Smith writes,

Unless we accept Zwolinski and his fellow Bleeding Hearts as the final arbiters in this matter, we will need clearly defined criteria by which we can identify legitimate exceptions to the NAP, and distinguish them from purported exceptions that cannot be justified. Again, it will not do for Zwolinski (or anyone else) to proclaim that x is an exception to the NAP unless he also explains and justifies the criteria we should use to distinguish between legitimate and illegitimate exceptions.

This seems wrong, though I might not quite grasp what Smith has in mind here.

Most of us have a significant degree of moral knowledge, even though most of us cannot articulate a theory that codifies, justifies, or explains this knowledge. And, it’s not at all clear one needs to have a theory in order to have moral knowledge. In fact, one can probably subscribe to the wrong theory and still have moral knowledge. So, for instance, MLK jr. probably thought that morality depended in some way on the existence of God, but even though that’s wrong, MLK jr. still knew that racism was wrong. Ayn Rand advocated ethical egoism, but even though ethical egoism (including her esoteric version of it) is wrong, she still knew it is wrong to kill people for fun. If pressed to articulate a full moral theory, I’d say some version of Rossian pluralism comes closest to the truth, but even if I’m wrong about that, it doesn’t mean I don’t know that I’m obligated to feed my kids.

Just as we can have knowledge about what to do in particular situations, we can also know that a principle is flawed or has exceptions, even if we cannot articulate a principle that explains all the exceptions. So, for instance, most normal people know that it is usually wrong to lie. And most people also can articulate cases where lying is permissible. Most people know that you are permitted to lie to the murderer at the door. But probably most people cannot give a good account of just when lying is permissible. Their moral knowledge is implicit and tacit.

I once refereed a paper in bioethics, a field I know little about. The paper articulated a theory of when cells count as human individuals. I noticed, contrary to the author’s intent, that the theory had a perverse implication: if I decide to scrape off some living cells from my tongue, thus killing the cells, I thereby commit murder. Now, I don’t have a theory of just when human cells become sufficiently person-like to acquire rights. If I had to articulate such a theory on the spot, experts on those questions could probably rip my pet theory to shreds. Still, it was easy for me to see that the principles advocated by the author had serious problems and couldn’t be right.

Similarly, the students in my introductory class can all identify that “The Ones Who Walk Away from Omelas” presents a strong objection to certain crude forms of utilitarianism, even if none of these students can articulate a good theory of when it is and is not okay to try to promote general happiness.

Smith’s argument seems to presuppose that Rossian pluralism is wrong. If Rossian pluralism is true, then it will turn out that competent moral agents will be able to identify exceptions to general moral rules, but there simply is no way in principle to perfectly codify all of those exceptions.

A good chunk of my dissertation was on the issue of moral knowledge without moral theory, so if there’s interest, I might press this further.

Note an area of agreement: I agree with Smith that choosing a theory is comparative. All moral theories–in fact, probably all theories in all fields–will have counterexamples and problems. It’s part of the nature of theorizing. For a good account of why, see the introductory chapter of David Schmidtz’s Elements of Justice.