Previously I had developed an argument to the effect that, because stronger prudential reasons can always clash with moral reasons, morality and prudence are relative to each other. [1] I suggested this relativity might manifest in either direction: maybe some karmic force ensures that it is always prudent to be moral, or maybe morality is fundamentally egoistic so that it is always moral to be prudent. Both horns of this dilemma seem implausible, and so I took this to be a problem for the moral realist. My intention for this argument was, in fact, to knock it down, for I think I have come up with an attractive theory of moral obligation that invalidates it. I was wanting to show that my theory was worthwhile, and that by constructing it I had answered an important and difficult metaethical problem.
I was excited, then, to come across Matthew Flannagan critiquing Erik Wielenberg's atheistic moral realism on exactly these grounds, and rallying behind other philosophers in the process. [2] Flannagan quotes William Lane Craig saying, “if God does not exist, then prudential reason and moral reason can and often do come into conflict, in which case there is no reason to act morally rather than in one’s self-interest.” [3] He also cites Layman arguing that God must exist because, if there was no God and no afterlife, then prudential reasons could on occasion override moral reasons, leaving us without decisive reason to do our moral duty. [4] To this Craig says, “I agree with Layman that on atheism, what he calls the overriding thesis (namely that moral value always trumps prudential value) is not true, for one can have extremely strong prudential reasons for not acting morally, and there seems to be no common scale in which to weigh moral against prudential considerations”. [5]
Essentially, I had been defending Flannagan’s criticism of Wielenberg: that atheism and robust moral realism do not quite fit together. It seems implausible, given atheism, that there would be any karmic force ensuring that it's always prudent to be moral. And robust moral realism does not seem compatible with egoism.
Wielenberg's responses don’t quite hit their mark, and so it seems he has been pinned down. It looks as if that atheists and naturalists cannot accommodate robust moral realism. But these first appearances are illusory. As I had mentioned, there is an error in this argument being brought against Wielenberg. True, it may always be possible to have even stronger prudential reasons that clash with ones moral reasons. But that does not, in of itself, entail that those weaker moral reasons can always be defeated. They might be helped along by further normative reasons that defeat those threatening defeaters, and this is exactly what my theory is intended to explain. In the end, I will identify moral obligations with a kind of composite normative state of affairs that amount to indefeasibly decisive moral reasons. This will then explain why we should always be moral. It falls out of the very nature of a moral obligation, on my view, that you necessarily should do whatever is morally obligatory.
The overlooked mistake is that rebutting defeaters are not the only sort of defeater. There are also undercutting defeaters, which are more commonly spoken of in the context of epistemology. Where a rebutting epistemic defeater might be a strong bit of counter-evidence, an undercutting defeater would be one that takes a piece of evidence out of the equation. You might show that an argument is invalid, or that an experiment is flawed, or that a method used is unreliable. In doing so, you are not providing an epistemic reason that clashes with the original evidence. Rather, you’re providing an epistemic reason to not be motivated by it, or to not hold any belief on the basis of it. This is exactly the sort of defeater that arguments for extreme skepticism seek to defend.
There are clear examples of undercutting defeaters for practical reasons as well. Suppose you want to eat junk food, but you also want to not want to eat junk food. You want to no longer be motivated by your desire for junk food. Again, we’re not seeing two reasons clashing. Rather, we’re seeing one reason undercut the other. Your wanting to not be motivated by that desire only gives you reason to not be motivated by that desire, and not reason to abstain from eating junk food altogether. This can be clearly seen by supposing you had a second, different reason to eat junk food. Maybe, for example, you’ve developed an anxiety disorder over your diet, and your therapist thinks it would be beneficial for you to start eating junk food every once in a while. Your wanting to not be motivated by your desire to eat junk food gives you no reason to not be motivated by your therapists wisdom or your waning mental health. Talk of higher order normative reasons may be clunky, but it’s not outside of ordinary normative discourse.
For a case in the moral domain suppose you’ve been caught doing something wrong and, wanting to avoid punishment, you decide to lie to get out of trouble. Generally, wanting to avoid punishment is not a good reason to lie. In fact it seems like a pretty bad reason, especially when you deserve that punishment. Often times when people talk about “good” or “bad” reasons they really mean “valid” or “invalid”, and it’s easy to see that the latter meaning could be intended here. But in this case the reason seems both morally bad as well as invalid. In fact, we might think it’s invalid precisely because it’s morally bad. That is, the moral badness of lying to get out of trouble gives you reason to not lie to get out of trouble. You thus have a moral reason that undercuts your prudential reason to lie.
