Does Moral Realism Require Objective Moral Reasons For Action?

An alternate approach to the relationship between rationality and ethics

Pictured: Warren Quinn, because I can’t find a single photo of him online

All moral realists believe that objective moral facts exist. They disagree about many different features these moral facts might have: Whether they’re natural or non-natural, whether they motivate intrinsically or only in conjunction with some other psychological fact, whether they deal only with actions or also with outcomes, and so on. But everyone agrees that they exist, or else they aren’t a moral realist. In fact, plenty of people would say that moral realism just is the belief that objective moral facts exist, along with maybe some other minor assumptions about how they relate to our own moral discourse. If you believe moral facts exist, but that they are totally unknowable and have nothing to do with the statements we make about what things are right and wrong in our daily lives, then you probably aren’t a moral realist. But I don’t think anyone believes that.

As far as I can tell, pretty much all moral realists also believe that we have objective or stance-independent moral reasons for action; that is, we have reasons to do certain things or avoid certain things, and those reasons (at least sometimes) relate specifically to the moral qualities of the actions in question and not just to our own personal desires or interests. Here is an example: If you see someone drowning in a lake, a moral realist thinks you should save them, and that the reason you should save them has nothing to do with whether or not you’d like to. Of course, the moral realist hopes you do want to save them. But that desire to save them isn’t the thing that makes it true that you should. (Also, many moral realists would say you desire to save them in the first place at least in part because you recognize that you have an objective reason to do so.)

I’m a moral realist, so I believe the first part: I think objective moral facts indisputably exist. I also believe the second part, in that I think moral facts generate stance-independent reasons for action. But recently, I’ve been considering a unique form of moral realism that would get rid of this last bit and accept that there are only stance-dependent, or “subjective,” reasons for action. I don’t know of any philosopher who explicitly holds this position, but there are definitely hints of it in the work of Judith Jarvis Thomson, Phillipa Foot, and especially Warren Quinn, who is tragically underappreciated. I’m not ready to adopt it myself - I haven’t fully fleshed it out, and I’m not sure it would all hang together if I did - but I think it’s very interesting to explore. So here, I’ll attempt to explain how it might work.

It might help to start with the paradigmatic case. Imagine a man is walking down the street when he encounters a woman in distress. The woman lets him know that she’s having some sort of medical emergency and needs five dollars to get a bus ticket to the hospital. The man has a five dollar bill in his pocket and has no particular urgent need for it, but he still refuses to hand it over and rather continues walking, leaving the woman in distress. On any plausible moral theory, the man here has acted wrongly. And on any particular moral realist theory, the wrongness of his action is explained by some set of moral facts. What exactly those moral facts are depends on the particulars of the theory and what sorts of moral facts it admits in general, but candidates include: The fact that helping her would maximize the welfare of everyone involved; the fact that a good person would help her; or the fact that we are obligated to give what we can to those in need.

Traditionally, moral realists would take the further position that these facts objectively count in favor of his helping her, which is to say that the man has a powerful reason to give her five dollars and that he would be irrational not to. The fact that he feels no urge to do so doesn’t matter, because the facts of the situation itself are what give him a reason to fulfill her request rather than any fact about his own dispositions or desires. When he doesn’t, he is acting wrongly. He shows himself to be callous and uncaring, and for him to be callous and uncaring just is for him to be unaware of, hostile towards, or unresponsive to the objective moral reasons generated by the woman’s distress. The man’s act itself is made wrong by whatever moral facts ground its wrongness (because it didn’t maximize the welfare of everyone involved, or because it wasn’t what a good person would do, or because it failed to meet some obligation, or whatever). But when we move from considering the act to considering the agent who acted, it would be more reasonable to say he acted wrongly as an agent because he failed to respond appropriately to the objective moral reasons for action that he had.

This last part is important for what comes next, so let me emphasize the point. In the most common realist accounts, it seems all moral defects can be ultimately reduced to persistent irrationality with regards to some set of objective moral reasons for action. A dishonest person, for example, is just someone who is chronically unresponsive to reasons generated by facts about the untruthfulness of what they intend to say, while a bigoted person is just someone who is chronically unresponsive to reasons generated by facts about the equality and dignity of all persons. This is also true of moral virtues; to be compassionate, for example, is to be appropriately sensitive to reasons generated by facts about the suffering of other people. I’m not aware of many moral philosophers who would explicitly frame moral qualities in this way, but I think it should be unobjectionable to those who believe that moral facts generate objective reasons for action and that morality consists in responding to them appropriately.

The view I’m looking to explore here would handle things differently. It would accept that the man acted wrongly, and that what makes his action wrong is some moral fact about the situation involved. Without this first step, I don’t think it could qualify as a realist view at all. But this view would diverge from the traditional account above by adopting a broadly Humean conception of rationality and therefore rejecting that the man had any stance-independent reason for giving the woman five dollars; that is, if the man truly had no desires, interests, or values that would be served in any way by helping her, then this view says he had no reason to do so, and in fact had a strong reason to refuse her request and continue walking. These two judgments - that he acted wrongly by not helping the woman, and that he had no reason to help the woman - seem to contradict each other. But they can (perhaps) be reconciled by adopting a different conception of his specific moral defect.

