Great post, I hadn't heard the comparison to "relations of things" before. I previously thought the best comparison for moral reasons were to normative rational truths (like if p implies q - then given p you should believe q). Feels like normative rational truths are in between the truthiness of "relations of things" and "moral reasons" in terms of how easy they are to swallow.
The issue with using examples is it may be controversial that it exemplifies a non-natural moral fact. In the case "I ought not kick a wall on a whim" you are in fact explicitly owing ("ought" is an archaic "owed") the avoidance of kicking the wall to an interest in avoiding pain. This looks just like stance-dependence (that is, the action is ought-to/owing-to some cares/concerns) to me! And of course you can't cash out with "pain" as the ultimate ground since there are cares/concerns for which pain is rationalized. We can even come to enjoy pain, e.g., the discomfort I have while jogging used to bother me very much, but now it gives me a sense of accomplishment in overcoming something. I wouldn't wish it away!
If we say "ought" is always "indexed" ("owing-to" some cares & concerns), we stay careful to safely express by what things actions are rationalized, prioritized, evaluated, etc. And it is nearly always the case that when these sorts of basic clear examples are provided, the cares & concerns to which they are indexed are quite easy to point out!
For clarity, a distinction should be made between different sorts of concerns one might have with the ability to communicate moral concepts. One might be concerned that moral non-naturalists are committed to the view that moral concepts are incapable of being communicated in a way that is different than other concepts, which might be cause for suspicion. This sort of concern, which is partly a priori, is how I interpreted Lance Bush's remarks. But this concern is addressed by pointing out that moral non-naturalists can simply say that they can be communicated by means of examples, which is one of the main ways concepts are communicated -- and like those other cases, it cannot be communicated by definition.
But once that concern is addressed, the moral anti-realist might have an a posteriori concern about the ability of moral non-naturalists to communicate their concepts -- the anti-realist might from his own experience conclude that these concepts cannot in fact be communicated. I acknowledged this in the original essay, saying that the moral anti-realist might simply fail to grasp any relation in the examples at all, or be convinced that the only sort of relation is a stance-dependent one. I did give some reasons for thinking that it is not a stance-dependent one.
Anyway, you seem to be using etymological reasoning as evidence that the relation is in fact stance-dependent. First, I don't think that etymology is a reliable guide to how words are currently used. The suffix 'phile' derives from a word meaning love, but is often used for sexual interests, for example. But beyond that, I don't think what you said actually supports anti-realism anyway, although it would maybe support ethical egoism. A moral non-naturalist could say there is an owing relation such that a person owes making a certain choice to a certain interest he has (such as avoiding pain). If all moral obligations are dependent on interests you have, then I suppose that this would entail a form of ethical egoism. But you could still hold that facts about owing are not further reducible to non-normative facts, and be a moral non-naturalist. For a metaethical theory to be stance-dependent in an anti-realist way, it must be that all moral facts are reducible to facts about stances, not merely that stances play some role in the moral facts. One could be a moral non-naturalist and hold to a normative ethical theory in which you should maximize your experiences involving positive attitudes and minimize the negative ones.
Regarding your point about pain, I am actually quite open to thinking that the perception of pain is not intrinsically bad but that it depends on what you feel regarding the pain. However, again, this is consistent with moral non-naturalism; and in this case, it is even consistent with rejecting ethical egoism -- one could hold that you should maximize not only your own experiences where you have a positive attitude and minimize the ones where you have a negative attitude, but that you should also do this for other people. That being said, I'm not sure if the example you gave is much evidence for such a position; it might be that there is something bad about the experience, but also that the sense of accomplishment outweighs it.
What is your explanation for how we come to have moral knowledge? For example, I could imagine someone observing that evolution does a pretty good job explaining why our beliefs about many universals track the truth, but not why our beliefs about morality would track the moral truth.
I'm a substance dualist, so I don't think that our mental faculties are purely the product of evolution.
However, I don't see why moral knowledge would pose more of a problem for physicalists than non-moral knowledge. There is no need to suppose that there is some unique faculty for moral knowledge. And while moral knowledge might not be evolutionary advantageous, having a general rational faculty that allows us to come to know a variety of truths presumably is.
