Defining evil in the philosophy of religion

The problem of evil is most intuitively expressed in the form of a personal lament: ‘if a perfectly good God exists, how could God allow evil things to happen to me?’ That is, why has God not eradicated the evils that have happened to me? Less personally, J. L. Mackie (Reference Mackie1955, 201) presents the problem of evil as the logical impossibility of reconciling two beliefs: (1) a perfectly good God exists, and (2) evils exist. If we presuppose that evils exist, then God does not. He adds another premise to create the supposed contradiction, which he considers a ‘quasi-logical’ rule: ‘a good thing always eliminates evil as far as it can.’ He is not alone in this. Stephen Wykstra shares with William Rowe the belief that ‘a wholly good God must be “against” suffering in this sense: such a being would allow suffering only if there were an outweighing good served by so doing’ (Wykstra Reference Wykstra1984, 91). Now the evil things that this problem refers to are usually differentiated into natural evils and moral evils. Natural evil – or what Mackie (Reference Mackie1955, 206) calls ‘physical evil’ or ‘first-order evil’ – is evil that results from natural processes, lacking an agent (usually a human) that could be held morally responsible for it, whereas moral evil is evil that results from the misuse of free will by moral agents, which makes said agents morally blameworthy. Classic examples of natural evils include natural disasters and diseases, while classic examples of moral evils include intentional wrongdoing like murder or fraud. Most philosophers of religion treat both natural evils and moral evils as part of the problem of evil, although different theodicies are usually called for in response to each type. However, we argue that philosophers should reject the term ‘evil’ for natural evil. In doing so, what the literature calls ‘natural evil’ may no longer factor into the problem of evil.

Since this is a critique of the use of the term ‘evil,’ we will use ‘X’ to refer to the problematic thing, whatever it is (usually assumed to be ‘evil’), which is supposed to generate the so-called problem of evil for the existence of God. This will distinguish the local use of the term ‘evil’ from what we think the term actually should represent, without presupposing that X is necessarily problematic for God’s existence. Most philosophers who have written directly in response to the problem of X have read X, or what they refer to as ‘evil,’ as equivalent with suffering or pain. Mackie (Reference Mackie1955, 206) straightforwardly equates ‘physical’ (aka natural) evil with pain, and implies that moral evil (‘second order evil’) is the pain caused immorally by ‘malevolence, cruelty, callousness, [and] cowardice….’ Jim Stone (Reference Stone2003, 266) does the same with suffering: natural evil is ‘the rest of suffering’ not caused by ‘the evil men do.’ Rowe (Reference Rowe1979, 335) equates suffering with badness and evil: ‘Intense human or animal suffering is in itself bad, an evil.’ Wykstra (Reference Wykstra1984, 73) doesn’t even need to use the word evil (although he occasionally does), because for him the problem of evil is equivalent with the problem suffering poses for theism. C. S. Lewis (Reference Lewis1940) simply called this ‘the problem of pain.’ David O’Connor, C. Stephen Layman, A. M. Weisberger, and Nick Trakakis all claim that without pain, no evils occur. For example, if O’Connor’s cat falls off a window ledge but has an improperly functioning ‘neurophysiological system’ such that the cat feels no pain, discomfort, or panic from the fall, then the cat suffers no evil, for it is ‘the raw feel of pain, its phenomenological dimension, [that] is always a natural evil….’(O’Connor Reference O’Connor1998, 89). For Layman (Reference Layman2003, 11), if a natural catastrophe causes no suffering or loss, no natural evil occurs either.

This, in Weisberger’s (Reference Weisberger1999, 82) words, is ‘evil in its most mundane sense’; pain or suffering is either a necessary or sufficient condition for evil. While Rowe allows for ‘disagreements about whether something is good,’ including the existence of ‘unfamiliar goods, goods we haven’t dreamed of,’ he concedes no corresponding leeway for defining evil in the mundane sense (Wykstra Reference Wykstra1984, 336-337). Layman (Reference Layman2003, 10), however, expansively defines evil as ‘badness in general, including such items as moral wrongdoing, suffering, loss, and being under systematic illusion about important matters.’ This opens the possibility that suffering may not be sufficient for evil. In particular, suffering is equivalent with an awareness of ‘an unsatisfied (or apparently unsatisfied) desire or instinctual drive,’ but unsatisfied wicked or unhealthy desires or instinctual drives do not constitute suffering. Regardless of how phenomenologically painful it is, being unable to satisfy the desire to murder ‘a rival in love’ or the ‘pica victim’s [having an unsatisfied] desire to eat harmful substances’ is not a bad thing. Rather, having such desires is. Evil, for Layman, is therefore defined in relation not to the phenomenon of suffering or pain, but some other axiological standard by which suffering or pain is to be evaluated as good or bad.

