Why People Mistake Feelings for Moral Insight | Alex O'Connor

It's a very difficult thing to define, famously, like concepts like good and bad, but we employ them all the time. Like, what is the difference between a murder just occurred and it's wrong that that murder just occurred? If you try to isolate what the actual difference is there, I think it has to just be an attitude, it's an expression, it's the way you feel. And so, the only thing that can be added by saying that something is wrong is the expression of that attitude. My name's Alex O'Connor. I'm the host of the Within Reason podcast. I think I'm best described as a philosophy YouTuber and former edgy atheist. How emotivism shapes ethics. Emotivism is an ethical theory. It's also a theory of language, really, but it belongs in this discussion around what is good and what is bad. Broadly speaking, people might be familiar that when it comes to morals, what's good and what's bad, there are kind of two schools. You have the objectivists who think that there are things which are good and bad, whether you like it or not. And there are the subjectivists who say, it's kind of just all your opinion. It's your preference. Some people like murder, some people don't in the way that some people like cake and some people don't. Emotivism is more on this subjective side, but emotivism is something a little bit different. It suggests that when people make ethical statements, what they're doing is expressing an emotion. When you say that murder is wrong, you are literally expressing an emotional state towards murder. That sounds very similar to subjectivism. In the subjectivist framework, the phrase murder is wrong translates to something like I don't like murder. A phrase like I don't like murder is a claim about my psychology. I could be lying. I could say I don't like murder, but actually I do secretly like murder. It could be something that I'm lying about. It could be true. It could be false that I don't like murder. It seems to be a claim about my brain, about my psychology, about my attitude towards murder. Emotivism is not that. It's not the reporting of an emotional state or an attitude. It is literally the expression of that attitude. It's the difference between telling you I don't like murder and literally going like ugh, murder, or just like pulling a nasty face, or as A.J. Ayer famously had it, going boo, boo murder. The emotivist thinks that that is what's going on when we say murder is wrong, meaning that these ethical claims are literally not the kind of thing that can be true or false. There are some kinds of expressions and statements that have what's called truth value. Truth value means that it can be true or false. A statement like the sky is red has truth value. It's just that the truth value is false, whereas some statements like commands, for example, go over there doesn't have truth value. It's not the kind of thing that can be true or false. A phrase like I don't like murder could be true. A phrase like the objectivist theory that murder is wrong might be true, but for the emotivist it's literally like going boo, murder, which isn't true. It isn't false. It's just an expression. It's the same as me kicking my foot and going ow. Is that true? Is it false? Am I lying? That kind of doesn't apply here. You're just expressing something. That's what the emotivist believes anyway. That's the broad position of emotivism. It belongs in this category of non-cognitivism about ethics, which means that these statements don't have truth value. There are cognitivists who think that ethical statements have truth value and non-cognitivists who think that they don't. Emotivism is a form of non-cognitivism. For example, another non-cognitivist school is prescriptivism, which thinks that ethical statements are essentially commands. When I say murder is wrong, what I mean is something like don't murder. That's the kind of statement that I'm saying. It's not exactly expressing an emotion. It's giving a command, but it's also non-cognitivist because it doesn't have truth value. You're just saying something that isn't true or false. Reasoning can be connected to ethical statements on an emotivist framework. There is this idea that if we're emotivists, then suddenly the possibility for all disagreement on moral issues vanishes because how can you have a debate if there's no truth value? And that's true to a degree and I understand the criticism. It's not so much a criticism as an observation. It might just be true that we can't do that, but a lot of people aren't satisfied with that. But it's way more restricted than people realise. AJ Ayer pointed out in Language, Truth and Logic, he sort of is the founder of this emotivist school. He points out that the vast majority of moral debate is not moral debate at all. It's debate about descriptive facts that we then apply our sort of moral intuitions to. So for example, if you imagine a debate about guns in America and people are having a back and forth and one person says, did you know that if we criminalise guns, it would reduce the number of gun related deaths by 10,000 a year? And another person says, ah, yeah, but did you know that you haven't taken into account this statistic that if you get rid of good people with guns, then less