeling indignation and feeling horror is not necessarily based on the level of harm inflicted. Kramer makes this clear in his rejection of Morton’s example of small-scale evil involving the example of a professor who embarrasses his student to save face during a lecture.Footnote 16 While he agrees that the retort is a clearly ‘cruel and self-indulgently deceitful riposte’, it is nevertheless not an evil actFootnote 17. Kramer insists that embarrassing a student through a wrongful and self-serving retort whatever the level of unintended harm it may cause, lacks the ‘grotesque perversity of evil’Footnote 18. While the perversity of evil does often involve terrible harm, this need not be the case. Immoral actions can entail very severe harms while evil actions arising from repugnant motives might not. The judgment that needs to be made is a difficult one and Kramer admits to the inherent vagueness of this process. Considering Morton’s example, we can see that the retort given to the student lacked that qualitatively different heartless frame of mind. But, if it had possessed this frame of mind then presumably Kramer’s judgment would change whether the harm it intended to inflict obtained or not. The level of harm caused, or expected to have caused, would guide us in identifying the kind of evil act rather than whether this act was correctly identified as evil.

Given Kramer’s insistence on the need for sever harms in all evil conduct it is a pity that Kramer does not provide a taxonomy or hierarchy of different kinds of harmful actions. However, it is possible to ascertain a reasonable grasp of the wide range of harms he considers to be severe from the examples he uses. He rightly identifies life destroying harms or those that make any kind of decent life impossible as very severe. Examples of such harms are the great evils of torture, murder, enslavement, forced starvation, and rape. However, it is worth noting that Kramer drops the term severe when referring to the harm enacted on the voyeur’s victims who are ‘subject to the indignity of furnishing’ him with pleasure and amusementFootnote 19. This harm occurs whether the victim is unaware of it or not and applies even to those who are deceased.Footnote 20 Treating a person as a mere means to one’s own perverse ends, violating their right to be respected and treated with dignity, need not be a severe harm. As he eloquently puts the case:

‘Though their anguish has not been intensified (ex hypothesi), they have – unbeknownst to them – been harmed by being subjected to the indignity of furnishing (the Voyeur) with sadistic amusement. His use of them and their tribulations for his own perverse enjoyment has harmed them by demoting them to the status of playthings. Their posture as moral agents entitled to respect, and as sentient creatures entitled to concern, has been besmirched by his abominable behaviour. Though he has not increased their pain, he has subordinated them grotesquely to his own purposes. Subordinated in that fashion, they have been harmed by the contravention of their dignity’Footnote 21.

There are two important points to stress here. Firstly, we know that Kramer claims that the voyeur’s actions are indeed evil and not simply immoral. And, secondly, he states that the voyeur’s victims have been harmed by the contravention of their dignity. But for Kramer, the voyeur’s act is evil not because of the actual harm it causes in treating the victim as a mere means to his pleasure but because the act itself is connected to the severe harm of torture committed by someone else. What makes it a low-grade evil is that the actual harm done by the voyeur himself is not severe. Kramer raises this point about the degree of harm done by the voyeur, firstly, to rebut Garrard’s claim that since such an act does not add to the suffering of the victim no further harm has been done. And secondly, the degree of harm done enables him to place such conduct on the spectrum of evil actions from the worst to the very low-grade. Therefore, for Kramer the act is evil because it is connected to severe harm in the relevant way but it is low grade because the act of the voyeur himself only contributes a small quantity of harm to the victim.

So Kramer consistently remains wedded to his claim that all evil actions are connected to severe harms and that small-scale evils are impossible if we mean by small-scale, as Garrard and Morton do, that there is no need to have any connection to severe harms. However, this is not a sustainable view if the level of harms indicates the type of evil we are facing – purgative, severe or low-grade. If it is a necessary condition for such acts to be connected to severe harms in virtue of being classified as evil, then the term ‘Low-grade evil’ is an oxymoron. The voyeur’s evil act is connected to severe harms and cannot therefore be low-grade. The fact that the voyeur has also added to the harm inflicted on the victim makes this even more so the case.

Kramer, however, is right to point out that we do grade evil acts in part by their impact on their victims. Actions that increase suffering, ceteris paribus, are worse, morally speaking, than evil actions with the identical motives and intentions that do not. However, even if the connection to severe harm in the relevant way is one of the conditions for defining a particular type of evil action, it does not exclude the possibility of small-scale evils obtaining due to other relevant conditions that involve no, or a small amount of, harm to the victims. A more plausible way to understand some low-grade or small-scale evil conduct is to focus on the culpable evil motives with the level of harm providing the intensifier, so to speak. Or to put it another way, if our concern with an action’s relevant connection to severe harm is a necessary condition for evil conduct, we will be unable to identify evil actions arising from the most repugnant motives and intentions but which cause very little or no harm to their victims. The issue of harm is important and relevant but this is because it is needed to categorise different kinds of evil conduct as either small-scale, quintessential, or at its worst to follow Kramer’s description, purgative evilsFootnote 22.