What that standard might be is not often made explicit, but Jaime Mayerfeld, at least, identifies it with Epicurean hedonism against Aristotelian eudaimonism, and with Bentham and Sidgwick’s (but not Mill’s) utilitarianism (Mayerfeld Reference Mayerfeld1999, 11-12). Some philosophers of religion openly eschew defining evil. Michael Peterson (Reference Peterson1998, 10) thinks that attempting to offer a specific definition of evil ‘frequently ladens the meaning of evil with preconceived ideas and thus hinders objective discussion,’ but he does not consider whether a widespread utilitarian identification of evil with suffering itself constitutes a preconceived idea that hinders objective discussion. Layman (Reference Layman2003, 2), meanwhile, embraces the unavoidability of ‘mak[ing] controversial assumptions about a series of issues,’ such that there can be no definition of evil that we could expect all reasonable persons to accept.

Even so, Bruce Reichenbach (Reference Reichenbach1979, 207) recognises that at least with respect to what we imagine as the best possible world, much depends on our theory of value: ‘Hedonism suggests states of affairs which produce pleasure; utilitarianism suggests utility; eudaemonism, happiness. Other possibilities would include the production of the most virtue, moral goodness, or goodwill in the world.’ O’Connor (Reference O’Connor1998, 51) notes that ‘if our conception of human goodness is broadly utilitarian, then we will tend to see God as the supreme maximizer of utility; if our conception of human goodness is other than utilitarian, so will be our conception of the goodness of God.’ On this point, Marilyn McCord Adams (Reference Adams1989, 298) bluntly states that ‘Christians never believed God was a pleasure maximizer anyway.’ Indeed, what Mackie (Reference Mackie1955, 203, 204) considers fallacious solutions to the problem of evil utilise ‘equivocation’ and ‘vagueness’ to surreptitiously ignore his quasi-logical rule of evil’s opposition to good, and the (supposed) fact ‘that good and evil are…qualities of things.’ On account of this, he thinks that neither greater-good theodicies nor Alexander Pope mean what they say when they refer to omnipotence or evil, respectively (Mackie Reference Mackie1955, 202, 204). Therefore, definitions matter. What evil is, or how to define it axiologically, appears important for the problem of evil, even if it is not something that has been dealt with extensively in the philosophy of religion.

Some refinements of evil as suffering exist in the literature, however. Paul Draper (Reference Draper1989, 336) asserts that, unlike other biological or teleological organic functions, pain and pleasure are intrinsically valuable, pain negatively so. Beyond that, there is disagreement about the level of intensity required for pains to constitute suffering or, for that matter, evil. For Trakakis (Reference Trakakis2005, 35, 56), natural evil is physical pain of any magnitude, including that of a minor toothache. For Layman (Reference Layman2003, 10), however, ‘a desire must have a certain intensity before an awareness of its non-satisfaction can count as something we would naturally call suffering,’ although it is not easy to specify where that line is. Yet while the notion of suffering (in English, at least) implies a level of intensity beyond mere pain, it is difficult to find suggestions in the philosophy of religion that evil is also more intense than mere badness. Richard Swinburne (Reference Swinburne1998, 163) says that pain ‘caused by humans (and not negligently allowed to occur by them)…is a natural evil because it is a sensation strongly disliked,’ and that any ‘state of affairs…disliked as strongly, would be just as bad.’ Pain, if it is disliked strongly enough, is evil. That it may have ‘just as much badness’ as a non-painful but equally disliked state suggests that evil is a particular degree of badness determined by strength of dislike. Layman says that ‘physical suffering’ is having a ‘strong unsatisfied desire’ that a sensation cease (or, as he adds parenthetically, ‘a strong unsatisfied instinctual drive to avoid it’) and being aware of the dissatisfaction (Layman Reference Layman2003, 10). Presumably pain is a less strong desire or instinctual drive for a sensation to cease. Regardless, for all the philosophers mentioned, what determines X as such – i.e. the problematic thing in the problem of evil – appears to be the phenomenologically immediate or instinctually negative self-assessment of subjective states.