people will be saved? And then someone says, oh, but did you know that swimming pools kill more children than guns do? This is the kind of debate that we're having back and forth, but none of those claims are moral claims. They're just descriptive, factual claims. They're statistics. You can go and test them. It's either true or false. A lot of what we think is moral debate is actually just debate about descriptive facts, which once those are settled, we then apply our moral intuitions to. So for example, in the gun debate, one person has one moral view and another person has another moral view and they're battling it out. But actually they might have the same moral view, so to speak. For the emotivist, it's some kind of emotional expression. And so what might be the emotional expression? Well, suppose one person says, if we allow guns in society, innocent people are going to die. And another person says, but like, if we don't allow guns in society, then only the bad guys are going to have guns and they're going to kill innocent people and there's going to be no one to protect them, so innocent people will die. Both of them agree. Innocent people dying. Boo. And what they're debating is a descriptive truth of the matter as to whether, literally speaking, more guns, less guns, more deaths, less deaths. And so the idea that instantly comes to mind that we can't have any kind of reasonable discussion about moral issues, it's true, but I just want to point out that it's way more restricted. It's not like you can't have this kind of gun discussion. You can't have a discussion about, I don't know, about abortion and healthcare and you can't talk about, I don't know, pick another pet moral issue, speed limits, you know, like how can we have a moral discussion about speed limits? Well, we can talk about how many people will die, the statistics about increasing by 10 miles an hour and all this kind of, that's all just facts. And then we'll apply our moral intuitions. And I think that when we actually break down and isolate the moral element alone, we'll find that there is actually a lot more agreement than there is disagreement. That doesn't do away with the problem. The challenge is what if there really is disagreement on the moral stuff? What if two people genuinely just have conflicting boos and yays? And in that case, there are sort of two strategies. You either say, well, if you really broke those down further, you'd realize that everybody ultimately has the same emotional expression on fundamental ethical issues because we're all human. We all have a shared evolutionary heritage that will have placed within us certain emotional predispositions that ultimately will be the same if we just break them down far enough. The other strategy is to say, no, people do just have fundamental value conflicts. And that's just true. And that's just the way it is. And there's nothing we can do about it. Whichever of those you find plausible is up to you. But obviously, the former is a bit more of an optimistic idea. I think people will be more attracted to that because it allows us to sort of say, well, everybody agrees that boo innocent people dying. And so, you know, we can just act in accordance with that. But it seems more likely the case that that's not true. I mean, on that point, it's quite clear that there are people who do not have an adverse emotional reaction to innocent people dying. And so that can't be the most foundational ethical principle that we all agree on. But it might go deeper than that. You know, it might be that's not foundational. Innocent people dying cannot be a foundational ethical expression because it contains too many concepts. The concept of innocent, the concept of death, you know, the reason why you might think the reason why you might have that emotional reaction to innocent people dying. It can be broken down much, much further. If somebody thinks that their fundamental ethical principle is innocent people dying, boo, I can think of 10 situations off the top of my head in which you would be perfectly fine with an innocent person dying. It can't be that. It must go deeper than that. And it might genuinely be that if you actually go deep enough into people's moral emotions, they will ultimately, at root, be identical across the human species. And that at a higher level, when someone has an emotional reaction to something, that emotional reaction, you can't say it's wrong because there's no truth value, but you could say that it betrays a mistake that somebody has made. This is really interesting. Like, suppose you were really upset because it was raining on your wedding day, but it turned out it wasn't actually raining. You'd been accidentally watching a tape recording of the news forecast, right? And you believed it was raining and you're really upset about it. And I come over to you and I say, hey, you know, it's not actually raining, right? And suddenly you're happy again. Those emotional states, the sadness of the rain and the happiness that it's not raining, they're just emotional states, right? They're not true. They're not false. They're literally just expressions of emotion. But I can still come in and give you information that will change your emotional state and literally convince you to sort of feel the