Kramer rejects this intensifier account I have outlined it above. He dismisses Garrard’s example of small-scale evil concerning a tyrannical state that charges the parents of a boy for the bullet used to kill him. Garrard argues that the charging of the bullet is not a great harm in itself but strikes us as evil nonethelessFootnote 23. For her this is evidence for the possibility of small-scale evil. Kramer agrees that such an act is evil but consistently with his theoretical account maintains that this is so because of the connection to the grievous harm inflicted on the boy. If the bullet had missed or merely grazed him, then charging the parents for the bullet would be ‘wrongful and cavalier but would not have been wicked’ since it would not be linked to any serious harm.Footnote 24 The act would be ‘haughtily intimidating’ but not meet the right ‘attitude … of utter disdain for the humanity and elementary well-being of someone else, who is regarded as a mere plaything to be destroyed or otherwise grossly mistreated for one’s own sinister purposes’Footnote 25. However, Kramer’s argument here misconstrues why charging for the bullet, even when it merely grazed the boy, is rightly categorised as evil. Such actions by a tyrannical state are more than haughtily intimidating even when the boy managed to escape and suffered no injury. Charging for the bullet seeks (among other things) to demonstrate the state’s dominance and unchallengeable power over individuals who are unable to defend themselves against its abuses. This act is evil because it treats individuals as playthings to be grossly mistreated for the State’s own sinister purposes. It exploits the sacrosanct parent/child relationship that has love and care at its core to demonstrate its power and callous disregard for a widely accepted civilised boundary. So, even when the state charges for the bullet that missed the boy it is an evil act, albeit a small-scale one.Footnote 26 The connection to the degree of harm caused categorises the type of evil conduct involved given that the core motives and intentions here were culpably evil.

Kramer interestingly makes what seems to be a parallel point elsewhere when discussing moral luck and the non-materialization of a riskFootnote 27. He states that there is an important ethical difference between reckless conduct involving an unmaterialised risk and the same conduct where the risk is materialized. For Kramer, the difference is significant since the latter may designate a case of purgative evil while the former is a lesser evil. He insists that both acts would be evil but the materialization issue changes the way we understand the action and deal with it. With this I concur but why then is it not similarly the case when considering the degree of harm caused? If materialization of risk makes a difference to the type of evil involved, it is hard to see why different levels of harm do not make a similar modification. In most cases of immoral conduct, and this would include the category of evil actions, an ethical difference arises from the degree of harm caused. Kramer accepts that evil is a scalar property but rejects the view that the magnitude of harm caused is an appropriate metric for determining how this categorization varies. However, it is hard to see why he adopts this position. The level of harm inflicted coupled with the relevant culpable motives or intentions do make a difference to the seriousness and deplorability of the action that we identify as evil. And, if this is right, the way is open for what Morton and Garrard call small-scale evils - evil actions where the level of harm is non-existent or moderate.

Two Examples of Small-Scale Evil

The second response to Kramer rejection of the possibility of small-scale evils is to take up his challenge. In what follows I offer two examples where an action is motivated by a culpable evil motives but causes moderate or no harm to their victims. Consider, then, the first, which I call ‘Tantalus Redux’. Imagine a contemporary version of the Greek mythological story of Tantalus. An undertaker, Fred, is tasked with preparing recently deceased Tim for burial. He organizes the funeral reception in line with the last wishes of the deceased, which involves providing food for the family and other mourners after the burial. Fred, unknown to the family, takes ghoulish delight in following Tim’s final secret instructions to extract several of his body parts prior to closing the coffin and cook them into a stew to serve to the family members and guests. This is done, and the mourners eat the stew having no idea that the meat is actually Tim’s body parts. Fred does not tell anyone and neither the family nor the other mourners ever find out. Tim is not harmed posthumously through a rights violation as it is his wish that this happens, and those who eat the stew are not harmed physically, and, due to ignorance, experience no psychological trauma either. However, Fred knows the taboo on eating human flesh and the horror and revulsion it would provoke if the mourners discovered this had happened. He revels in the knowledge that he has ghoulishly tricked the mourners. Given this malicious delight, it seems plausible to consider Fred’s conduct evil, although it causes no or very little harm to his victims. Fred’s conduct, then, is best described as a case of small-scale evil.

