Morality Isn’t Objective

By Noah Birnbaum @ 2025-06-26T13:42 (+42)

In response to Matthew's post about the objectivity of morality, I'd thought I'd throw out a(n initially -- this became much longer than I was expecting) short post explaining why I think this view is pretty implausible. 

I have another post on my Substack that shows why I think some other arguments for moral realism fail (though I'm not sure I endorse everything I argue there anymore...). If you like this one, check it out! 

Note: The formatting of this article got a little messed up (inconsistent about ordering with the numbered and lettered bullets, etc), but it should still be understandable and fairly readable. Also, I didn't spend much time editing because I wanted to get this out so apologies in advance.  

1. Evolutionary debunking arguments: I see there being two main versions of these arguments:
   1. Even if moral realism were true, there is no causal connection from the facts about morality to our beliefs about them (because our beliefs about morality stem from cultural and biological evolution); therefore, we shouldn't think our beliefs about the facts are reliable. (See Street)
   2. We can explain all of the data without positing the existence of these facts by appealing to already known facts about evolutionary game theory -- societies who cooperate are more likely to survive, etc. Not only does it predict the types of behavior we see, it often predicts the exact way that we expect to see it (i.e. people are more likely to behave altruistically in front of others, people care more about their family, people care more about themselves than others, etc). (see Joyce) Since we can explain all the data without positing the existence of another substance, via Occam's razor (or some similar heuristic), we shouldn't posit the existence of special moral facts!
      1. In his article, Matthew says: "You could tell a similar debunking story about our belief in the law of non-contradiction. But I think in such cases, we just have to consider the plausibility of the belief and see that, even though they can tell a consistent story of how you come to mistakenly intuit some fact, their account is less plausible."
         1. Say we gave some ontology of the world and need to account for why people have a belief that there can't be contradictions. Unlike moral facts, I think this is the exact type of fact evolution would want you to know -- knowing whether there can be contradictions or not is actually really helpful for survival (unlike moral facts). The mere fact that you say that it comes from a source is insufficient for debunking - you must show that we have reason to think it comes from some unreliable source.
      2. "Like, suppose that I give the theory that everything in the world was created by a brain worm. You point out that that’s crazy—a brain worm being fundamental is very complicated, it can’t make the world. I say that the brain worm is fundamental and misleads you into thinking it’s complicated plus that complexity is a virtue plus that brain worms can’t create the world. I point out that people often are misled by brain worms. It’s true that I can tell an internally consistent story of how you come to be mistaken across the board, but the story is just not at all plausible. Same with the story in which all of our beliefs about morality are wrong—random side effects of blind evolution."
         1. I don't want to strawman you here, and I doubt you believe something like "evolution is as simple and/or plausible about brain worms." I think I see the point that you're making here more about the intuitiveness of the hypothesis being reason to believe it, which is a point I will come back to.
      3. "Or suppose that I try to debunk the existence of love. I note that it would be evolutionarily beneficial to think you’re in love because that aids in reproduction. Adding love to your ontology is an extra posit. While I could tell an internally consistent debunking story, one would need to evaluate its plausibility. And such a story wouldn’t be plausible—it would be very unintuitive, just like the debunking story of the anti-realist."
         1. I don't think that it's at all implausible that love is totally the creation of evolutionary pressures and not real. I actually think this is probably a widely held view on the forum - if you want to posit otherwise, you're gonna have to start with a much more primordial belief in epistemology rather than appealing to some much 'higher-level' intuition that people have. I also don't know what it means to "evaluate it's plausibility" here aside from evaluating how intuitive it is, which I don't think is a reliable method to attain truth.
      4. "Now, is it true that our evolutionary beliefs are the byproducts of blind chance so that it would be a huge coincidence if they were true? No, I don’t think so. Here’s my account of how we have true moral beliefs: evolution makes us super smart, and then we figure out the moral truths. This is the same way we come to have true beliefs about modal facts, logical facts, mathematical facts, and so on. There’s no special challenge for moral facts (now, I think us having such rational capacities is surprising on atheism, but to account for how we know tons of other things, we should already grant that we have those rational capacities even if we’re atheists)."
         1. I don't think we come to facts about math or logic this way -- I think cognitive science has a pretty good (approximate) story of how we come to those beliefs about math, which I'm not gonna go into here (though I do discuss this a bit more in part a of the intuitions aren't sufficient section).
         2. However, I'll bite: (there's more to say, but I'm simplifying for the sake of this post) imagine that you wake up one day, and you see that some of our beliefs about higher level math don't continue to reflect what we see in the empirical world. In fact, they contradict what we see. Would you still think that they reflect the truth? I think you shouldn't.
         3. Separately, imagine that we didn't have the intuitions we do about some higher level math, but we see empirically that they follow all these rules that we can start to get a grasp on and make derivations from. In this case, should we think that math reflects reality? I think yes.