opposite way just based on the information that I've just given you. So in the same way, it might be that people have, so to speak, mistakes in their emotional expressions at a higher level. In the same way that you might be sad that it's raining, even though it's not really raining, somebody might have an emotional reaction to a surface level moral issue. They might be like, oh, I think it's wrong that that gun law passed. But if you actually just explained the facts to them, they might be like, actually, yeah, fair enough. No, I think it's good that that gun law passed. So you're sort of having this what looks like rational discussion about moral issues, but all you're actually doing is discussing the facts. And it might be the case that if we actually made every human being on earth consistent internally, emotionally, with all of their foundational emotional motivations, that there would be a lot more unity than disunity. But we don't know because, of course, that experiment has not and probably cannot ever be done. I'm not sure there is a distinction between emotions and attitudes. I think these are essentially synonyms. Simon Blackburn disagrees with me about this. He's an important metaethicist. He founded a metaethical position which he calls quasi-realism, which is a kind of non-cognitivism about ethics. But he thinks that, for example, you can have an attitude towards something without having an emotional state or being in an emotional state with relation to that thing. So you might have an attitude that, you know, that thing is bad or that you shouldn't do that or something like that. But it doesn't mean that you're like emotionally invested. I'm just not sure that that's possible. I think that to have an attitude towards something is kind of, literally speaking, to have a way that you prefer it to be. And I think that that belongs in the category of emotion. But this is essentially semantics. Like if you want to draw out a distinction, you could probably do it and say that, you know, an attitude towards something is something like your desire for the way that you want it to be, whereas your emotional state is the reaction you have to it not being the way you want it to be. Yeah, you could draw out some distinction like that. But I think that in practice, we're kind of talking about the same thing. One thing that's important to point out is a mistake that people make. Emotivists don't think that wrongness or badness or something like that like maps on to a pre-existing emotion like disgust or like anger or upset. It just belongs in that category. But it's its own unique emotion in the same way that anger is a little bit like sadness, but they're not the same thing. They just belong in the same category. They're just two unique kinds of emotions. Wrong is its own kind of emotion. It's unique, but it's still an emotion, you know. It just belongs in that category. So, well, why might we think that? Well, how do people talk about ethics? Okay, we talk about things being wrong, but when someone does something absolutely awful, like really, really bad, what do we say? Disgusting. You know, you say that's just disgusting. It's despicable. And the sort of emotional reaction just comes up in your face. Now, some people will say that that emotional reaction is your reaction to the wrongness. There is this objective wrongness and you're having this reaction of disgust to the wrongness. I'm more inclined to say that the wrongness in its totality is the feeling of disgust. A great example for emotivism is the incest taboo. It's the most universal taboo that exists in human societies. Everywhere we look, there is a prohibition against any kind of marital or sexual relation with your immediate family. Some tribes, I think we've found, have some interesting variants whereby you can marry your older sister but not your younger sister, you know, stuff like that. But across the board, there is some kind of incest taboo. It's quite easy to explain evolutionarily why this is the case. Clearly, there are all kinds of problems with inbreeding that will be evolutionarily deselected. But evolutionary explanations aren't a justification. I can also evolutionarily explain why people tend to be a bit racist. I can explain that in terms of your evolutionary biology, but that wouldn't be a justification for racism. That would just be an explanation for it. So I can explain why there is this incest taboo. But most people want to say that incest is wrong. Why? It's a famously difficult thing for any, at least secular ethicists, to do, especially if you're some kind of utilitarian who thinks that the only thing that matters is the minimization of suffering. Oh well, you know, it's because you can have disabled children. A few questions jump out of that for me. Firstly, what's wrong with having disabled children? Secondly, obviously, you can control for these factors. Okay, so you make sure that it's, you know, two women, or it's two men who've been sterilized, or something like that. Or a man and a woman who've been sterilized, or two men, you know. Okay, well then it's about the power imbalances in a family. Okay, well just control for that. So often we waste so much time on these irrelevant factors. You can literally just control for these factors, you