A quick sketch of that different conception would be this: Rather than seeing the man as callous and uncaring because he was unaware of, hostile towards, or unresponsive to the objective moral reasons generated by the woman’s distress, we can say he was callous and uncaring because the woman’s distress failed to generate any reason for action in him in the first place. In other words, rather than saying his decision to refuse the woman’s request was a sort of irrationality, we would say the very fact that his refusal was rational for him marks him off as an immoral agent. What makes a person good in this view, then, is just having the sort of motivational set that is properly sensitive to the morally relevant features of the world; what makes an action good is whether it was the result of rational deliberation with regards to the reasons such a set generates. Similarly, an immoral or wicked person is just someone with a motivational set that makes immoral or wicked action rational, and an immoral or wicked action is just one that is made rational by a morally deficient motivational set. This view would say about one person, “He was so virtuous, he had a reason to help everyone he met,” and about another, “she was so depraved, even the greatest suffering of others gave her no reason to intervene.”

Having asked around, it seems to me that some people think this view doesn’t qualify as realist because it doesn’t allow us to say that the man should help the woman, only that he acts wrongly when he doesn’t and that, in doing so, he displays morally deficient traits. But I think there are good reasons to accept that not all “shoulds” are the result of reasons to begin with, or at least I don’t think the idea that all “shoulds” are the result of reasons is self-evidently true. Judith Jarvis Thomson, for example, has a very plausible conception of normative obligation that centers the notion of defect; to say someone or something should do something is just to say that it’s what they would do were they not defective in some way. You could also simply assert that some moral obligations are brute and fundamental and not derived from any prior rational basis. I’ll return to this idea at the end.

There are a few main reasons this view appeals to me, or at least that I hope it would appeal to me even after I did the work to really lay it out in detail. The first reason - a very obvious one - is that it avoids the controversial assertion that stance-independent reasons for action exist. I personally don’t find objective reasons like these to be particularly concerning; in fact, I tend to believe the distinction between so-called hypothetical and categorical reasons for action to be unhelpful and confused, as I’ve discussed previously here. But there are many people out there more skittish than me, and it would be nice to have a form of moral realism available to them.

Another virtue of this view is that it avoids reducing all immoral behavior to some species of ignorance or apathy. On the traditional account, people who act wrongly have true reasons to do otherwise that they fail to incorporate into their reasoning process, either because they don’t recognize that they have such reasons or because those reasons fail to move them. But when I consider paradigmatic examples of moral failure, it doesn’t seem clear to me that this is what’s happening. I don’t think Hitler, for example, failed to deliberate proficiently enough regarding his reasons for not engineering the Holocaust; if those reasons really were there in some real sense, they were obviously completely and totally inaccessible to him. Maybe that can still be explained as a sort of permanent, culpable irrationality. But it just seems easier and more immediately, viscerally correct to say Hitler engineered the Holocaust because he was morally depraved, period, and not because his moral depravity caused him to be irrational in some way. The rationality of it all seems to be beside the point.

The final feature of this view that I find most appealing is also what most people will find objectionable: That it upends the traditional relationship in which moral behavior is seen as a subset of rational behavior more broadly. On the traditional view, we condemn immoral behavior by finding some sense in which it demonstrates irrationality. But why think this is the right way to frame things? This is a point Warren Quinn makes very forcefully in his essay “Putting Rationality in its Place.” Here is a short excerpt from the conclusion:

In much of contemporary moral thought, rationality seems to be regarded as the basic virtue of action or motivation, one that grounds all the other virtues. This, I have been arguing, is a mistake. … A virtue isn’t a virtue because it’s rational to have it. A good action isn’t good because it’s rational to do. On my view, the only proper ground for claiming that a quality is rational to have or an action rational to do is that the quality or action is, on the whole, good. It is human good and bad that stand at the center of practical thought and not any independent ideas of rationality or reasons for action.

Of course, Quinn’s is a different view than the one I’m suggesting here. He holds, like the traditional moral realist, that all moral behavior is rational; he just thinks the logical priority is inverted, such that to be rational just is to be good (as opposed to thinking that to be good just is to be fully rational). But it seems that we could, if we wanted, introduce an even more radical discontinuity that allowed morality to sit in judgment of rationality itself. Then we could say things like, “Yes, that act was rational, but it was deeply immoral” or, “It’s a shame that, for him, the right thing to do was irrational.” These statements might seem a little odd to some of us, but I find them very refreshing as a sort of uncompromising, morality-first approach to practical action. I would even go so far as to say that this sort of view feels the most “moral realist-y,” in the sense that it makes objective moral facts about our reasons the fundamental foundation of our judgments.

As I said previously, I haven’t taken the time to really think through every implication of the view I’m laying out. It could very well have some serious issues that make it implausible or incoherent. I’m not sure I find it more appealing than the traditional moral realist analysis, even in the absence of those issues. But I do think it’s a view worth considering, and I think the common assumption that objective reasons for action are necessary for moral realism is a mistake. If we want to take morality seriously, we should at least reconsider the default assumption that moral action is a subset of rational action. It could be that things are the other way around, as Quinn argues, or even that the two can come fully apart. I don’t think that view undercuts the force of moral realism; if anything, it puts morality in a preeminent position that other theories can inadvertently degrade.