         4. Given that we've shown both that (ii) intuition alone isn't sufficient to believe in the existence of math and (iii) empirical observation alone is sufficient for reliable beliefs, we shouldn't believe that we have knowledge of math and logic by intuition alone! Similarly, having knowledge of morals via intuition alone isn't sufficient.
2. Moral motivation: One part of the moral realism debate that always befuddles me is moral motivation (some part of this is called the internalism vs externalism debate, but I'm not interested in going into the semantics of this).
   1. Suppose we grant that there were these things in the world that were just morally true (say, unnecessary killing is bad). I still don't think that this means they should be action guiding. Why is this? Well, suppose there was a god, and he (yes, this god is a he) said that people have an obligation to do the Macarena every Tuesday. In addition, everyone knows that god believes and says that we should do this. However, this type of god also has no enforcement power (everybody also knows this) -- in other words, he's not going to do anything if you don't do what he says (he won't send you to hell, give you karma (on the EA Forum), make your life shorter, or anything else). I would argue that we have no reason to do what god says. Similarly, even if facts about morality existed, I still don't get (I do get it, but this is rhetorical) why people should be motivated to follow those facts.
   2. To pump this intuition further, imagine a psychopath who only cares about themself. Is there really something irrational about what they are doing (i.e. that he is not using his best information properly; not that we can define being moral as rationality and then say he is being irrational)? Sure, we don't prefer it, but I'm not sure we can say that whatever faculty he doesn't have/ doesn't motivate him is actually a failure of rationality of any kind. What would you say to them to make them a realist? It seems like they merely have a different psychological propensity, and there is not much that can be done to make them intrinsically motivated by the moral facts.
   3. Now imagine the psychopath but for math (as in, they don't have the intuitions we do about math) -- I think we would be able to convince them about much of the math that we do today on the basis of empirical observation and maximizing whatever goals they value. This, I believe, is different in the case of morality: in what sense are the moral facts helpful for goal maximization like the mathematical facts are (which are instrumentally important for understanding the world!) such that we'd be able to convince those without the faculties of its truth? (Note: my guess it that they would likely even accept mathematical principles that aren't clearly subject to empirical investigation--like non-euclidean geometry-- because the usefulness of following these mathematical principles creates a good meta-inductive argument that they would generalize to empirical phenomena that have not been tested yet, but this is a side point).
3. Intuitions aren't sufficient: Much of the initial argument for moral realism comes from the strong intuitions we have about a bunch of moral claims (i.e. unnecessarily torturing babies is actually bad, we can say that Hitler was doing something wrong, etc). However, unlike Matthew and others, I don't think that intuitions alone are a very good basis for epistemology.
   1. Cognitive science suggests that the reliability of our intuitions depends heavily on the presence of consistent empirical feedback. In domains where such feedback is weak or absent, our intuitions are less trustworthy, and we should be more skeptical of relying on them without external calibration or formal reasoning. Connection with the moral facts is not something that we have much empirical feedback to (if any), so we shouldn't trust out intuitions about it.
   2. I think you need to be able to tell a story about not only having intuitions but also why they're reliable. However, I have yet to see a scientifically/ evolutionarily/ empirically informed story that makes me think that we would have reliable morally-truth-tracking intuitions. This is related to a general critique I have for the moral realists: namely, what is the causal story one has about how we get from the truth of the moral item to our accurate beliefs about them?
   3. A lot of the reason that we can trust our intuitions about math or the empirical world is because of the "No Miracles Argument" from the philosophy of science. The "no miracles argument" is a philosophical argument for scientific realism, suggesting that the success of scientific theories, particularly their predictive power, is best explained by their approximate truth. If our best scientific theories were not even close to the truth, the argument goes, their success would be an improbable coincidence, a "miracle."
   4. Similarly, if these theories weren't capturing truth, we would expect to see cases where science doesn't work all the time. The fact that it's consistent and understandable likely implies that there's probably some underlying truth that we're capturing.
   5. These same types of great arguments for scientific realism, however, don't hold in the case of intuitions alone -- we have inconsistencies all the time such that we shouldn't be so shocked when we learn that intuitions aren't the most truth-tracking process of reasoning.
4. Convincing Hunter-Gatherers:
   1. If Matthew (or any other moral realist) went back in time, I think he would have quite a hard time trying to convince some early hunter gatherers with very different psychological dispositions to care about the out-group (though it would be funny to see him try to get them to donate to shrimp welfare, and I would pay lots of money to watch this).