Well, the reason it would pose *more* of a problem for physicalists, in my mind, is that there seems to be a clear advantage offered by understanding other kinds of truths (e.g., the relative positions of things or mathematics), whereas there is no clear advantage to knowing moral truths. Of course, as you say, we could evolve a general faculty that happens to arrive at moral truths, just as we evolved eyes to see things on Earth but can also see the moon — although I’m not sure if the faculties that allow us to understand different kinds of universals are very related at all/I’m not sure what the connection would be (whereas it’s relatively clear why “seeing the moon” and “seeing things on Earth” require similar faculties). What would your account be of why it’s plausible that the same general faculty allows us (for the physicalist) a kind of epistemic access to many different universals?
General mental faculties allow for handling novel situations; might be more probable due to fewer adaptions in total (if developing a general ability to pick out colors has the same probability as developing the ability to pick out blue, then having the ability to pick out multiple colors would be more likely with a general faculty); and might be more efficient (not all humans are going to end up in the same situations; even if the situation is not entirely novel from the perspective of humanity, it could be inefficient to have innate abilities that are never going to be used).
Putting aside evolutionary speculation, it seems pretty clear that we do have mental faculties with some degree of generality. For example, it seems likely that we have some sort of general ability to recognize different colors rather than have innate color concepts. If I am not mistaken, the Japanese previously did not distinguish between blue and green, but have since adopted the distinction, which suggests that they had the ability to recognize the different colors if it was pointed out to them. You might suggest that we have different mental faculties for different classes of things (like colors, or relative positions of objects), but I don't see any reason to think that. If you admit that there is some degree of generality, why be skeptical of a greater degree?
Maybe what you are asking for is some metaphysical account of how physicalism can account for a general rational faculty. I don't have an account of that, and I doubt whether there would be a satisfactory account. But I would see that as much more of a problem for physicalism than for moral platonism, or for rationalism and platonism more generally.
To clarify, I think the etymological linkage is helpful, but not prescriptive. When we think about it "ought as owing" it reminds us to look for one or more motivating "debtees," which our contemporary language & thinking habits tend to neglect.
I do not know why you said that for a metaethical theory to be stance-dependent in an antirealist way it must be that all moral facts are reducible to facts about stances, and not merely that stances play some role in the moral facts. Under antirealism stance-dependence just means that stances are unavoidably involved, but stance-independent facts (like how germs work, how tides work, etc.) can also be involved in a moral evaluation or proposition. Perhaps that is the hangup here.
I think that your notion of stance-dependence is too broad. Consider preference utilitarianism, a pretty widespread normative ethical theory. According to this theory, all first-order facts about moral obligations are dependent in a sense on preferences. For example, if it is the case that I should not rape someone, it must be because this goes against someone's (like the rape victim's) preferences. I guess that preferences could be considered a type of stance. So if preference utilitarianism is true, any fact about what one is obligated to do involves stances. I don't think that preference utilitarianism is incompatible with moral realism, though. And even if we say that all the facts depend on the agent's preferences instead of people's preferences in general, such a view could still be consistent with moral non-naturalism (though it would be a form of ethical egoism).
For the stance-dependence to entail anti-realism, there must be something more than the "stance-dependence" involved in preference utilitarianism. You must say that moral statements are translatable (at least in principle) to stance statements, and that moral facts are simply facts about stances. For example, 'I should not rape this person' must mean something like 'I dislike the idea of my raping this person' or 'My culture dislikes the idea of my raping this person'. Even if you think I am going too far by saying that it must be reducible, at the very least, there should be a biconditional relation between how some person or group thinks about the choice, and whether it is right or wrong (anyway, once one acknowledges such a biconditional relation, there would be less reason to deny reducibility). What you said about owing seemed to entail that moral facts were dependent on how the agent feels about things, but not necessarily dependent on how the agent feels about the choices themselves, so someone might dread and otherwise think negatively about going to the store, but still owe it to his interest in having delicious food to do so, for example.
Preference utilitarianism is typically considered a form of moral antirealism by virtue of its stance-dependence.