Objective and subjective suffering

Mayerfeld, who works in political science rather than the philosophy of religion, takes note of this subject-relative ‘mundane sense’ of evil when he helpfully distinguishes between two kinds of suffering. There is suffering in the objective respect, and suffering in the subjective respect. The former equates to tragic events, misfortunes, and the like, whereas the latter is constituted by the psychological experience of pain (Mayerfeld Reference Mayerfeld1999, 11-12). The distinction between objective and subjective suffering is not the same as between pain and suffering. Some philosophers of mind distinguish ‘between “mere pain”: a purely sensory experience, and “suffering”: the state one’s in when one minds or is disturbed by their pain,’ but in both mere pain and suffering, the subject is psychologically conscious of the negative stimulus. In the latter case, though, suffering is when the sensation passes the threshold whereby the subject ‘cares about it’ (Johannsen Reference Johannsen2021, 12-13, 24-25). With objective suffering, by contrast, the subject of the tragedy could experience nothing at all. While it is interesting to consider whether reflex nociception in non-sentient organisms would constitute objective suffering in Mayerfeld’s sense (i.e. whether it is a tragic occurrence when an organism responds to noxious stimuli even though the organism has no awareness of it), what matters for our purposes is that both mere pain and ‘disturbing’ pain (aka suffering) would constitute subjective suffering in Mayerfeld’s sense. Hereafter, we use the terms ‘pain’ and ‘suffering’ interchangeably.

Objective and subjective suffering share some elements in common, but have significant differences as well. If X is equivalent to objective suffering, then we can consider all cases of premature painless deaths as evils. If X is equivalent to subjective suffering, then we would have to say that anything between a minor toothache and being tortured over a long period of time is evil (so long as they’re sufficiently disliked), but not painless premature death since it is painless. Additionally, since objective suffering correlates with misfortune, if something is a common experience among people, then it would not be considered a misfortune. This means objective suffering would exclude common, or even ‘naturally’Footnote ¹ experienced instances of psychological suffering, such as growing pains, toothaches, and period cramps. On the other hand, being tortured would be considered objective suffering in addition to subjective suffering, since it is both painful and unfortunate. Whether or not suffering of either type is X or, in the words of Diogenes Allen (Reference Allen1980, 444), constitutes ‘evidence that runs counter to a theistic world view’ is yet to be determined, even though Mayerfeld (Reference Mayerfeld1999, 11-12) claims that the subjective suffering of ‘feel[ing] bad’ is a ‘paramount evil’ with ‘towering moral significance….’

What we have established, however, is that philosophers of religion generally equate X with suffering. Next, we should consider exactly which kind of suffering it is. We argue that X cannot be subjective suffering. Philosophers generally admit natural disasters to be natural X, but disagree on why they qualify as such. If natural X were about pain but not painless misfortunes, then we would have to say that Trakakis’s minor (or appropriately intense) toothache is better evidence against the existence of God than painless premature or instantaneous deaths caused by lightning strikes or falling debris. The latter might only be considered X only as far as it causes psychological pain to the dead’s loved ones, but is not X in and of itself. This compels one to say that the appropriate lament in this context is ‘How could God allow me to suffer from the death of my child’ rather than ‘How could God allow my child to die?’ If that is the case, the solution to it would be either God creates a world in which one does not feel pain from the painless death of their child, or the child dies painlessly without anybody knowing and hence nobody is emotionally affected by the loss. Thus, a lonely but painless death that no one mourns (which is difficult to imagine, but not impossible for an omnipotent being to actualise) is not an X, but a toothache is. This is highly implausible. Therefore, X must be more than just subjective pain. This does not mean that pain is absent from all notions of X, but that pain, which constitutes subjective suffering, itself is unnecessary and insufficient to define X. What we have left as the contender for what X actually refers to is objective suffering, which is closely related to the concept of misfortune.