know. What's wrong with it? When it comes down to it, most people, if they search their inner experience, I mean the sort of enlightened rationalists will go, well then I don't think it is wrong, you know. And okay, maybe they're right in being consistent there. But even such people tend to still feel uncomfortable with it, you know. Somebody actually once put me through my paces on this, because I used to be a bit of a utilitarian, and I used to talk about this and say, well I suppose I can't see anything that's objectively wrong with that. And someone I know pretended that his new girlfriend, he introduced her to me as like his cousin or something, and then like later on it became clear that they were an item, and he was just toying with me to see how I'd actually react. Because it's easy to say, oh yeah, you know, there's nothing wrong with that. But you're still going to be like, what? Because it's so universal. So what's going on there? We just think it's disgusting. We're just like, gosh no, I mean that's your brother, that's your sister, that's your mother. It's like that's all you've got, ultimately, when you break it down. There will be people who say, no, you know, like I believe in God, and God has enshrined into the moral law of the universe that this is wrong. Interestingly, still doesn't tell you why it's wrong, just gives you confidence that it's wrong. But fine, shelving that at least for the time being, like for at least the secular ethicists among us, like even secular ethicists will say, no, emotivism is wrong. Okay, well then try to account for what's going on here. Maybe you will just say that in this case people are just having an emotional reaction, and for that reason it's not wrong. But I think that's just what's going on across the board. Think about the worst, most homophobic person you've ever met. What do they say about gay people? They don't just say abstractly, like completely dispassionately, I think that is immoral. They say it's disgusting. It's disgusting. No, no, no son of mine will be getting gay. No, no, you know, it's got this emotional component that everybody knows. Everybody knows about this emotional component that these things have, but I think there is just the emotional component. I think that's what this ethics is. That is the difference between a father saying my son is gay as a matter of fact, and saying it's wrong that my son is gay. I think if you isolate what the difference is between those two statements, it's that in one case he's feeling something in relation to that fact about the world, and it's not like another fact that he's adding on. There's only one fact there, which is that his son is gay, and the only thing that's being added on and saying that it's wrong that his son is gay is how he feels about it. You might remember that I said that emotivism is a theory about ethics, but really it's a theory about language. It's more about what people mean when they say particular words. And so to say that when someone says like, you know, the world ought to be this way, what they're doing is they're expressing an emotional state about the world. They're expressing their emotional reaction to the world, but they don't like it, you know. And you can still make sense of what's going on there. It's kind of like if we were looking for some food, and I said let's go to Wetherspoons. I don't know if you know what a Wetherspoons is, but it's a glorious pub chain in the United Kingdom. And you looked on your phone and said, oh, they're all shut. And I went, oh, God's sake, you know. I'm expressing a preference about how I wish the world were, you know. And crucially, I'm not just reporting this. I'm not just saying, you know, I wish we lived in a world where Wetherspoons were open all night. You know, that would just be a fact about my psychological state. It's the expression itself. And I think that if I said something like, it's just so wrong that they shut early or something like that, I use these moralizing terms, that what I'd be doing is expressing an emotion. So you can still say something like, you know, the world really ought not be that way. But the emotivist just translates that language into the language of ethical expression in the same way that, like, I don't know, you might translate when, a lot of the time in like relationships with people, you kind of have to translate what they're saying. If your wife says to you, like, oh, you're back 10 minutes later than you said you'd be. What she's actually saying is, you know, stop taking me for a fool. Like, you're lying to me and you're never on time and you're lazy. But it's sort of this, there's something more going on under the surface. It might even be a kind of expression of emotion when she says, oh, well, oh no, it's just that I thought we could maybe, you know, get dinner together tonight. On the surface, she's making a descriptive claim about her psychology. I'm just reporting to you, I thought, I had a belief that we were going to get dinner. That's just a fact about the world. But clearly what she's actually doing there is expressing an emotional state to you. She's expressing her upset, but she's sort of disguising it in the language of truth claims about her psychology, right? And the same thing is going on with ethics