Your view of moral antirealism appears to be something like expressivism, which is just one of several different antirealist views. Moral antirealism is just the position that the number of true moral propositions that are completely stance-independent is zero.
Expressivism is the view that moral sentences are not truth-apt but merely expressions of attitudes. For example, 'Murder is wrong' would mean something like 'Boo murder!'; the latter sentence is neither true nor false, similar to how 'Please shut the door' is neither true nor false.
The view I had in mind was instead moral relativism/subjectivism, which affirms the truth-aptness of at least some moral sentences. Moral relativism is not the only type of moral anti-realism, but it is the one that is distinguished from realism by its stance-dependence, so it was the relevant one.
But even if I grant your (as far as I can tell, idiosyncratic) definition of 'moral anti-realism', what is the significance of this? Surely preference utilitarianism is consistent with saying that there are irreducible normative facts, and that moral statements can be true regardless of the stances that people hold with regard to the choice that is being evaluated. The sort of 'stance-dependence' you have in mind is consistent with the sort of moral ontology and epistemology that I defend -- one can still say that there is an irreducible relation of obligation, and that having an obligation to make a choice is not merely having a certain attitude towards doing it, and one can also deny that it is biconditionally related to anyone or any group's having a certain stance towards the choice.
Further, moral naturalists, relativists, non-cognitivists, and error theorists will have the same sort of objections against a non-naturalist "anti-realism" that is 'stance-dependent' in your sense, as they would against other forms of moral non-naturalism. While the non-naturalist "realist" might now be less opposed to "anti-realism" as a whole due to the expansion of the meaning, he can be equally opposed to moral naturalism, non-cognitivism, error theory, and relativism.
What you said about an 'owing' relation is also consistent with the broad sort of moral ontology and epistemology that a non-naturalist would endorse. Actually, although I previously suggested it would entail ethical egoism, this is only true if one can only have the owing relation to one's own interests; if one can 'owe' things to other's interests, it is even consistent with denying egoism.
Really good post! I find these sorts of criticisms very frustrating too and I'm glad there are more realists on here pushing back on them.
Great post, I hadn't heard the comparison to "relations of things" before. I previously thought the best comparison for moral reasons were to normative rational truths (like if p implies q - then given p you should believe q). Feels like normative rational truths are in between the truthiness of "relations of things" and "moral reasons" in terms of how easy they are to swallow.
The issue with using examples is it may be controversial that it exemplifies a non-natural moral fact. In the case "I ought not kick a wall on a whim" you are in fact explicitly owing ("ought" is an archaic "owed") the avoidance of kicking the wall to an interest in avoiding pain. This looks just like stance-dependence (that is, the action is ought-to/owing-to some cares/concerns) to me! And of course you can't cash out with "pain" as the ultimate ground since there are cares/concerns for which pain is rationalized. We can even come to enjoy pain, e.g., the discomfort I have while jogging used to bother me very much, but now it gives me a sense of accomplishment in overcoming something. I wouldn't wish it away!
If we say "ought" is always "indexed" ("owing-to" some cares & concerns), we stay careful to safely express by what things actions are rationalized, prioritized, evaluated, etc. And it is nearly always the case that when these sorts of basic clear examples are provided, the cares & concerns to which they are indexed are quite easy to point out!
Thanks for the comment.
For clarity, a distinction should be made between different sorts of concerns one might have with the ability to communicate moral concepts. One might be concerned that moral non-naturalists are committed to the view that moral concepts are incapable of being communicated in a way that is different than other concepts, which might be cause for suspicion. This sort of concern, which is partly a priori, is how I interpreted Lance Bush's remarks. But this concern is addressed by pointing out that moral non-naturalists can simply say that they can be communicated by means of examples, which is one of the main ways concepts are communicated -- and like those other cases, it cannot be communicated by definition.
But once that concern is addressed, the moral anti-realist might have an a posteriori concern about the ability of moral non-naturalists to communicate their concepts -- the anti-realist might from his own experience conclude that these concepts cannot in fact be communicated. I acknowledged this in the original essay, saying that the moral anti-realist might simply fail to grasp any relation in the examples at all, or be convinced that the only sort of relation is a stance-dependent one. I did give some reasons for thinking that it is not a stance-dependent one.
