One of my favorite philosophers, Michael Huemer, has a fun paper in which he presents an ontological argument for moral realism. I think that the argument, while technically successful, is misleading, and points to a relatively trivial conclusion. Here, I’ll explain why.
Huemer argues that torturing people is objectively immoral, writing:
1. If the following conditions hold –
a. If S knew that P, this would provide a reason for S to Φ,
b. If S knew that ~P, this would provide no reason for S not to Φ, and
c. S has some reason to believe that P
– then S thereby has a reason to Φ.
2. If we knew torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.
3. Even if we knew torturing babies was not objectively wrong, this would provide no reason to torture babies.
4. We have some reason to believe that torturing babies is objectively wrong
5. Therefore, we have a reason to avoid torturing babies.
Distinguish two theses: objective moral realism and subjective moral realism. Objective moral reason claims that you have certain objective stance-independent moral reasons in the sense that there really are reasons out there to take actions that diverge from your desires. In contrast, subjective moral reason claims that an agent should be responsive to moral reasons which may or may not be out there.
Here’s a way to helpfully distinguish them: imagine that someone secretly poisoned my coffee (perhaps a rabid Camus fan). I have objective reason not to drink the coffee—drinking the coffee would truly be a bad thing to do. But I have no subjective reason not to drink the coffee—I don’t know that it’s poisoned, so I wouldn’t be acting foolishly if I drink the coffee. A subject reason is a reason based on the things you know about, an objective reason is a reason that someone who knew everything would be responsive to.
Huemer’s argument does establish subjective moral realism. But subjective reason is, I claim, totally trivial—of course you have some reason not to torture babies if moral realism might be right. That reason might come from the fact that it might be objectively wrong.
I’m a subjective moral realist about deontological moral reasons, for instance. I think that, because deontology might be right (say 1 in 10^10^123 odds) if all else is equal, you should afford some weight to deontological reasons. But I’m not a realist about deontological reasons in any meaningful sense—I think almost certainly there are no such reasons, but because there might be, I should afford them some weight.
So Huemer’s argument does establish subjective moral realism. But that’s a perfectly trivial position. Heck, we should all be subjective moral realists about torturing people—you have some reason to torture people, on the grounds that it might be good, but very weak reason because it’s quite bad in expectation. In contrast, when framed as being about objective reasons, 1 is false—what you have objective reason to do has nothing to do with your state of knowledge.
Huemer’s argument, therefore, does prove a trivial conclusion. But that’s not very interesting. When people disagree about objective morality, they’re disagreeing about the presence of objective stance independent moral reasons, not subjective ones.
There are two upsides to Huemer’s paper, however. First, it shows that subjective moral realism is true. That’s important—it shows that rational anti-realists should afford some weight to stance independent reasons. Second, it shows that humeanism about reasons is false. Humeanism about reasons says all of one’s reasons come from their desires. This shows that there’s sometimes a reason to do something even if you don’t want to do it.
I think your objection to 3 doesn't work. He's not saying that even if something is not objectively wrong there can't be a reason to do it, he's saying the lack of wrongness itself doesn't provide a reason to do it. Rightness would provide a reason, sure, but that's not the property he's talking about. He's saying "not wrongness" doesn't provide a reason. Like how the fact that it's it not wrong to cross the street doesn't itself provide a reason to cross the street
I wrote a very similar argument a while ago about utilitarianism. Seems like we converged on the same point. But yeah, subjective/objective distinction.
https://www.parrhesia.co/p/probabilist-reasons-to-reject-utilitarianism