Misfortunes, not evils

We will now show why any direct response to the problem of X is about misfortune instead of evil. X has to encompass both natural Xs and moral Xs, because philosophers of religion have generally assumed that both types are parts of the problem of evil. What does becoming permanently disabled because of an earthquake and being poisoned by my spouse have in common? The former would commonly be accepted as a natural X whereas the latter would be commonly accepted as a moral X. Both of them are misfortunate events for the victim. The adjective ‘misfortunate’ is predicated to the victim. It is I who is misfortunate, regardless of who or how the misfortune is brought about. This is how X can be understood as a unified category. If we were to understand X as evil, then I would be a victim of evil in both cases. However, when we look at me being poisoned by my spouse and say that is evil, it is natural for us to predicate the adjective ‘evil’ to my spouse. It is my spouse who is evil, hence evil has befallen me. But if I were to say that my spouse is kind-hearted and only unintentionally caused food poisoning by using expired ingredients, then one might hesitate to say that my spouse is evil after all. If my spouse were not evil, was what had befallen me still evil? Not a moral evil. It does not seem to be the case that I could experience a moral evil without there being a morally evil agent.

Perhaps my spouse mistakenly poisoning me is a natural evil? David O’Connor (Reference O’Connor1998, 10) would consider this event as NE∼RNP (natural evil not resulting from natural processes) rather than NERNP (natural evil resulting from natural processes), but even so, neither my spouse nor anyone else is blameworthy for my mistakenly being poisoned. If it is evil, it is strange that evil has befallen me without there being an evil actor. This suggests that an equivocation occurs in the shift from moral to natural evil. For moral evil, the evil resides with the cause – the malintending agent – but with natural evil, the evil resides only in the effect – the unfortunate victim. Put another way, moral evil pertains to the grammatical subject of the action, whereas natural evil pertains to the grammatical object of the action. It might be said that the locus of evil can be both the subject and the object in the case of moral evil, but that is implausible, because victims of intentional poisoning are not evil thereby. Moreover, it cannot be the case that natural evil pertains to both the subject and the object of the action. The entire purpose of the concept of natural evil is to remove blameworthiness from the amoral subject of the action. As such, the meaning of ‘evil’ shifts depending on whether it is predicated by ‘moral’ or ‘natural.’

Similarly, who exactly is evil in the case of an earthquake? Apart from God, there is no morally evil agent intentionally using the earthquake to cause my permanent disability. Even with a deity, it’s not clear that the earthquake occurs in order to cause permanent disability. Nor is there any sense in saying that there was an evil natural agent – such as an earthquake – unintentionally or innocently causing my permanent disability. An evil though innocent natural agent is conceptually incoherent. We conclude, therefore, that evil is characteristic of an actor whose actions some being suffers from,Footnote ² but not characteristic of something that some being suffers independently of a corresponding evil actor’s action. Therefore, evil describes the state of the actor (and only derivatively, that actor’s actions), not the state of the one acted upon. However, the appropriate term for X should work whether or not we have an actor. The actor is unimportant since the problem of X is a problem of the experience of the victim as they cry out ‘how could God exist if I were allowed to experience this and that?’ It is unfortunate for me to experience bad things, but bad experiences themselves are not necessarily evil.

To rephrase the argument above, evil necessarily implies an evil doer. That means if we were to examine the terms ‘natural evil’ and ‘moral evil,’ we would realise that the former is an oxymoron whereas the latter is redundant. Natural evils are supposed to be evils by natural or non-intentional causes. If no intentional agents have caused something misfortunate to happen, there is no one to take the blame of being evil. Nature itself cannot be evil because nature lacks moral agency. The only way to maintain the term natural evil is to say that God is the evil agent since God either allowed or caused natural evils to occur. Yet the problem of evil states that evils exist, therefore God does not, since evil is in direct contradiction with the benevolence of God. In the case of natural evils, if God does not exist, there is no one to blame for natural evils. To maintain the term ‘natural evils,’ either God has to coexist with natural evils or there are no actualised natural evils. Either case does not disprove God’s existence as implied by the problem of evil. One could even say that the term presupposes God’s existence, while the argument using that term attempts to disprove that very presupposition. Therefore, the concept of natural evil implies a contradiction.

Put another way, the problem of natural evil is itself a paradox. The problem of natural evil can only exist if God exists, because natural evil (qua evil) can only exist if God exists. But given the problem of natural evil, if natural evil exists, God cannot. If God cannot exist, then neither can natural evil, and so no problem was possible in the first place. Like the liar’s paradox, with respect to natural evil, ‘God exists’ is false only if ‘God exists’ is true. To presuppose, therefore, that natural evils can be evils because they can be attributed to a divine perpetrator (who would be evil) is to beg the question. To eliminate the paradox, one must deny that natural Xs are evil, with the remaining option being that they are misfortunes.

