Singer's Unstable Metaethics
Here, I make problems for metaethical views that Peter Singer used to somewhat endorse.*
[* From “Singer’s Unstable Metaethics” in Peter Singer Under Fire, ed. Jeffrey Schaler (Open Court, 2009), pp. 359-79.]
I. Singer’s Views
Ethics
Peter Singer is the most famous living ethicist. He’s known for defending preference utilitarianism, the view that the right action is always the action that maximizes the quantity of preference-satisfaction in the universe.
Here are some controversial views he has taken in applied ethics:
Failure to donate to famine relief is morally comparable to murder.
Animals merit equal moral consideration to human beings.
Bestiality is permissible (provided it does not harm the animal).
Infanticide of severely disabled babies is permissible, because newborn babies have no right to life.
Regarding #1: Singer says that if you saw a child drowning in a pond, you would obviously be morally obligated to save that child, even if doing so would inconvenience you and ruin your clothes.
He thinks this is relevantly like saving poor people in the developing world by donating to charity. Most people fail to donate just because the starving children are far away from us and we don’t actually see them. But that’s just a bias that we should overcome; geographical distance and visibility are obviously morally irrelevant.
Metaethics
Singer said little about metaethics, but he did make a few remarks about it, which, I argue, were in tension with his ethical views. The early Singer tentatively supported these ideas:
The Humean Theory of Reasons: This is the view that the only reason anyone can have for doing anything is that it will satisfy some desire that the agent has (or increase the chances of satisfying some desire, according to the agent’s beliefs).
Thus, you have a reason to behave morally only if you want to be moral, or being moral will contribute to getting something else that you want.
Ethical Non-Cognitivism: This is the view that moral statements, such as “x is wrong” and “y is good”, do not express propositions and thus cannot be true or false. Instead, they express a speaker’s emotions or desires, or function as disguised imperatives, or something like that. Thus, “stealing is wrong” is similar to “Boo on stealing!” or “Don’t steal!”
II. Problems
Understanding Revisionary Ethical Claims
Given non-cognitivism, what is going on when Singer makes ethical claims (1)-(4) above? He must be expressing his subjective feelings and desires. Thus:
Failure to donate to famine relief is morally comparable to murder.
Here, he would be expressing disapproval toward failure to donate to charity that is equally strong as his disapproval of murder.
I very much doubt that he feels that way, though. When you tell a utilitarian that you didn’t donate to charity last year, or you didn’t donate as much as you should have, they do not react in the same way as (I imagine) they would if you announced that you committed several murders last year.
Animals deserve equal moral consideration to human beings.
Here, he’d be expressing his own equally strong desire for animals to flourish as for humans to do so.
Again, I’d be surprised if he really felt that way. I have no trouble believing that he believes that the desires of animals matter as much as human desires. But it would be very odd if his feelings and desires about his own species were equal to his feelings and desires about any other species. Hardly anyone is like that.
Bestiality is permissible (provided it does not harm the animal).
Here, he’d be expressing toleration for, or a lack of desire for people to avoid, bestiality.
Perhaps that is in fact his attitude. Many other people would feel differently, though, so they might equally aptly say that bestiality is unacceptable.
Infanticide of severely disabled infants is permissible, because newborn babies have no right to life.
Here, he’d be expressing a lack of regard for newborn babies, or something like that.
Again, it would be very unusual if he felt that way. Most people feel sympathy and protective impulses toward newborn babies, for obvious evolutionary reasons.
Why Listen to Singer?
Anyway, even if Singer has all the attitudes that he’s supposed to be expressing, why would the rest of us take any interest in that? So Peter Singer has a really unusual set of feelings and desires. The rest of us don’t. Why would we start talking like Singer and expressing these weird preferences that he has?
Moreover, on the Humean Theory of Reasons, we only have reasons for action provided by our own desires. And there’s no such thing as correct or incorrect desires. So there’s no reason why we should change our behavior as long as it fits with our desires.
My point: Non-cognitivism does a terrible job of explaining what’s going on in Singer’s ethical discourse. But there’s nothing wrong with Singer’s first-order ethical discourse. Non-cognitivism is just false. What Singer is trying to do is rationally revise our moral beliefs, to get beyond the biases produced by our feelings and desires. He’s not just expressing his own beliefs and desires.