   2. Aside from the fact that he can't speak their language, I think he won't be able to convince them because a lot of his (and my, to be fair) intuitions about the expansion of the moral circle come from psychological dispositions from historically contingent circumstances (for instance, a lot of beliefs about egalitarianism probably come from Capitalism, Christianity, and other historical factors). One can complain about the arbitrariness of certain moral principles all they want (about how they're inconsistently applied, can be dutch-booked, etc), but I still think that the sorts of arguments he might make will only be susceptible to those with certain psychological dispositions (especially ones towards systematizing). Even if these dispositions are correlated with being correct about a bunch of other things (i.e. science, math, etc), this is still insufficient to show that these intuitions are truth-tracking.
   3. Presumably, a moral realist would argue that these hunter-gatherers are just being irrational, and there are some facts that they either know or aren't taking seriously. While this is definitely true for some other things in this case (science probably, etc), I find this hard to believe in the case of morality. What are the facts they are missing? Can one point to the cognitive faculties that they are missing? Can we test this out?
   4. Once there is disagreement and no clear way, in principle, to evaluate who is correct in a certain domain, I think we should give up on believing that there are facts about the matter. If I kept arguing that chocolate was the best flavor and my friend kept arguing that it was vanilla and neither of us could rationally convince the other despite the deep intuitional truths in both directions, I think we should give up and think that there is no right answer. Similarly, if there is moral disagreement with no clear way to resolve it, we should think that there are no facts about morality!
5. Anticipating different experiences: When disagreement arises about a philosophical position, I often think that it is useful to ask what the two positions actually differently anticipate about the world (Note: this is not to say that all disagreements can be cashed out in these terms; I am merely making the point that this is often useful). Let's go into some ways that moral realism, then, might cash out:
   1. All else being equal (i.e. without incentives to form a social contract, cooperate due to repeated games, etc), the orthogonality thesis is false; as agents get smarter, their beliefs converge on the moral facts. This includes AGI. I look forward to testing out this empirical claim, and I would be quite surprised if it holds.
   2. In our evolutionary history, in addition to us learning more about math/ logic/ other stuff via pattern matching, there should be some evidence that we learned to generalize about morality in similar ways that track the moral truth. It also wouldn't be sufficient to show that the generalizations we make from math/ other types of reasoning alone started to apply to morality -- we would need to give some plausible story of how we achieved these beliefs. 
6. In favor of future Tuesday indifference: It's much easier to say that moral facts exist if one could say that there are rationality constraints on preferences (ways that one's preferences can be irrational from a perspective that is not merely from their own preferences). Once one accepts this, it becomes easier to talk about good and bad outside of one's preferences (a position which I don't think makes much sense). But here's my crazy take: I don't think there are rationality constraints on preferences. Famously, Derek Parfit thought this was false and his argument goes as follows: imagine there is a man who walks into the doctor's office to get a surgery on Tuesday, and the doctor tells him "oh, sorry man. We won't get the anesthetic until tomorrow." The man then replies "that's totally fine -- I can still do it today." the doctor, looking considerably confused, says "I don't think you understand; this surgery is incredibly painful, and without anesthetic, you will suffer intense agony for many hours." The man once again calmly responds "nah, it's fine. I'm future Tuesday indifferent, and there will be some pain tomorrow vs immense pain today, but I am totally indifferent to the pain today. Therefore, I'd prefer to endure the pain today." "Whatever you say," the doctor says. He goes throughout the surgery and is screaming in agony and regret the entire time. The question now is: do we think the man did something irrational.
   1. My answer: sorta; I mean, it depends on the case, right? If we're talking about someone who is merely saying that they have future tuesday indifference but  internally doesn't (as in, someone's preferences do not actually align with what they claim/ do; think about the irrationality of the heroine addict), I would say that they are not being rational (insofar as they'd like to accomplish their goals, they shouldn't make decisions as if they are future tuesday indifferent when they are not). However, if the man is actually indifferent (and therefore has a very different psychological profile than the rest of us), I'm not sure on what grounds one can claim that this person is irrational.
   2. I think, oftentimes, our intuitions about these matters go wrong because we over-extrapolate from our own preferences and the preferences of people around us. For instance, we probably wouldn't say that a robot who was programmed in such a way to respond like this but who's reward function was actually indifferent to pain on Tuesday's is being irrational. Similarly, in the case of a very foreign alien. Similarly, we should think of the guy with future Tuesday indifference as someone who merely has a very strange psychological disposition or is confused about their own preferences. Once we start viewing it like this, I think, it becomes much harder to talk about the irrationality of their preferences.
      1. For those that think one can't be wrong about their own preferences, imagine the following case: a poor and depressed girl named Tara grows up being told that money will solve all her problems. Tara then becomes extremely wealthy early on and starts to feel happy again. Tara, however, becomes so addicted to money that she starts sacrificing her happiness for money; she works long hours, spends no time with family, and is totally depressed. Here, you probably wouldn't say that her actual preference is to like money more than happiness and that she is just being rational with respect to her preferences; more likely, you'd say that she is confused about her true preferences. 

