An Objection to This Whole Supererogation and Obligation Business

Nothing is over the call of duty -- there are just differently weighted moral reasons

Supererogation — a word I’ve only just learned was spelled this way — is the idea that some actions go above and beyond the moral call of duty. Some actions are nice, but they’re not obligatory. For example, maybe it’s morally good to donate a lot of money to charity, but it’s not obligatory — if you don’t donate a lot of money to charity, you’re not violating any moral obligations. Even the non-charitable can be entered into the good people’s club, as long as they don’t steal, murder, or torture anyone.

In contrast, obligations are things that you must do. For example, you’re allegedly obligated not to torture people or randomly assault people. Thus, even if you’re a very nice charitable guy, if you commit Honda-Esque random acts of flogging, you’re excluded from the good people’s club. They will not invite you to the parties — only those who live up to their moral obligations get invited.

I think that this entire way of thinking about morality is wrong. There is no precisely delineated good people’s club, that has its lines at obligations. There are no obligations or supererogation — just actions that are differently worthwhile. As Norcross argues, actions are just rated from best to worst based on ones reasons to take the action — there’s no precise threshold at which something becomes an obligation. Morality comes in degrees.

I’ll present two reasons to think this. Both of these reasons are incredibly persuasive and should convince any rational reader. Though some of my readers may be irrational — for example, I hear some type A physicalists read my blog — and thus they may not be convinced by just one argument, however rationally compelling it may be¹. Thus, I’ll provide a whopping two arguments; if the first one doesn’t work, the second one must! Though if you’re still not convinced, you may just have to read Shelly Kagan’s The Limits of Morality, in which he advances a more complicated argument for the same conclusion.

One first argument for the view that there is no supererogation appeals to an analogy between morality and the epistemic domain or the prudential domain. The prudential domain is about what’s in one’s interests. So, prudentially I ought to do X if X benefits me. The epistemic domain is about what we should rationally believe.

Both cases, like morality, are reducible to reasons. In all three of these cases, there are certain things that provide reasons. For example, the fact that prodding a bear with a stick would result in the bear eating me gives me a prudential reason not to prod the bear with a stick. The fact that Aristotelian metaphysics is vouched for only by medieval astrology gives some reason not to believe Aristotelian metaphysics. In all of these cases, facts provide us with various reasons. In the epistemic case, the stronger the reasons, the greater one’s epistemic justification is for a belief. For example, if there are 500 good arguments for atheism, that gives good reason to be an atheist. But there’s no point at which a belief becomes obligatory or supererogatory. Instead, in the epistemic case, facts just provide us with reasons to hold various beliefs, and those facts are of differing strength.

The same is true of the prudential domain. The fact that some act would lead one to be eaten by a bear gives an even better reason to avoid it than the fact that the act would just result in getting mildly mauled by a bear. The fact that some act would cause one to be so brain damaged that they find type a physicalism plausible gives even stronger reason to avoid it than the fact that it would result in getting eaten by a bear. At no point are there prudential obligations — there are just reasons of different strengths.

Maybe one would want to say that if there’s a belief that no person could reasonably believe it’s epistemically impermissible. Likewise, they could say that if there’s some action that’s obviously prudentially irrational, not doing it is prudentially obligatory. But then this will be a reducible account of obligations. On this account, an obligation in some domain is just a case where the reasons are so strong that you have to be really stupid not to recognize those reasons. But if this is all that supererogation and obligations are, then a utilitarian can believe in supererogation and obligation. As long as an obligation is just a type of scenario in which one has strong moral reasons, and supererogation is a scenario in which we can expect reasonable people to sometimes fail to do what they have most reason to do, the utilitarian can agree with this account.

Thus, supererogation and obligations are analogous in the moral domain to the prudential or epistemic domains. And in those cases, they don’t really exist in any deep, ontological sense — they’re either non-existent or higher-order descriptions that we can give reducible accounts of.

The second argument against obligations and supererogation is the following. It seems obvious that it’s morally worse to take a supererogatory action and an action that one is obligated not to take than to take no action at all. After all, the supererogatory action is nice, but the action that one is obligated not to take can’t be taken under any circumstances — thus, based on the idea of the concepts, it’s worse to take both an action that one is obligated not to take and a supererogatory action than to not act at all.

This seems true if we accept the idea of obligations and supererogation. But it admits of a counterexample. Suppose we say — as almost everyone that beliefs in supererogation does — that donating money above a certain threshold is supererogatory, not obligatory. Thus, a person is deciding whether to take two actions — one of which is supererogatory and one of which is morally impermissible — or to not act at all. The actions in question are

A) Donating 5000 dollars to the malaria consortium to save a child’s life.

B) Beating up a child who was saved by the donation.

It seems obviously better to take both actions than to take neither. If you were advising a friend on which one to take, you should advise they take both rather than neither. And if you were deciding which one to take, you should take both rather than neither. After all, doing nothing leaves the child dead, who would otherwise be beat up but very much alive.

But this contradicts supererogation. If it’s worse to take an impermissible action and a supererogatory action than to do nothing, then it would be worse to do both than to do neither. But this is clearly false. It’s obvious that it’s false when the supposed victim would be dramatically better off if you do both rather than neither.

Thus, the idea of supererogation and obligations are nonsense. They don’t mesh well with our best ideas of reasons. They run into the super secret problems that you’ll have to explore Kagan’s book to find out. And they run into problems with a combination of impermissible and supererogatory actions. We can give a perfectly adequate utilitarian friendly reduction of them. As we should.

1. I am of course kidding — Mr. Coase’s Ghost and I are friends!