when someone says, I think that's wrong. What they're doing is having an emotional reaction to something and putting it into descriptive clothing. I was always already just deeply suspicious of the idea that ethics is this, it's a very difficult thing to define famously, like concepts like good and bad, but we employ them all the time. Like outside of this kind of conversation, you know, if I'm a bit late to this interview and somebody says, oh, I just think that's really bad. That's really immoral that you were late. Like without doing any meta-ethics, we all kind of get what's being said there, right? Technically speaking, if I asked everybody in this room, well, what's your definition of wrongness? What does that literally mean? Do you mean that it causes suffering? Do you mean that people might have different views and whatnot? But like on the surface level, we all sort of use these terms and we kind of get along using these terms and we know what they mean. So I was just interested in the fact that it's a kind of lexicon that I was already frequently employing, but also deeply suspicious of the idea that there is this thing in the universe called goodness. It's just like, where does it exist? How does it exist? Does it exist in people, in actions? Can something even exist within an act? Where is it? What is it? It's like, okay, well then let's start paying attention. I mean, this is not an uncommon thought for people to realize that this good and bad stuff is essentially just people saying stuff. It's just like viewpoints and subjective preferences. But let's get more specific. What exactly am I saying? Because I use these terms. It's not like I'm using a term without knowing what it means. It's not like I've just used the word obnoxious, but not known the definition of obnoxious. I know what I mean when I say these terms, right? But when I reflect on it, but what is this good, this bad? I must mean something. So what do I mean? And that's when I sort of do this exercise of really trying to isolate the moral element. What is the difference between it's raining and it's bad that it's raining? What is the difference between a murder just occurred and it's wrong that that murder just occurred? If you try to isolate what the actual difference is there, I think it has to just be an attitude. It's an expression. It's the way you feel about that particular situation. And so the only thing that can be added by saying that something is wrong is the expression of that attitude. That's sort of the way of thinking about it. That's what led me to emotivism, I think. It's hard to remember literally how I sort of first started becoming convinced by this, but I suppose I was convinced that there is no objective good or bad. I couldn't even make sense of such a concept. And once that's gone, it's not even like you're really doing ethics anymore. Now you're just trying to work out, okay, then what do those words mean? What is the referent for the word good? It can't be literally meaningless. It must have some kind of referent. That's what led A.J. Ayer to it in the first place. He had this whole philosophy about meaning in language. He thought that statements could only be meaningful if they were either something you could verify, so prove or disprove through empirical observation, or if they were just analytically true, just true by definition. Things like 2 plus 2 is 4, something like that. Analytically false. 2 plus 2 is 5. It's a meaningful statement. It's just meaningfully false. This philosophy, by the way, has completely fallen out of favour, this logical positivism as it's called. This idea that statements are only meaningful if they are verifiable or analytically true or false, which means that any kind of statement which does not belong in that category, it's not just that it's wrong or false. Ayer thought it was literally meaningless. It was the same as just going blah, blah, blah or something. It just literally had no meaning. And so the big problem in chapter 7 of Language, Truth and Logic is what about ethical statements? When I say murder is wrong, it doesn't seem like it's something I can empirically verify. I can't go out and test with scientific experiments whether murder is wrong. But it also doesn't seem analytically true. It doesn't seem true by definition. It seems like I can meaningfully ask, is murder really wrong? And so Ayer's philosophy led him to this position where he either had to say that ethical statements were therefore literally meaningless, again, saying murder is wrong would be the same as just saying murder is blah, blah, blah, like just a meaningless statement. But Ayer was like, no, it definitely means something. So if it's not analytically true and it's not empirically verifiable, then what do these ethical statements mean? Well, they're expressions of emotions. They're not the kind of things that are true and false that we can test scientifically or might be analytically true. It's not in the realm of truth value at all. It's literally just an emotional expression. So it's meaningful in that sense, in a slightly different sense that the literal meaning of the expression is just that, an expression. The biggest problem for emotivism is this so-called Frege-Gieg problem, named after two philosophers. It's