There are many other reasons to lie that might not be all that bad: to avoid hurting someone’s feelings, to make someone believe in a benign fantasy, to preserve a surprise that will bring them happiness, to confound a wrongdoer and protect his victims, and so on. Of course each is controversial to some extent. Many people don’t think it’s right to tell children that Santa Claus exists, while others think it’s harmless fun. And Kant was infamous for saying that you should never lie, not even to protect people's lives. This controversy can easily be seen as being over whether or not these are morally good or bad reasons to lie. Those who think that lying is categorically wrong will no doubt think that there are no reasons to lie that are not morally bad reasons. So the moral value of the motivating reason factors into a good explanation of why the motivated act is wrong.
It is worth noting that when looking for an example of something that is categorically wrong, people often describe actions in such a way that encapsulates the motivating reason. To use an extreme example, it’s always wrong to kill people for the fun of it. But, uncontroversially, it’s not always wrong to kill people out of self defense. Again the difference here can be seen as a difference in the sort of defeaters potentially at play. A person's life is intrinsically valuable, and so there is always moral reason to not kill people. But, wanting to defend yourself from an attacker doesn’t seem to be a morally bad reason to kill. And so this prudential reason remains undefeated, and clashes with and defeats your moral reason to not kill. In this case, then, we do not have decisive reason to abstain from killing. On the other hand, mere entertainment is certainly a morally bad reason to kill someone. Nothing could be more obvious. And so there is moral reason that undercuts this prudential reason to kill, leaving only the original moral reason to abstain from killing. In such a case, then, we do have decisive reason to abstain from killing.
But what about higher order prudential reasons? Couldn’t someone always have a higher order desire, that would defeat higher order moral reasons? What if, for example, someone wanted to do evil for its own sake? We should keep in mind how extraordinary this is. The vast majority of people are at worst apathetic to morality. It would take a real monster to be motivated by evil itself. Because this case is so unusual, I don’t think it would be biting much of a bullet to admit of a loophole here. That being said an answer can be given anyway. Surely wanting to do evil for its own sake is itself evil. That’s a pretty bad desire to have, and a bad reason to be motivated by. Even in these sorts of unusual cases, surely there would always still be further higher order moral reasons to not be motivated in this way.
We are now seeing what I had promised: a sort of composite normative state of affairs that amounts to an indefeasibly decisive moral reason to act. For an action to be morally wrong, or morally obligatory that we not do it, is for it to be both morally bad (so that there is defeasible, indecisive reason to not do it) and motivated by a morally bad reason (so that the reason to do it has been undercut).
The property of wrongness, then, is actually a property of action and motive pairings. It’s wrong to do something for some reason. There are two questions that naturally arise. Aren’t some actions categorically wrong, regardless of what motivates them? It does seem so, and in such cases any motivating reason would be a morally bad reason. So this is something my view can easily accommodate. But what if an action has no motivating reason at all? Well, if an action is wholly unmotivated then it’s like a hiccup or a sneeze. It’s either random, or caused by some involuntary process within the person. Such actions are hardly the sort of thing that we can be held responsible for, anyway. A prerequisite for doing wrong is being in control. If someone is acting truly without motive or reason, then it’s hard to see in what sense they are in control of that action. If they are able to prevent it, like we can hold back a sneeze, then their failing to do so may be wrong. But involuntary actions themselves, however morally bad, are never morally wrong.
On a second glance this theory is not all that novel. Condition (b) could be rephrased to say that the action is unjustified. For an action to be justified, it seems, is for it to have been motivated by some reason that is morally good or, at least, not bad. Indeed, this seems to just be what people ordinarily mean by “justified” in a moral context. So really my theory can be stated as: for an action to be morally wrong is for it to be morally bad and morally unjustified. This again seems rather analytic. I expect this just is the content of our ordinary concept of moral wrongness. I am not under any illusion, then, that my theory is especially insightful from the top looking down. What is interesting is the normative consequences I have drawn out from an action being both morally bad and morally unjustified.
One might be satisfied with this, but now turn their attention to moral value. How might the atheist or naturalist then explain moral goodness and badness? This is where Wielenberg’s view shines. Moral goodness and badness may be irreducible, objective properties, that give us normative reasons to act in certain ways. If this is correct then the metaphysical explanation stops here. This, then, is a sketch of how atheists or naturalists might hold to robust moral realism.
[1] http://www.metaethical.com/2018/02/answering-why-should-i-be-moral.html
[2] http://www.mandm.org.nz/2017/03/erik-wielenberg-and-the-autonomy-thesis-part-two-standard-objections-to-the-autonomy-thesis-reasons-to-be-moral-without-god.html
[3] William Lane Craig “Q&A 230 Is Life Absurd without God?” available at http://www.reasonablefaith.org/is-life-absurd-without-god
[4] C. Stephen Layman “God and the moral order”, Faith and Philosophy, 19: 3 (2002) 304-16.
[5] William Lane Craig “This Most Gruesome of Guests” 183