Many are attracted to the view that what an agent ought to do depends (in part) upon what the agent is capable of doing:
OIC: If S has an obligation to Φ, S can Φ.
In the hopes of finding a sound defense of OIC, we might turn our attention away from the linguistic evidence and think about guiding reasons and the obligations they create. There is an argument that suggests that OIC follows from independently plausible claims about practical reasons and the relationships between practical reasons, the actions they call for, and the obligations they seem to create:
1. If S has an obligation to Φ, S has a reason to Φ.
2. If S has a reason to Φ, then Φ-ing is a potential action of S’s.
3. If Φ-ing is a potential action of S’s, then S can Φ.
C: If S has an obligation to Φ, S can Φ.
This argument is taken from Vranas' Philosophical Studies article, "I Ought, Therefore I Can" (here).
I suspect that many will be of the opinion that if we focus on just the linguistic considerations taken to cause trouble for OIC, we will end up saying things about reasons we ought not say. For example, we will end up saying that there are reasons for us to perform actions that are not potential actions of ours. Fleshing out the intuition, Vranas writes, “… reasons of any kind are conceptually linked to objects of the given kind: reasons for belief to potential beliefs, reasons for action to potential actions, and so on” (5).
In spite of its initial intuitive appeal, I worry that a premise crucial to the argument’s success is neither a truism about practical reasons nor true. Are reasons for S to Φ invariably reasons for S to perform a potential action? That is not entirely clear. Two examples should suffice for my purposes.
Jones lights a cigarette. In lighting the match to light the cigarette, he gives away his position to a sniper. According to a commonly held view about reasons for action, the reasons that count as reasons to, say, light and smoke a cigarette are considerations that favor lighting and smoking a cigarette. They show that there is something that makes the lighting and smoking worthy of choice, desirable, attractive, etc… Given the circumstances, lighting and smoking a cigarette is a potential action of Jones’. Given the circumstances, lighting and smoking a cigarette without giving away his position to a sniper is not a potential action of Jones’. Given his circumstances, by lighting and smoking the cigarette he will smoke the cigarette and give away his position. The considerations that speak in favor of lighting and smoking, however (e.g., the rich taste, the calming of the nerves, the coolness of the pose he’ll strike in photos while holding the cigarette) do not speak in favor of his giving away his position to a sniper. So, we have a threat to (2). There can be a reason for S to Φ in the very circumstances where that reason is not a reason to Ψ but Φ-ing without Ψ-ing is not a potential action of S’s.
Someone might be picky and say that Jones’ smoking and alerting a sniper to his presence is not an action of his. It is not if Jones knows nothing of the sniper’s presence. This problem is easily fixed. Jones is well aware of the fact that snipers are often watching the trenches for the flickers of light that indicate that an enemy is smoking a cigarette. I do not think it would be an abuse of the language to say that his running the risk of being spotted by a sniper is something he did and did willingly. The considerations that favor smoking a cigarette are not considerations that count in favor of running the risk of being spotted by a sniper. In general, it seems odd to think that the pros that favored Jones’ action also favor Jones’ imposing upon himself the foreseeable costs that come of pursuing that course of action. Reasons to smoke do not become reasons to alert snipers even when those things go hand in hand.
The point of this example is, hopefully, clear enough. When we know we cannot Φ without thereby Ψ-ing, see that there is much to be said in favor of Φ-ing and some things to regret about having to Ψ if one Φ’s, the considerations that favor Φ-ing do not seem to favor bringing about the cons associated with Ψ-ing. I like to jog. I hate wearing down the soles of my shoes. I know I cannot jog without thereby wearing down the soles of my shoes. The reasons that favor jogging do not favor wearing down the soles of my shoes. If I do not jog, I know I have failed to do what the reasons spoke in favor of and reasonably regret that. By not wearing down my shoes in the way I would if I jogged, I do not have the same cause of regret.
Smith is asked to give some alms for the poor. Fishing around in his pockets, he realizes he is completely out of alms. He has some keys, some matches, a few stamps, and a pocketknife. Seeing a passerby, he knows that he could use that knife to brutally kill the man and give his alms to the poor. Reasons associated with the duty of beneficence demand Smith to give alms. Given the circumstances, there is no potential action of Smith’s that is his giving alms to the poor that are rightfully his. Given the circumstances, there is the potential action of Smith’s that is his killing another in order to give alms to the poor. Reasons associated with the duty of non-maleficence and considerations of justice, however, speak strongly against doing that. It seems that while the reasons associated with the duty of beneficence are reasons to give alms to the poor but they are not reasons to give alms to the poor at the expense of considerations having to do with non-maleficence and justice. So, we have a situation in which it seems Smith has reasons to give alms to the poor. The situation is such that Smith cannot give alms to the poor without murdering a passerby in order to give alms to the poor, in which case if Smith gives alms to the poor he will do so at the expense of considerations having to do with justice and non-maleficence. Reasons associated with the duty of beneficence, however, are reasons to promote the welfare of others but not to promote the welfare of others at the expense of justice and non-maleficence. To see this, suppose Smith decides to murder the passerby and distribute his alms to the poor. Can we really say that while he acted wrongly, he did what beneficence asked of him? I think not. However, beneficence did ask of him that he distribute alms to the poor. Again, we have reasons for an agent to perform an action that are not reasons to perform a potential action of his.