The problem is that Singer’s ethics is radically revisionary. It’s hard to see how a non-cognitivist could have revisionary ethical views. Perhaps Singer would say that he’s just pointing to inconsistencies in our actual desires, and a way of rendering them coherent. But
a) There’s nothing wrong with having inconsistent desires. The problem with inconsistent beliefs is that they have to be false. But if there are no moral facts and no right or wrong desires, then there’s nothing wrong with inconsistent desires. (Notice, e.g., that it doesn’t matter if your musical tastes are “inconsistent”.)
b) Anyway, there is no inconsistency in conventional ethical judgments. There’s nothing inconsistent, for example, in holding that people whom you don’t see who live far away from you don’t matter. On a non-cognitivist view, in saying that they “don’t matter”, all you’re doing is expressing your actual attitudes, attitudes which Peter Singer knows very well that almost everyone has. If you were to say that those people “matter equally” to people whom you can see in front of you, you would just be being insincere.
Why Be Moral?
According to Singer, we’re all obligated to donate a huge portion of our income to charity. On a utilitarian view, you actually have to keep giving until you have the minimum amount of money needed to sustain your life and your income-earning capacity.
Why would anyone follow Singer’s moral recommendations?
Again, on the Humean Theory, the only reason for doing anything is that it will satisfy your own desires. The rational thing to do is to maximize your preference-satisfaction.
But it seems clear donating a huge portion of your earnings to charity will not maximize your preference-satisfaction. So it seems that we have no reason to do what Singer says we’re obligated to do. (Or if you prefer, we have a reason that is obviously outweighed, so that it would be irrational to behave morally.)
By the way, one of the key arguments usually cited by non-cognitivists is that non-cognitivism explains the alleged fact that moral judgments inherently provide reasons for action. So it wouldn’t be good for Singer to just agree that we have no reason to be moral.
Singer has discussed why we should be ethical. He says that committing to an ethical life can give one a sense of meaning, which contributes to one’s happiness in a robust way. There is no danger that it will come to seem trivial or pointless, as can happen with many other life projects (say, trying to amass personal wealth).
But,
a) Given how much you have to sacrifice to lead an ethical (utilitarian) life, it seems obviously not worth it.
b) There are many other things you could try if you want a sense of fulfilment that would be much cheaper—e.g., join a religion, devote yourself to raising a family, start a philosophy blog, etc.
c) Living an ethical life by utilitarian lights is so difficult that probably no one has ever done it. So if you decide to adopt the ethical perspective, most likely you’ll just wind up condemning yourself for immorality, which won’t lead to much happiness for you.
d) Why couldn’t ethics come to seem trivial or pointless? On the non-cognitivist view, there are no ethical facts, only subjective preferences. There is no meaning or value in the universe. So if you decide to “live ethically”, that’ll just be satisfying your personal preference, which is no more right than any other personal preference.
Singer often says that one person’s interests matter no more or less, from the standpoint of the universe (or: from an objective point of view), than any other person’s interests. But that’s because, per anti-realism, from the standpoint of the universe (or: from an objective point of view), no one’s interests matter at all, and neither does anything else.
Ethical Methodology
In normative ethics, Singer says that we should not rely on ethical intuitions about concrete cases, because these are easily biased, i.e., influenced by morally irrelevant factors. Instead, he thinks we should rely on self-evident moral axioms, such as the principle that each person’s interests matter equally from the standpoint of the universe.
This fits with his revisionary ethical views mentioned above. But it does not fit with being a non-cognitivist. If you’re a non-cognitivist, what could possibly be meant by saying that a feature of a situation is or is not “morally relevant”? Since there are no moral facts, only subjective preferences and attitudes, it seems that the “morally relevant” features of a situation would just be the features that affect our subjective preferences. So it’s incoherent to claim that our attitudes are influenced by morally irrelevant features.
III. Try Intuitionism
Ethical intuitionism makes much better sense of what Singer is doing in normative ethics. If there are objective ethical truths, our emotions and subjective preferences may be (indeed, almost certainly are) often out of step with the objective ethical truths. Thus, it makes perfect sense to make ethical assertions that fail to fit with your and other people’s emotions and preferences. It also makes sense to talk about intuitions being biased by those emotions and preferences.
Lastly, these ethical facts might provide us with reasons for action that are independent of our desires. Hence, there could be a good reason for donating to famine relief even when doing so obviously doesn’t increase your own overall desire-satisfaction.