On the other hand, it is redundant to predicate ‘moral’ to ‘evil’ because evils are necessarily (im)moral. Only moral agents are capable of doing evil things, therefore all actually evil things are caused by moral agents. This, however, disqualifies evil from being what X is, since X has to encompass both natural X and moral X. Whatever natural X turns out to be, it cannot be ‘natural evil’ – the incoherent concept – for there are no agents available to make it evil. If, however, the problem of X were restricted to moral evils only, this would disregard all the discourse on the problem that has already taken place using the problematic language of natural evil, even though the intuitions underlying that discourse still need to be addressed. Therefore, we propose identifying X as misfortunes, however they are caused.

The understanding of X as misfortunes does not lead us to intuitively shift the blame to God as the previous understanding of X as evil led us to do. Good is directly opposite of evil, therefore Mackie’s assumption that a good being eradicates evil as much as possible makes sense at least linguistically. This, however, does not apply to misfortunes. Goodness is not quasi-logically or intuitively opposed to misfortunes like the good-evil pairing is. Furthermore, the understanding of X as misfortunes is not an arbitrary linguistic imposition. We can see such an understanding in the Chinese cognate expressions for ‘natural evil’ and ‘moral evil.’ The expression is 天灾人祸 [tiān zāi rén huò], which literally translates into heavenly calamities and human calamities, and can be more idiosyncratically translated as natural calamities and man-made calamities. The idea of a calamity is synonymous with a disaster, and is inseparable from notions of misfortune. It is an expression that refers to unfortunate events that lead to losses in general, with the former referring to those with non-human causes and the latter to those with human causes. Calamity is closest to Mayerfield’s idea of objective suffering. The expression in and of itself does not presuppose anything about God, which we saw is a fault of the term ‘natural evil.’ Man-made calamities also do not imply anything about the person or people who caused such calamities to happen. They may be caused by evil intentions or not, and so not all humanly caused misfortunes are evil. The expression is simply a statement about what happens to the victims, which may or may not include psychological suffering. This more appropriately captures what the problem of X is all about, when expressed in terms of lament.

The problem of evil, then, should be recast as the problem of misfortunes: ‘why would an all-good and all-powerful deity allow misfortunes to befall me?’ From this lament, one inquires as to the character or motivations of the deity. However, there is no logical or evidential implication that an all-good and all-powerful deity cannot exist on account of misfortunes alone. Only on the basis of additional premises might one draw that conclusion. One might seek comfort or reassurance instead, potentially from the deity itself. The problem of misfortunes is still a problem, but is less a classically philosophical one, and more of a psychological, spiritual, or therapeutic problem.

What is evil?

The term ‘evil’ as it is used in the philosophy of religion is therefore ambiguous,Footnote ³ and we think the problem of evil gets at least some of its plausibility from equivocating on this ambiguity. In ‘moral evil,’ the evil pertains to the actor (and, derivatively, the actor’s actions), whereas with ‘natural evil,’ the evil is not derivative of an actor but rather describes the effect on a victim. The first sense is, we think, the more intuitive understanding of evil, for few people are likely to call an earthquake evil, but if the first sense slides into the second by switching its grammatical location from agent to recipient, ‘natural evil’ arises, borrowing its valence from moral evil.Footnote ⁴

Evil is also a morally ambiguous term because it can be used either as an equivalent with badness or specifically as an extreme form of badness. For this reason, David Oderberg (Reference Oderberg, Hüntelmann and Hattler2014, 63 n.2; cf. Reference Oderberg2015, 84 n.15) has to define his use of the word evil ‘to exclude any connotation of essential heinousness or egregiousness that attaches to the contemporary sense of the term.’ The scholastic sense of malum he relies on may not have had any associations with heinousness or egregious wickedness, but etymologically the distinction between evil and badness dates to Middle English, which obviously precedes modern English.Footnote ⁵ The distinction is also present in 19th century GermanFootnote ⁶ and, as we shall see below with Jean Nabert, in contemporary French. In stipulating evil as not extreme badness, Oderberg recognises that evil is a multivalent concept, but it remains to be seen how it should be used in the philosophy of religion. These ambiguities require us to determine who or what can be designated evil, and if evil is any badness at all or just very bad badness?