Kramer might object to my claim that Fred has caused very little or no harm to his victims since deceiving them, especially on such a matter, just is to cause severe harm. This harm is cashed out as a violation of their rights and subordinating ‘them grotesquely to his own purposes’Footnote 28. In short, Kramer argues they have been harmed by the violation of their dignity. But Kramer cannot plausibly claim this is severe harm since he rejects the violation of rights in the voyeur scenario as such a case. If there is any harm done here it is small or moderate at best. Furthermore, Kramer’s claim that harm is done by the violation of rights seems implausible in the Tantalus Redux example given a common sense understanding of what harm is to others. Harm typically involves the victim suffering pain or humiliation or anguish. We harm persons when deceiving them or treat them as mere means to our own ends because in almost all the cases this results in deleterious consequences for them. In circumstances where this does not happen, we think the deception less serious and in some cases trivial. If Fred’s victims are never aware of what has been done to them, claiming they have been severely harmed stretches the notion of severe harm (or a connection to severe harm) beyond any plausible ordinary understanding of the term. Even if we agree with Kramer that there is some level of harm done to Fred’s victims, it is hardly severe in comparison to torture, for example, or to rape. To hold such harms as equivalent or in the same category as other severe harms is to stretch the notion beyond credibility or render it vacuous. However, as pointed out above, Fred’s conduct is rightly labeled evil since maliciously and gloatingly feeding the organs of a deceased relative to the mourners is qualitatively different from a merely immoral act. It was carried out with the relevantly culpable malice and perverse pleasure which we rightly ascribe to evil conduct.

Now consider the second hypothetical case, which I call ‘Robot Torture’. Imagine a situation where Bert tortures and then kills Sam in a secret dungeon. Bert does this out of extreme malice, heartlessness, and rampant sadism. However, Bert knows that he is ‘torturing’ and ‘killing’ a robot, albeit a very lifelike one that responds as a real human being would do in such circumstances. Sam screams, bleeds, and even ‘dies’ as a human would. But Sam has no feelings or consciousness beyond its computer programming. No one knows about nor observes Bert’s actions, and no harm is caused by setting an example that might influence or encourage others to behave in this manner. Yet it seems plausible to claim that Bert’s actions toward Sam is a case of small-scale evil. His malicious, sadistic pleasure and enormous gratification from deliberately causing ‘pain’ and ‘humiliation’ arises only because he pretends to be torturing a real defenceless person. The simulated realism of his actions gratifies him. He would torture a real person if he were not scared of the consequences of being caught.Footnote 29 Yet his conduct towards the robot does not harm anyone nor is there a connection to severe harm of others, dead or alive. The fact that his actions are unobserved, that no person’s dignity or rights are violated, and that there is no traducing the rights of a deceased person, provides a case of evil conduct without harm. What is more Bert’s conduct cannot be classed as harm towards another’s property if we assume this is a robot legally owned by Bert who can do with it as he pleases. And even though Bert wishes and pretends to be torturing a real person, wishing and pretending to severely harm someone is clearly not the same as actually harming him.

Kramer might offer two responses here. Firstly, he could argue that the pretence that Bert is torturing a real person does provide the relevant necessary connection with the infliction of severe harm. However, this use of connectedness is far too concessive and implausible. As mentioned above, it is a considerable stretch to claim that pretending to harm a real person is relevantly connected to actual harm in anyway. Secondly, and this is the more likely and consistent position for Kramer to adopt, he would simply deny that Bert has engaged in evil conduct. Yet this denial runs up against strong intuitions to the contrary. Bert’s conduct given the qualitative nature of his sadistic motives and intentions do seem evil although properly categorised as a small-scale one. Again, what this hypothetical case strongly suggests is that what makes an action evil is the culpable evil motives and intentions that provoke it. The issue of the harm caused raises the stakes, so to speak, enabling a distinction between small and large-scale evils.

If I am correct, we have at least two examples of small-scale evil where motives, intention and sadistic pleasure render the act evil without the necessary condition of severe harm. Kramer, therefore, is too quick to dismiss the possibility of small-scale evils. To be fair to Kramer, this issue is not of paramount importance to his project in defining evil conduct. He is interested in cases of purgative evil that may justify the death penalty. For such cases, it is right to claim that they will always involve very severe or catastrophic harms for the victims. Nevertheless, Kramer is incorrect to dismiss the possibility of small-scale evils articulated by Garrard and Morton, which are an important part of our understanding of evil conduct.Footnote 30