also known as the embedding problem. The reason for that is because it deals with this problem that when you take the meaning of a word or a statement and you embed it in the English language into a larger sentence, the meaning needs to stay the same. So if I say, it's raining outside, and then I say something like, I wonder if it's raining outside. So I've embedded that phrase, it's raining outside, that atomic phrase inside a larger sentence. That isolated element of it's raining outside has to mean the same thing. I'm it's raining outside, or I think that it's raining outside, or if it's raining outside, then I will bring an umbrella. Those statements, I will bring an umbrella and it's raining outside, have meaning of their own, and they retain that meaning when you embed them into a statement like, if it's raining outside, then I'll bring an umbrella. The problem for Frege and Gieg is that with emotivism, if ethical statements are just expressions of emotion, that is, when I say murder is wrong, I literally mean something like boo murder. Then how can I make sense of a statement like, I wonder if murder is wrong? What does that statement mean when someone says that? Because if murder is wrong just means boo murder, that would mean someone saying, I wonder if boo murder. That doesn't make very much sense because you're expressing an emotion and then saying, I wonder. It's almost like you stub your toe and you go, I wonder if, ah, it doesn't make sense. That's not the kind of, it's the only way you could make sense of someone saying, I wonder if murder is wrong, is if in that context, in the embedded context, murder is wrong means something different. Then in that context, it means it's not an expression of emotion. Oh, it's a, you know, a discussion about their mental attitude or something like that. Meaning doesn't just change like that for sentences. And so the embedding problem says, well, if you look at a statement like, I wonder if murder is wrong, or if murder is wrong, then murdering James is wrong, which seems to be a sensible statement that I can make sense of. Then given that in that embedded context, those statements can't just be expressions of emotion, means that outside of those embedded contexts, they also can't just be expressions of emotions. It's a bit of a technical problem. I hope that kind of makes sense. It's a little bit difficult to explain actually, but I think hopefully that comes across and people can do further reading on this. But I mean, if you think about an argument like, if murder is wrong, then murdering James is wrong. Premise two, murder is wrong. Conclusion, therefore, murdering James is wrong. That conclusion follows from those premises. We don't know if the premises are true, but we can say that if the premises are true, the conclusion follows. It's a valid argument. But logical validity relies on the truth value of the premises. To say that an argument is valid is to say that if the premises are true, the conclusion is true. That's what it means to say an argument is valid. That's like the definition of a valid argument. So if we take that argument, if murder is wrong, then murdering James is wrong. Premise two, murder is wrong. It follows that murdering James is wrong. That seems valid. But that would require us being able to conceive of premise two, murder is wrong, as being something which is true or false. If we're an emotivist and we say, no, no, no, that's an expression of emotion. It can't be true or false. Then we'd have to say that that argument that I just laid out is not valid. That doesn't seem right. Seems perfectly valid. I mean, it has to be valid, surely, if you see it written down. The conclusion obviously follows from the premises. So it's as though when you take these ethical expressions and you embed them into bigger contexts, like syllogisms or I wonder if, or whatever, if-then statements, that suddenly they're just not emotional expressions anymore. Suddenly they have truth value and we talk about them differently. That doesn't seem to make any sense. So that's the Frege-Geach problem. If you're interested in emotivism, the place to start might be A. J. Ayer's Language, Truth and Logic, which is where this all kind of comes from. It's chapter seven, in which he talks about what it means to use moral terms. The book as a whole explains his logical positivism as a philosophy, which like I say, no one really believes in anymore. It was kind of popular for a while in the 20th century and then just completely fell out of favour. But that chapter seven is still the origin of this emotivist framework. But a more modern philosopher to read on this stuff would be Simon Blackburn, I think. Blackburn is famous for his introductions to philosophy. He's written a few sort of popular introductions for just getting into philosophy, but he's also done some important work in metaethics. He's not himself an emotivist, but he's written compellingly about emotivism and non-cognitivism more broadly. He's also compiled or constructed quite an influential response to the best objection to emotivism. Those would be the two names that jump out at me to begin exploring emotivism. Want to support the channel? Join the Big Think members community where you get access to videos early, ad-free.