Anyway, you seem to be using etymological reasoning as evidence that the relation is in fact stance-dependent. First, I don't think that etymology is a reliable guide to how words are currently used. The suffix 'phile' derives from a word meaning love, but is often used for sexual interests, for example. But beyond that, I don't think what you said actually supports anti-realism anyway, although it would maybe support ethical egoism. A moral non-naturalist could say there is an owing relation such that a person owes making a certain choice to a certain interest he has (such as avoiding pain). If all moral obligations are dependent on interests you have, then I suppose that this would entail a form of ethical egoism. But you could still hold that facts about owing are not further reducible to non-normative facts, and be a moral non-naturalist. For a metaethical theory to be stance-dependent in an anti-realist way, it must be that all moral facts are reducible to facts about stances, not merely that stances play some role in the moral facts. One could be a moral non-naturalist and hold to a normative ethical theory in which you should maximize your experiences involving positive attitudes and minimize the negative ones.
Regarding your point about pain, I am actually quite open to thinking that the perception of pain is not intrinsically bad but that it depends on what you feel regarding the pain. However, again, this is consistent with moral non-naturalism; and in this case, it is even consistent with rejecting ethical egoism -- one could hold that you should maximize not only your own experiences where you have a positive attitude and minimize the ones where you have a negative attitude, but that you should also do this for other people. That being said, I'm not sure if the example you gave is much evidence for such a position; it might be that there is something bad about the experience, but also that the sense of accomplishment outweighs it.
What is your explanation for how we come to have moral knowledge? For example, I could imagine someone observing that evolution does a pretty good job explaining why our beliefs about many universals track the truth, but not why our beliefs about morality would track the moral truth.
I'm a substance dualist, so I don't think that our mental faculties are purely the product of evolution.
However, I don't see why moral knowledge would pose more of a problem for physicalists than non-moral knowledge. There is no need to suppose that there is some unique faculty for moral knowledge. And while moral knowledge might not be evolutionary advantageous, having a general rational faculty that allows us to come to know a variety of truths presumably is.
Well, the reason it would pose *more* of a problem for physicalists, in my mind, is that there seems to be a clear advantage offered by understanding other kinds of truths (e.g., the relative positions of things or mathematics), whereas there is no clear advantage to knowing moral truths. Of course, as you say, we could evolve a general faculty that happens to arrive at moral truths, just as we evolved eyes to see things on Earth but can also see the moon — although I’m not sure if the faculties that allow us to understand different kinds of universals are very related at all/I’m not sure what the connection would be (whereas it’s relatively clear why “seeing the moon” and “seeing things on Earth” require similar faculties). What would your account be of why it’s plausible that the same general faculty allows us (for the physicalist) a kind of epistemic access to many different universals?
General mental faculties allow for handling novel situations; might be more probable due to fewer adaptions in total (if developing a general ability to pick out colors has the same probability as developing the ability to pick out blue, then having the ability to pick out multiple colors would be more likely with a general faculty); and might be more efficient (not all humans are going to end up in the same situations; even if the situation is not entirely novel from the perspective of humanity, it could be inefficient to have innate abilities that are never going to be used).
Putting aside evolutionary speculation, it seems pretty clear that we do have mental faculties with some degree of generality. For example, it seems likely that we have some sort of general ability to recognize different colors rather than have innate color concepts. If I am not mistaken, the Japanese previously did not distinguish between blue and green, but have since adopted the distinction, which suggests that they had the ability to recognize the different colors if it was pointed out to them. You might suggest that we have different mental faculties for different classes of things (like colors, or relative positions of objects), but I don't see any reason to think that. If you admit that there is some degree of generality, why be skeptical of a greater degree?
Maybe what you are asking for is some metaphysical account of how physicalism can account for a general rational faculty. I don't have an account of that, and I doubt whether there would be a satisfactory account. But I would see that as much more of a problem for physicalism than for moral platonism, or for rationalism and platonism more generally.