Perhaps this is why in later work, Singer has come around to intuitionism, following Sidgwick (the great utilitarian intuitionist).
If you're going to go after Peter Singer's former metaethics, you might as well be clear about what those were and directly engage with the specific brand of non-cognitivism he accepted since the different versions of it have different problems. He accepted Hare's universal prescriptivism—something you kind of acknowledge when you say, "This is the view that moral statements function... as disguised imperatives, or something like that" and that "stealing is wrong" is similar to “Don’t steal!”
When you hone in on universal prescriptivism as Singer's metaethics, it's not surprising that he thought consistency mattered. For Hare, and previously Singer, moral claims are prescriptions which imply principles with consistency requirements. So, "don't steal" isn't just an isolated claim with no other implications. Supposedly "don't steal" implies a moral principle which those who say "don't steal" commit themselves to, and must follow to all its logical conclusions.
It's also not surprising that Singer's ethics were revisionary. "Famine, Affluence and Morality" and "Animal Liberation" both show how universal prescription can be revisionary. In "Famine, Affluence and Morality," Singer argues that we all accept that it is wrong to let a child drown in a shallow pond. He says there is a principle linked with this--"if you can prevent something bad from happening...." etc--and that this principle commits us to saving far away lives too.
In "Animal Liberation," Singer says most of us are against racism and sexism, and that the principle underlying this--supposedly the equal consideration of interests principle--commits us to anti-speciesism as well. So, universal prescriptivism can be revisionary by claiming that the seemingly non-controversial moral prescriptions we all make imply our commitment to principles which have radical implications when we follow them to their logical ends.
If we believe in universal prescriptivism, one way out of this is through what Hare called "fanaticism." Someone can say "actually I don't prescribe that we must save children from drowning in shallow ponds, so it is consistent for me not to donate to charities" or "actually I don't prescribe against racism, so it is consistent for me to be speciesist." When he was a universal prescriptivist, Singer would have nothing to say to fanatics like this, so long as they were being sincere.
You make a good point about morally relevant factors. For non-cognitivists, there are no factors which are objectively morally relevant. But I suppose Singer as universal prescriptivist would have tried to show that other prescriptions we make and thus principles we accept imply that distance is not morally relevant, or something like this, so it might still have been a coherent concept for him back them.
I don't know much about Singer's former Humeanism, so I don't know exactly how that fits in here. Maybe he thought our emotions guide which prescriptions we make and thus the principles we commit ourselves to. So, for someone to prescribe that it's wrong to be racist, and therefore commit themselves to anti-speciesism, they would first need to have an emotional reaction against racism. If they didn't have that emotional reaction, and thus didn't prescribe against racism, Singer would not have been able to commit them to anti-speciesism.
However, Humean Singer might only have needed emotions as a foothold into an initial prescription, which would then commit the prescriber into a principle which required them to behave in certain ways even when they didn't have the relevant emotions. In other words, universal prescriptivist Singer might not need the anti-racist to have an anti-speciesist feeling as well for them to be committed to anti-speciesism. The emotional feeling against racism commits someone to the equal consideration of interests principle, supposedly, which then logically commits them to anti-speciesism even if they don't have any negative feelings about animals being harmed. Then again, Singer as Humean should have recognized that even if you manage to logically commit an anti-racist to anti-speciesism, they aren't actually going to act on that commitment if they don't have a feeling motivating them to avoid harming animals. And, again, I'm just speculating because I'm not very familiar with what Humeanism meant to Singer.
Universal prescriptivism is false, so Singer was right to abandon it. However, it's interesting that he hasn't formally disavowed his previous arguments in Famine, Affluence and Morality and Animal Liberation which clearly rest on universal prescriptivist foundations. If it is wrong to exploit animals and wrong to let faraway children suffer, this is not because of our prescriptions which imply principles which we commit ourselves to and must follow to their logical conclusions. Singer needs to make new arguments now.
https://fantasticanachronism.com/2022/10/10/against-effective-altruism/ , I think lots of the EA types etc. want to commit themselves to something like moral realism, because they like the kind of force moral arguments have, the weird irreducible normativity or stance independence or action guidingness etc. but at the same time they feel uncomfortable with how spooky and magical these properties seem, so they invent some sort of weird complicated meta ethical theory that lets them have their cake and eat it too, as opposed to becoming intuitionists or error theorists, who both seem to take moral utterances at something close to face value.