We think an equation of evil with badness in general robs it of much of its axiological bite. Oderberg (Reference Oderberg, Hüntelmann and Hattler2014, 64) defends a privation theory of evil, where an evil is any ‘absence of something on which some aspect of the world has what we might call a prior claim or title….’ However, he does not think that all of these privations are ‘essentially evaluative,’ for example deafness or disease. Elsewhere, he refers to ‘pre-moral evil’ of which ‘immoral acts’ are but a species (Oderberg Reference Oderberg2015, 84, 86). What this amounts to are non-moral, non-evaluative ‘evils’ that may not constitute X. It is, at least, not immediately clear why something non-evaluative should be thought to generate a problem for the existence of an omnibenevolent being. Trakakis’s toothache is indeed a privation, but it may only generate a ‘problem of minor irritations’ that a deity may not be required to address.

Oderberg’s terminology is consistent with Rowe’s, however, for the latter is able to countenance justifiable evils:

Here, not only is ‘evil’ a state of the recipient(s) rather than an actor, evil is also something non-heinous, non-egregious, or non-wicked enough to be justifiable. As such, it is difficult to see evil as a term of strong disapprobation. It allows one to claim that ‘I will commit this evil action, but my doing so is morally justified.’ On Oderberg and Rowe’s definition, it is not necessarily wrong to bring about intrinsically evil effects. On the other hand, if evil is a strong, even radical term of disapprobation, then evil may in fact be constituted (in part) by a belief that ‘evils’ are justified.

So is evil the mere privation of a subject (some of which are non-evaluative and some of which may be justifiable), or is it something far more serious than that? For the philosophy of religion, we suggest the latter. What Oderberg and Rowe identify as an ‘evil’ is more appropriately called ‘badness,’ and to that end, we shall consider what precisely and essentially constitutes evil. To recapitulate thus far: we first distinguished between subjective and objective suffering. Subjective suffering is the disturbing experience of (generally agonising) pain, which may be present in objective suffering, while objective suffering is coextensive with tragedy, calamity, or misfortune. However, not all tragic, calamitous, or misfortunate events are painful, and so there may be painless instances of objective suffering. Inversely, not all objectively negative events are tragic, calamitous, or misfortunate, and so there may be instances of painless or painful negativities (e.g. Oderberg’s non-evaluative privations) that do not constitute objective suffering. Therefore, objective suffering would exclude Trakakis’s minor toothache but include painless premature death.

We used the term ‘misfortune’ to refer to objective suffering, but neither subjective suffering nor painless privation were necessary conditions for it. They are likely disjunctively necessary, but insufficient for misfortune, for a misfortune is discontinuous with the common or natural course of life on Earth. Having a toothache is not a misfortune, nor is aging; getting struck by lightning or being accidentally poisoned by a loved one is. There can be natural misfortunes, then, either misfortunes with natural causes (e.g. ‘acts of God’ in the insurance industry, or NERNP in O’Connor’s terminology) or misfortunes with accidental causes (e.g. accidental food poisoning, or NE∼RNP in O’Connor’s terminology). Natural misfortunes as natural X is still a possibility for the philosophy of religion, but not natural evil, because evil requires an evildoer. We thus determined that the problem of X deals with the reality of misfortunes befalling someone or something, regardless of how those misfortunes arose. Only now are we prepared to analyse misfortunes that come about intentionally.

John Rawls (1973, 439, quoted in Dews Reference Dews2008, 4) distinguishes between a bad person and an evil person. Both of them cause their victims to experience social injustice. The bad person does so out of a ‘desire for arbitrary power because he enjoys the sense of mastery’ that such exercise of power gives them. In contrast, the evil person causes social injustice out of ‘the love of injustice.’ The evil person is depicted as one who gains first-order (or immediate) pleasures directly from the humiliation and powerlessness of their victim and also in their victim’s recognising them as the source of their misery, whereas the bad person gains second-order (or indirect) pleasures for which the victim’s misfortunes are but a means. Similarly, Arthur Schopenhauer describes the bad person as one who causes misfortunes to others out of self-interest or what he calls egoistic desire. Suffering ‘caused to others are merely a means, not the end,’ and is therefore only a by-product of the bad person’s actions. In contrast, Schopenhauer (Schopenhauer, On the Basis of Morality, §14) ascribes malicious desire to an evil person, causing misfortunes to others as an end in itself, often accompanied by their enjoyment of the victims’ suffering. Both Rawls and Schopenhauer capture a distinction between two different kinds of moral wrongdoing, one for self-gain, and the other for the suffering of their victim. The bad person is egoistic, whereas the evil person is malicious. Rawls and Schopenhauer’ definitions track significant distinctions between badness and evil that are typically lost in the philosophy of religion, but sometimes found in common usage of these terms.

