English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
Introduction
Doing evil is sometimes the only way to prevent a greater evil. According to moral purism, even in such dire situations doing evil is always impermissible. In this article I review several versions of moral purism and show that they are implausible.
This article is structured as follows. Section II defines ‘evil’ and ‘lesser evil’, and explains the difference between moral purism and the other views on the permissibility of choosing the lesser evil (moral absolutism and permissivism). Section III reviews the two main versions of moral purism (purism about outcomes and purism about acts) and argues that both versions are incomplete, in the sense that they are unable to provide practical guidance in all cases.
Preliminaries
In a broad sense, ‘evil’ refers to any bad state of affairs, wrongful act, or character flaw. In a narrow sense, ‘evil’ refers only to the most morally despicable states of affairs, acts and character traits (Calder, 2015). In this article I employ the broad sense, because I aim to address choices involving not only the worst kinds of acts such as murder, but also less serious wrongs such as lying. From the broad sense I focus exclusively on acts and states of affairs, leaving character traits aside.
Moral purism is just one view within a spectrum of views about the permissibility of doing the lesser evil. The fundamental distinction is between permissivism and non-permissivism. Permissivism is the view that choosing the lesser evil is always permissible. Non-permissivism comprises a range of views that deny that choosing the lesser evil is always permissible. The most extreme form of non-permissivism is moral purism, which claims that you should never choose evil, even if it is the lesser evil.1 Moderate forms of non-permissivism include absolutism (the view that choosing certain kinds of evil is always impermissible, even in those cases in which they are the lesser evil), and limited permissivism (the view that choosing all kinds of lesser evil are permissible, but only if the greater evil is substantially greater).2
What does it mean that an option is the lesser evil? Since ‘evil’ can refer to either wrongful acts or bad states of affairs, the idea of a ‘lesser evil’ is ambiguous. When an agent faces a choice between two options A and B, option A can be the lesser evil in at least two senses:
(1) option A results in a bad state of affairs, but option B results in an even worse state of affairs. (Both options are not wrongful, but simply bad)
(2) options A and B are wrongful, but option B is even more wrongful.
There is however a third category of ‘mixed’ cases, i.e., cases in which one option is the lesser evil in one sense, while the other option is the lesser evil in the other sense:
(3) option A is wrongful, while option B is not wrongful but right; option A results in a bad state of affairs, but option B results in an even worse state of affairs.
The first kind of lesser-evil choice-between two bad outcomes-can be illustrated with a choice between amputating your infected toe or losing the whole foot. The first option brings about a bad state of affairs (loss of a toe) and the other option brings about an even worse state of affairs (loss of the whole foot). Each act is permissible, so what makes this a lesser-evil choice is the outcomes, not the acts.3
The second kind of choice-between two wrong acts-can be illustrated with the following example. A couple has two adolescent girls. The youngest has been diagnosed with a mental illness. The illness is hard to treat, especially because it requires from the parents a lot of initiative and persistence. After a while the healthy girl notices that her parents are not doing all that is necessary. She realizes that their failure is not due to ill will, but simply to the fact that the situation would be very hard to manage for any human being. She tries to convince them of taking more action by reminding them that postponing proper treatment is very dangerous for her sister. The parents do not react. She finally realizes that the only means she has left to protect her sister (short of requesting a social worker) is to try to get to them emotionally by using deception: she could fake to be crying in despair, she could exaggerate the danger her sister is under, she could falsely threaten to leave the house if they do not react, etc. In less extreme situations those actions would be clearly wrong. In this particular case, however, those actions are the lesser evil.
In this lesser-evil choice the healthy daughter has two options: lying to the parents, which is wrong and it also produces a bad state of affairs (because the lying daughter may experience painful guilt after the lie, the bond between her and the parents may be harmed if the lie comes to light, etc.); and not lying to them, which is even more wrong (because, let us assume, it implies neglecting urgent duties of care towards the ill sister) and it results in an even worse state of affairs (mental health deteriorates and eventually ruins her life). The first option is the lesser evil in two senses: in terms of the act it involves, and in terms of the resulting state of affairs.
The third kind of choice-between a wrong act and a bad outcome-can be illustrated with a case in which your neighbour is away and you can save her lettuce plantation from being devoured by a hare only if you trespass in order to drive the hare away. Trespassing is wrong, while not trespassing is right (you do not infringe any duties) but it results in a bad state of affairs (the plantation is lost).
Versions of Purism and their Problems
Purism claims that you must never choose evil, even if it is the lesser evil. As explained above, ‘evil’ can be either a wrongful act or a bad state of affairs. So there are different versions of purism depending on what kind of lesser evil it refers to. In this section I argue that no version of the purist claim is plausible.
Purism about outcomes
If we interpret purism as a view about choices that bring about bad states of affairs, it would look like this:
Purism1: you should never choose an option that results in a bad state of affairs, even if all alternative options result in even worse states of affairs.
To see the problems with this view suppose that two glasses are rolling on the table and about to fall, and you can only stop one. One of them is very expensive, the other one is cheap. Purism1 implies that you are not allowed to save the expensive glass because this option results in a bad state of affairs (the other glass breaks). For the same reason, you are not allowed to save the cheap glass from falling either. Purism1 even implies that in this situation you are not allowed to refrain from acting, since doing nothing would result in a bad state of affairs too (it results in both glasses breaking). So purism1 is an incomplete view, in the sense of being unable to provide practical guidance for all possible cases.
