Bernard Williams discusses the problem with having conflicting obligations and a potential contradiction that might be accepted if we accept conflicting obligations. He argues that two things we ought to do can conflict because choosing to do one of them will not absolve us of the fact that we also ought to do the other. A closer look at the logic involved can reveal why no contradiction will be accepted. I will present three objections to his argument. The first questions whether or not obligations can conflict, the second admits that nonmoral “oughts” might conflict while moral ones don’t, and the third questions whether or not we can reject that beliefs conflict but accept that “oughts” can conflict.
The problem is that sometimes we can’t do both things that we ought to do. Perhaps we made a promise to someone to meet them at 5:00 pm and we ought to call the police to help catch a criminal at 4:50 pm. If you call the police, then you will have to help the police with an investigation that will make you miss your 5:00 meeting. In other words, we ought to call the police and we ought to get to the 5:00 meeting. But since it is impossible to do both of these things, we ought not do one of them. (We only ought to do what is possible.) We now have found a contradiction: We ought to do two different things, but we ought not to do both of those things. (A contradiction is found because something cannot be true and false at the same time in the same respect.)
Here is another way of forming this argument:
- I ought to do A and B.
- It is impossible to do A and B.
- It can’t be the case that I ought to do what is impossible.
- So, it is false that I ought do A and B.
- Therefore, a contradiction is found. I ought to do A and B, but it is false that I ought do A and B (53).
The question is, can it be true that we ought to do two different things, even if we cannot do both? If not, then how do we avoid the conflict?
Simply put, Williams agrees that it can be true that we ought to do two different things, even if it is impossible to do both. The discovery that two “oughts” (or obligations) conflict does not weaken either of them (46). Also, if we decide to live up to one obligation rather than another, it is a cause for regret. We can try to “make up” for the decision (47). In other words, Williams agrees that two obligations can conflict. Realizing that two “oughts” conflict does not allow us to absolve ourselves of one of them. If a person felt regret about acting on one “ought” rather than another, it certainly wouldn’t prove that person to be morally inferior.
On the contrary to Williams’s understanding, he discusses the fact that most philosophers do not agree that two “oughts” can conflict (49). They have said that we can choose one “ought” and be absolved of the other. “[T]hey eliminate from the scene the ‘ought’ that is not acted upon” (49). Philosophers have adopted our belief that two conflicting beliefs cannot be true, so two “oughts” cannot be true. One belief (or “ought”) must be rejected.
Williams argues that when we have conflicting “oughts” we might learn to avoid situations where we have conflicting “oughts” precisely because they both apply (50). Anyone who wants absolve everyone from having conflicting “oughts” will have to ignore the fact that we might learn a lesson to avoid these situations.
Williams argues that the problem with the apparent contradiction is that we cannot do actions to satisfy both of our “oughts,” but we ought to do two different things taken in isolation. I ought to do A and B, but I ought not do both A and B at the same time. (i.e. I ought to call the police and separately I ought to get to the 5:00 meeting, but I ought not do both of these things at once. Or consider Williams’s example: Tom ought to marry Susan and he ought to marry Joan, but he doesn’t think he ought to marry both Susan and Joan) (54).
Williams concludes that although it is true that “if I do A, then I ought not do B” in the sense that I ought not do both A and B (56). However, I am not entirely absolved of doing B in this situation because it is my fault that I chose to do A in the first place.
Williams then considers an objection: When I need to make a decision whether to act on one “ought” or another, no one would say that I ought to do both. I will have to ultimately decide that I should do one rather than the other. It would seem that I have to be absolved of acting on one “ought” in order to decide to act on the other.
Williams replies to the objection that to decide to act on one “ought” is not an affirmation that it really is something that I ought to do (57). To decide to act on one “ought” does not involve a denial that one of the “ought” statements is false. It is a new moral judgment that I ought to act on one of my “oughts” rather than another.
To conclude, Williams resolves the problem of contradiction by arguing that it is false that you are absolved from doing something you ought to do just because it conflicts with something else you ought to do. It is true that I ought to do each A and B, but I ought not do both because that is impossible.
My Objections to Williams’s Argument
Objection 1: Can “oughts” really conflict?