For Oderberg, however, it is impossible to aim at evil qua evil as, metaphysically, evil has no being and cannot be an object of orientation. There is no pure evil as all evil depends on the good it deprives. It would appear, then, that the Rawlsian/Schopenhauerian definition of evil cannot be actualised, for no one could possibly pursue a misfortune for itself. However, in Oderberg’s view, evil can be directly intended – in a manner of speaking. Oderberg thinks that all immoral action aims at some apparent good which is not the good that should be aimed at in that context. For him, a genocidal tyrant and an obese shut-in that refuses to take care of their health are both immoral, even though they both intend some supposed goods (e.g. racial purity or leisure). But neither of them intend ‘evil as evil [as privation] for the sake of evil [as privation]’ (Oderberg Reference Oderberg2015, 92). And yet, Oderberg (90) says that ‘a person can truly be said to intend evil [only if] they intend to produce some positive reality on which they know the evil supervenes….’ So a person may intend evil (as privation) in the sense of it being a known or anticipated by-product of their pursuit of some other apparent good. That is, an evil person appears unconcerned by the misfortunes their pursuit of certain goods will likely cause.

Integrating Oderberg’s account into Rawls and Schopenhauer, a bad person does not intend or take pleasure in misfortunes themselves, but intends or takes pleasure in something else which the misfortunes make available to the bad person. Badness is the intentional causing of misfortunes to others, acknowledged as bad but accepted as a necessity in the pursuit of some other goal. This acknowledgement can hold even in the face of vast suffering, provided the suffering was acknowledged as bad (not as good) but weighted as a trade off against a (mistaken) greater good. It is morally wrong for someone to cause misfortunes to befall somebody else for their own benefit, but it can at least be comprehended as badly justified.

A fictional example of this bad action could be Marvel’s Thanos. He wiped out half of humanity because he did what he evaluated to be the right thing to do, and at great cost to himself as he had to sacrifice his own beloved daughter in the process. What Thanos did was probably bad, but he was not evil. Schopenhauer goes so far as to claim that egoistic actions are not in themselves morally blameworthy or praiseworthy since they are balancing acts between self-gain and the misfortunes caused as by-products. While much of what is bad is in a dance with what is good and is part of a battle of values (and while Thanos was not an egoist), we think it is still morally blameworthy to be mistaken about weighing misfortunes against some other good. It is more morally blameworthy to lack such a noble motive for destruction, and that is where evil comes in.

For Schopenhauer, malicious actions are morally despicable without question as they cause misfortunes intentionally. While Rawls’s way of differentiating is not exactly the same, he at least feels the need to highlight a certain species of immoral motivation that warrants the term ‘evil’ to distinguish it from mere badness. While we can agree that all actions aim at some apparent good as Oderberg posited, we argue that there are certain apparent goods that when aimed at are practically considered evil, especially if so aiming involves a disregard for the misfortunes caused thereby. At the very least, actions intending to bring misfortune upon others, or what Schopenhauer calls disinterested as opposed to self-interested motivations, are evil. Actions done for sadistic pleasure of seeing another suffer is an obvious contender for this category. Pleasure is a good, but here it is the pleasure of regarding the misfortunes of others as good, not bad. Such pleasure is evil.