To clarify, I think the etymological linkage is helpful, but not prescriptive. When we think about it "ought as owing" it reminds us to look for one or more motivating "debtees," which our contemporary language & thinking habits tend to neglect.
I do not know why you said that for a metaethical theory to be stance-dependent in an antirealist way it must be that all moral facts are reducible to facts about stances, and not merely that stances play some role in the moral facts. Under antirealism stance-dependence just means that stances are unavoidably involved, but stance-independent facts (like how germs work, how tides work, etc.) can also be involved in a moral evaluation or proposition. Perhaps that is the hangup here.
I think that your notion of stance-dependence is too broad. Consider preference utilitarianism, a pretty widespread normative ethical theory. According to this theory, all first-order facts about moral obligations are dependent in a sense on preferences. For example, if it is the case that I should not rape someone, it must be because this goes against someone's (like the rape victim's) preferences. I guess that preferences could be considered a type of stance. So if preference utilitarianism is true, any fact about what one is obligated to do involves stances. I don't think that preference utilitarianism is incompatible with moral realism, though. And even if we say that all the facts depend on the agent's preferences instead of people's preferences in general, such a view could still be consistent with moral non-naturalism (though it would be a form of ethical egoism).
For the stance-dependence to entail anti-realism, there must be something more than the "stance-dependence" involved in preference utilitarianism. You must say that moral statements are translatable (at least in principle) to stance statements, and that moral facts are simply facts about stances. For example, 'I should not rape this person' must mean something like 'I dislike the idea of my raping this person' or 'My culture dislikes the idea of my raping this person'. Even if you think I am going too far by saying that it must be reducible, at the very least, there should be a biconditional relation between how some person or group thinks about the choice, and whether it is right or wrong (anyway, once one acknowledges such a biconditional relation, there would be less reason to deny reducibility). What you said about owing seemed to entail that moral facts were dependent on how the agent feels about things, but not necessarily dependent on how the agent feels about the choices themselves, so someone might dread and otherwise think negatively about going to the store, but still owe it to his interest in having delicious food to do so, for example.
Preference utilitarianism is typically considered a form of moral antirealism by virtue of its stance-dependence.
Your view of moral antirealism appears to be something like expressivism, which is just one of several different antirealist views. Moral antirealism is just the position that the number of true moral propositions that are completely stance-independent is zero.
Expressivism is the view that moral sentences are not truth-apt but merely expressions of attitudes. For example, 'Murder is wrong' would mean something like 'Boo murder!'; the latter sentence is neither true nor false, similar to how 'Please shut the door' is neither true nor false.
The view I had in mind was instead moral relativism/subjectivism, which affirms the truth-aptness of at least some moral sentences. Moral relativism is not the only type of moral anti-realism, but it is the one that is distinguished from realism by its stance-dependence, so it was the relevant one.
But even if I grant your (as far as I can tell, idiosyncratic) definition of 'moral anti-realism', what is the significance of this? Surely preference utilitarianism is consistent with saying that there are irreducible normative facts, and that moral statements can be true regardless of the stances that people hold with regard to the choice that is being evaluated. The sort of 'stance-dependence' you have in mind is consistent with the sort of moral ontology and epistemology that I defend -- one can still say that there is an irreducible relation of obligation, and that having an obligation to make a choice is not merely having a certain attitude towards doing it, and one can also deny that it is biconditionally related to anyone or any group's having a certain stance towards the choice.
Further, moral naturalists, relativists, non-cognitivists, and error theorists will have the same sort of objections against a non-naturalist "anti-realism" that is 'stance-dependent' in your sense, as they would against other forms of moral non-naturalism. While the non-naturalist "realist" might now be less opposed to "anti-realism" as a whole due to the expansion of the meaning, he can be equally opposed to moral naturalism, non-cognitivism, error theory, and relativism.
What you said about an 'owing' relation is also consistent with the broad sort of moral ontology and epistemology that a non-naturalist would endorse. Actually, although I previously suggested it would entail ethical egoism, this is only true if one can only have the owing relation to one's own interests; if one can 'owe' things to other's interests, it is even consistent with denying egoism.