In order to have conflicting “oughts” we must first accept a pluralism of conflicting values. Aristotle and utilitarians found that happiness is the only value, so we can never have a conflicing “ought.” The Stoics found virtue to be a value of uncompromising importance. If one action does the most good, then that is what we ought to do. If two actions do the same amount of good, then it is morally arbitrary.
In order to have conflicting “oughts,” we must live by “moral rules.” Even if we accept a pluralism of values, it isn’t clear why two “oughts” will conflict. If we ought to do whatever does the most amount of good, and two actions each do approximately the same amount of good, then why “ought” we to do each of them? One way might be to accept “moral rules.” If we find general moral rules that we ought to live by, then certainly those rules could cause conflicts. “Thou shalt not steal” could potentially conflict with “thou shalt stay healthy.” (Both of these commands are accepted by Kant.) Moral rules are suspicious because they are an abstract attempt to capture reality while ignoring the complexity of the situation. It might be true that insofar as property rights make people happy, we should not steal. However, what really matters in this case is what makes people happy, so we can find exceptions to the rule. Moral rules could be abstract rules of thumb, but the only moral rule could be something like “do the most good.” (It might be a bit severe to require that everyone do the most good, but it could be a form of advice.)
Williams claims that it is consistent with a view of morality that a truly admirable person could feel regret about acting on one “ought” rather than the other, but we might still find that a truly admirable person wouldn’t. Here a “truly admirable” person would be someone who fully understands ethics and has corresponding emotions to that understanding. However, we can reject regret if we find out that no “oughts” can conflict. (We can still “make up” to anyone harmed by the decision if that in itself will be “doing good,” perhaps in the sense of making people feel better who believe they have been wronged.)
What about how Williams said that we can learn not to put ourselves in an avoidable situation in which we can’t satisfy all of our “oughts?” Anyone who doesn’t believe in conflicting “oughts” could simply reply that we put ourselves in a situation where we couldn’t do as much good. Next time we should put ourselves in a situation where more good deeds could be done.
Objection 2: Conflicting nonmoral “oughts”
It should be noted that the word “ought” is vague, and what Williams said might be true of two non-moral “oughts.” It might be that giving myself obligations or goals could lead to conflicting nonmoral “oughts.” I ought to keep my 5:00 appointment in the sense that it is a goal of mine, but it might lack moral importance. (Of course, there might be a slight importance involving the happiness of the person I have an appointment with, but this misses the point.)
Whether or not a killer ought to use a gun to kill someone can have a similar conflict of nonmoral “oughts.” It might be that using a gun or using a knife could both be what the killer “ought” to do to kill someone and be equally effective at doing so, but the killer can’t do both. This is merely the word “ought” being used in the instrumental sense rather than the moral sense.
Williams might be aware of the fact that there are nonmoral “oughts” of this kind, and he might agree that he is using the word “ought” in a loose enough way to discuss these kinds of nonmoral “oughts” in addition to moral “oughts.” The point is that it might be possible for some philosophers to reject that moral “oughts” can conflict, but accept that nonmoral “oughts” can conflict as a trivial truth.
Objection 3: Aren’t “oughts” a kind of belief?
This third objection is one that will fail to be a problem for Williams’s argument.
Williams argues that “oughts” are different from beliefs in the sense that we can have two conflicting “oughts” but we cannot have two conflicting beliefs. We don’t want to allow ourselves to have conflicting beliefs because two conflicting beliefs can’t possibly be true. (Some roses are red cannot be both true and false.) However, Williams argues that we can have two conflicting “oughts.” It might be true that “I ought not to harm people” but it might also be possible that “I ought to protect myself from harm.” In some situations it might be impossible to do both. Isn’t “I ought to not harm people” something that is true or false? Isn’t this a moral belief?
Williams could admit that “oughts” are moral beliefs, but remind us that the fact that two “oughts” can conflict will not lead to a contradiction in and of itself. Williams rejects that conflicting oughts could absolve me of the fact that “I ought not to harm people.” The conflict is not about whether or not it is true that I have an “ought.” Rather, the conflict is one in which I must choose which “ought” to satisfy, so there is no reason to think that it is false that “I ought not to harm people.”