Perhaps purism1 can be saved from this objection by using the idea of causation: ‘you should never choose an option that causes a bad state of affairs, even if all alternative options result in even worse states of affairs’. Since in the example you do not cause the cheaper glass to fall, under this interpretation purism1 does not imply that you are not allowed to save the expensive glass, and it does not imply that whatever you choose you act wrong. However, it is easy to think of situations in which morality clearly allows us to cause bad states of affairs. For example I am allowed to buy the last bottle of juice from the store even if that causes your desire for juice to be frustrated. What matters morally is whether what you choose is wrong, not merely whether it causes bad states of affairs.
Purism about acts
Purism is more plausible if it refers to acts and not (only) to the resulting states of affairs. Consider then this new version:
Purism2: you should never choose a wrong act, even if all alternative options result in worse states of affairs.
In order to assess if purism2 is plausible we need to state it in more precise terms. Purism2 says that you should never choose wrong actions, but what sense of ‘wrong’ is being used here? An action can be all-things-considered wrong, pro tanto wrong, or prima facie wrong. When an action is all-things-considered wrong, there is no further claim to be made about its wrongness: the action is absolutely wrong. But when an action is pro tanto or prima facie wrong, there may be further considerations that affect whether that action is all-things-considered wrong. In the case of pro tanto wrongs, there may be competing pro tanto wrongs that need to be weighed against each other in order to determine a winner. In the case of prima facie wrongs, conditions may arise under which their wrongness is undermined (Reisner, 2013, pp. 1-3). I focus here on prima facie wrongs, but all of what I say applies to pro tanto wrongs as well.
If by ‘wrong’ purism2 means ‘all-things-considered wrong’, then purism2 is trivial: all theories are purist in that sense, because no theory claims that you are allowed to do what is all-things-considered wrong. Even consequentialism would deny that you are allowed to do what is all-things-considered wrong in order to bring about the best consequences. Consequentialism would simply claim that the set of actions that are both all-things-considered wrong and that bring about the best available state of affairs is an empty set. So even consequentialism could accept purism2.
Maybe by ‘wrong’ purism2 means ‘prima facie wrong’:
Purism3: you should never choose a wrong act, even if it is only prima facie wrong and all alternative options result in worse states of affairs.
This version of purism is rather obscure. As explained above prima facie wrong actions are actions that are wrong only in certain circumstances but not in others (when their wrongness is undermined). So it is not clear why purism3 would insist that we should never do those actions even in those other circumstances in which their wrongness is undermined.
Purism3 makes more sense if we interpret it as saying that prima facie duties cannot be undermined merely by the fact that they bring about bad states of affairs:
Purism4: you should never choose a prima facie wrong act merely because all alternative options result in worse states of affairs.
Purism4 does not claim that prima facie wrongs should never be chosen, but only that they should never be chosen to prevent something that is merely bad but not wrong. For example, suppose a very promising young violinist is about to drop her career because she cannot afford a decent violin. You have the opportunity to get her one by stealing it from a chain music store that would barely notice the violin’s absence. Letting her abandon her career would not be wrong, although it would bring about a bad state of affairs. Stealing the violin would bring about a less bad state of affairs, but it would be at least prima facie wrong. So purism4 claims that you are not allowed to steal the violin.
Purism4 is more plausible than the other versions. The problem again is that no one would deny it. No one would claim that you are allowed to do P (something wrong) for the sake of preventing Q (something bad) from happening, when it would not be wrong to let Q happen. Not even consequentialism would claim this. Consequentialists would say that if one option brings about the worse state of affairs, then it would be wrong to choose that option. But this claim is not incompatible with purism4 because purism4 is only about cases in which a choice must be made between a wrong option and a merely bad option (of course, for consequentialism this is an empty set).
Finally, a problem for all versions of purism (about acts) is that agents often face duty conflicts. Duty conflicts are choices in which all options are wrong in some sense (prima facie or pro tanto). In the sisters case, for example, I believe the healthy sister not only has a duty not to lie to her parents, but also a duty to promote her sister’s health. But she cannot fully comply with both at the same time. It is not clear what a purist would say about cases like that. Perhaps something like this:
Purism5: you should never choose a prima facie wrong act, even if all alternative options are prima facie wrong (and they are either less wrong or bring about less bad states of affairs).
The problem with this version of purism is that it is incomplete (it does not provide an answer to all cases). In cases in which two prima facie wrongs conflict it is impossible not to choose a prima facie wrong, so all choices would be wrong. Another problem with purism5 is that in a choice between two wrong actions it would be unreasonable to choose the wrong action that is more wrong or that brings about a state of affairs that is worse than all alternatives. If you are inevitably going to choose something wrong, why not do it in a way that minimizes wrongness or bad consequences?4
The purist may try to resist the incompleteness charge by claiming that there are no cases in which an agent must choose between prima facie duties, because there is no such thing as a duty conflict. This would be true if (a) all our general duties are negative duties (i.e. duties requiring omissions, not actions), because purely negative duties cannot conflict; if (b) we do have positive general duties but they are all imperfect, so a proper conflict between them and our perfect negative duties is impossible; or if (c) specificationism about moral obligation is defensible, and each duty has completely circumscribed conditions of application so that it never conflicts with another duty. I assume that the three claims are implausible.5
Concluding remarks
Moral purism is implausible in its two main versions. The first version claims that it is always impermissible to choose an option that results in a bad states of affairs, while the second version claims that it is always impermissible to choose an option that is even prima facie wrong. I have argued that both versions are incomplete, in the sense that they are unable to provide practical guidance for cases in which all available options result in bad states of affairs, or in which all available options are prima facie wrong.