On the other hand, there are a few arguments that I do take pretty seriously (that Matthew didn't go over) Here are a few: 

1. Deliberative Indispensability: David Enoch has what I take to be, one of the best arguments against anti-realism; it is called the deliberative indispensability argument.
   1. Before going into the argument itself, one should first understand the companions in guilt argument, which is doing some behind the scenes work in the deliberative indispensability argument. This is a common argument across philosophy, but Enoch makes this argument for moral realism by accepting normative realism. The structure of the argument goes as follows:
   2. Suppose you say that theory X is implausible because it has trait b. However, you also believe theory Y which also has trait b. In this case, you should either accept theory X and think that trait b is not defeating in itself or reject both theories on the grounds that trait b is defeating. To apply the trait b as a defeater in one case and not another, the argument goes, is applying a rule inconsistently and is therefore epistemically intolerable.
   3. Enoch argues that to deliberate about anything requires some form of normative realism. How does this work? I think this argument is best explained in dialogue format.
      1. David: grabs a bottle of water.
      2. Sarah: Oh David, I didn't know you were a normative realist?
      3. David: Wait, what? I'm not; I'm just drinking water.
      4. Sarah: Why are you drinking water?
      5. David: Because I'm thirsty.
      6. Sarah: Why do you care about thirst?
      7. David: Uhhh, because otherwise I would die.
      8. Sarah: Okay. Why should you not die?
      9. David: I don't prefer it.
      10. Sarah: Why do you act in accordance with your preferences?
      11. David: Because I like them... idk... can you stop...
      12. Sarah: Who cares if you like them? Can't you just do the thing that you exactly don't like?
      13. David: Yea, but I wouldn't like that.
      14. Sarah: That's circular.
      15. David: Idc. I like it and of course I should do my preferences; if not for my preferences, what else should I do?
      16. Sarah: So you are a normative realist!
      17. David: What? When did I say that?
      18. Sarah: You said you should do your preferences because that's the only way to ground why you should do your preferences as opposed to the exact opposite.
      19. (Maybe one can respond) David: I am merely making a descriptive claim. I am merely observing (rather than prescribing) that I will tend to take actions that achieve my preferences.
      20. Sarah: I think this is false. You could take actions that aren't in line with your preferences. Watch! Sarah bangs her head on the Table. It then seems like you are implicitly smuggling in a form of normativity whenever you deliberate!
   4. Once you accept this form of normative realism, the argument goes, you are subject to the same kind of evolutionary debunking arguments as the moral realist. In order to deliberate, then, one must admit that there is another kind of explanation (a normative one) associated with forms of normative realism that evolution itself can't merely debunk. Once you accept this kind of explanation, you are then admitting that evolutionary debunking arguments are actually not sufficient to reject an argument. If you accept that debunking arguments of this sort aren't sufficient, you shouldn't think that they are sufficient in the moral case.
      1. While I think this argument is pretty good, I think I get off the train at believing that any form of normative realism is required for deliberation. When I say that I do my preferences because I like them, I still think this is referring to a merely descriptive claim: I just tend to do my preferences. My sense is when one gets this deep into the "why did you choose that x thing as opposed to that y thing" language, it actually becomes the free will debate (Sarah is essentially asking if there are some things, on a deep level, which I can choose and are not merely descriptive claims), which I reject (I will not go into the reasons here). 
         1. The reason I still think this is a good argument is because it seems hard to exactly hone in on when this debate becomes about free will vs something else (if ever), though I should probably spend more time thinking about it.