Peter Dews (Reference Dews2008, 8) references Jean Nabert’s conception of evil as ‘l’injustifiable,’ paraphrased as ‘that which absolutely should not have occurred.’Footnote ⁷ This captures our earlier point that ‘evil’ is commonly used to denote extreme instances of moral badness, rather than being used interchangeably with it. Schopenhauer (On the Basis of Morality, §14) similarly claims that acting out of malice ‘constitute[s] a greater level of moral depravity’ in comparison to doing so out of egoistic desire, suggesting a difference in degree. This difference of extremity might map onto the difference of intensity between suffering and pain, but Nabert’s absolute unjustifiability suggests a qualitative discontinuity between evil and badness rather than one of degree, and is thus consistent with Rawls’s usage. Evil absolutely should not exist, for there is no comprehensible justification for it. In terms of Rowe, we might say badness could be justified, but not evil. The irrationality and utter unjustifiability of evil, as claimed by Nabert, would assign evil a wholly different ontological status from that of good-and-bad (Dews Reference Dews2008, 8).

By these above definitions, the evil person is somebody that causes misfortune to befall another for the pleasure those misfortunes directly give the evildoer, or with utter disregard towards said misfortunes. It does not matter to an evil person how much bad and how little good their actions cause; it is precisely the bringing about of the misery or calamity of others that is their concern – or, alternatively, no concern of theirs whatsoever. Evil is the intentional causing of misfortunes to others marked by moral erasure – that is, the nullification of another’s worth as a moral subject, thus treating their suffering as insignificant (indifference) or as a good in itself, perhaps something to be enjoyed (sadistic pleasure). Metaphorically, it can be expressed as treating people as toys (whereas a bad person treats people as tools).

A fictional example of this evil action could be DC’s Joker. The Joker is an archetypal evil character because he caused great misfortune to people just to torment his nemesis Batman. He may have gained nothing from these misfortunes except the amusement of seeing Batman suffer to uphold his ideal of not killing his enemies. It is more difficult to give a similar analytical treatment to historical figures like Hitler and Mao who are commonly labelled ‘evil’ because we do not have unhindered access to their motivations and intentions.Footnote ⁸ One can reasonably ask what is the point of deliberating over the intent of various alleged evildoers, given that they cause equivalent negative outcomes. We think one reason to deny this consequentialist view is that it ignores the degree to which the offender is teachable. If the role of moral education is to teach people to associate good and bad to their appropriate objects, then the bad person who mistakes bad for good is simply mistaken (Aristotle, Nicomachean Ethics, 1106b18–24). When we describe people who make ‘innocent mistakes’ or were deceived into doing what they thought were good, the language we use implies they can be corrected. They deserve to be empathised with, since nobody is faultless all their lives. They could be empathised with because we are united by a common pursuit of what is good, and a common respect for moral subjects. Some may be misguided, but that implies they could be good with the right guidance.

However, such treatment is harder to give to the evil person. A person who actively seeks out the misfortune of others is disqualified from being seen as just another human, but a monster to be avoided. One who aims to do good may miss the mark, but one who desires evil rejects the target itself. Unlike the bad person who is misguided, the evil person deliberately chooses evil. One may say that the evil person is mistaking the suffering of others for an apparent good, but that does not reduce the disapprobation of such evil. Assuming the apparent good at work here is sadistic pleasure, it is enjoyable for them to torment others precisely because it is the thing anyone would least desire for themselves – that is what makes it evil. To commit evil implies maliciously doing so with full knowledge of its depravity. For this reason, they are more appropriately described as ‘twisted.’ This is why the focus on intent is significant in distinguishing between badness and evil. The similarities in the consequences of their acts blind us from how radically different the perpetrators are at heart, between the evilly malicious and the badly mistaken person.

With the definition of evil as moral erasure, we could at least say that evil is opposed to good, because a good being (such as God) would not actively seek out and cause, let alone take pleasure in, the suffering of sentient beings. God can have his reasons to allow suffering, but as long as he does not cause suffering for the sake of seeing others suffer, he is not evil. God could be bad if he caused suffering or allowed it to occur for the sake of some illegitimate or mistaken view of the good (for God), but insofar as God is omniscient, this would not happen. God could only be bad if it were presupposed beforehand that God was not omniscient. This leaves us with the remaining problem of X: why would an omnibenevolent, omnipotent, and omniscient deity allow their beloved creatures to suffer misfortunes, however they are caused? Since misfortune is not linguistically or quasi-logically opposed to goodness the way evil is, the problem of X is not a deductive or inductive argument against the existence of such a deity. It is not the question of whether God is an evil or bad agent. It is rather an existential lament: a longing for an explanation of misfortune, the alleviation of suffering, or both.