2. The Normative Realist's Wager: Say there are two ways that the world could be: one where moral realism is true and one where it is false. Given that we're in a world where moral realism isn't true, there is no value associated with either acting as though moral realism is true or acting as though it's false. On the other hand, if we're in the world where moral realism is true, there is negative value in the world where we act like moral realism isn't true and positive value in the world where we act like it is true. Regardless of the world, then, it is as least as good (if not better) to act as though moral realism is true. (Note: While this argument resembles Pascal's wager, I don't think it is subject to many of the classical and good counter arguments: the many gods objection, pascal's mugging, the mixed strategies objection, the greater gods objection, etc).
   1. For any set of objective values associated with action x, one can make a set of objective values associated with y, where y is the exact opposite action of x. For instance, one action (x) is that killing babies is bad, and the other action (y) is that killing babies is good. Therefore,  we don't actually have dominance, and there are an equal number of worlds where killing babies is both good and bad.
      1. However, I don't find this objection plausible. Presumably, one's p(that it's bad to torture babies) > p(it's good to torture babies), even if only slightly (because of religion, deference to experts, etc). If this is the case, while you don't get weak dominance (as there are instances where acting as though moral realism is true for one action can be bad in another world), there are more worlds where not killing babies produces more value, giving you greater reason to act in accordance with it.
   2. Pascal's mugging.
      1. Your credence shouldn't be low enough for pascal's mugging to actually kick in. Most philosophers are moral realists - under almost all reasonable theories about deference to experts, your credence should then be well above the level where you start avoiding things for pascal's mugging reasons.
   3. This decision calculus already assumes some normative realism to begin with. It assumes, in some sense, that we ought to use the decision process with the highest expected value/ we should take actions that dominate in game theoretic terms. However, this is already assuming some form of normative claims, making it a circular argument (at least for those that don't already buy normative realism -- that we should act in ways that maximize expected value or that we ought to take certain actions that dominate others).
   4. When we are talking about making bets that maximize expected value or actions that dominate others, one needs probabilities and values that pay-off. In the moral case, however, it's hard to think about what exactly the pay-off you get out acting in accordance with the moral truth. Does it give you points in the moral dimension? Does God give you reward after death? Etc. Given that there are no actual payoffs, one might argue, there is no dominance when acting in accordance with moral realism.
   5. What if there are other forms of value outside of moral realism that are not being accounted for in this setup (like constructive value, which some philosophers buy).  While one can say that some form of objective value is better in EV terms because it is objective, it's not actually clear how to cash this out or even talk about how one of these values could be 'more important' than another. Being able to compare these different sorts of values gives moral uncertainty vibes, which is infamously quite difficult to deal with. (the awesome Joe Carlsmith makes this argument better than I do here, which you can read here). 

This is not to say that I'm not EA nor is it to say that I don't think others should be EA. In fact, quite the opposite; I run UChicago EA, plan to choose my career around EA principles, am vegan, and have made many other life sacrifices because of EA. While my anti realism has some influenced some of my normative ethics, I don't think my beliefs about cause prioritization actually differ that much from the average realist EA because of it. What I would say, however, is that this form of anti realism might imply EA os more of a preference than some objective truth; and I think this is totally fine! This doesn't mean that you can't or shouldn't convince others to be EA as well -- we argue with people about what follows from their stated preferences about how the world should be all the time! We need this type of reasoning for all types of institutions (designing markets, political institutions, etc). You can still think that most of the normal population underrate that they are always making trade-offs, have scope neglect, and that there are good arguments for thinking that, given peoples' values, wellbeing is by far the most important (or the only) value.

If anyone wants to share ideas more (constructive, critique, tell me how awesome these arguments are, etc) about this topic (especially on a call!), email me at dnbirnbaum@uchicago.edu, and we can set up a time! Also, as mentioned before, read my Substack!