2. Pluralisms and ideals

Value pluralism is standardly a metaphysical thesis about value. It claims that different values have no lexical ordering and cannot be reduced to some more fundamental value.Footnote ⁶ Since I want my project to have a broad metaethical appeal, I do not want to take sides in that debate. However, when we look at the most influential arguments for metaphysical value pluralism, we can see a form of epistemic value pluralism that should be philosophically fruitful.

Two popular arguments for pluralism move from phenomenological premises to metaphysical conclusions. One argument relies on the intuition that values just seem different. Chang (Reference Chang and Wright2001) cites the difference between the value of philosophical insight and the value of eating delicious cheesecake (22). Ross points to the difference between pleasure and virtue (Ross Reference Ross1988: 134). Intuitively, these values seem to have nothing in common. Hence, value is many and not one. Williams (Reference Williams and Williams1973) and Stocker (Reference Stocker1990) argue from the phenomenon of rational regret. Suppose you have to choose between marrying a childhood sweetheart or not. Say you marry and, many years later, are happy with the decision. It seems perfectly rational to regret missing out on all those other romantic interactions and other adventures. This rational regret shows you miss out on some value that is not compensated for by your many years of happy married life. Again, value is many, not one. In both cases, we go from the way things appear to the way things are. Now, none of the philosophers in this debate makes this flat-footed move. Rather, they claim that metaphysical value pluralism is the best explanation of these appearances. Value monists deny this and offer different explanations. And from there, things get complicated.

We do not need to settle this debate. We are doing moral epistemology, and these arguments open the door to a different kind of pluralism. Epistemic value pluralism is the claim that there are a variety of ways to come to know values. Epistemic value pluralism is a logically weaker claim, and it can do some interesting philosophical work. The move from phenomenological pluralism to epistemic pluralism seems easier to defend. The move is now justified by empiricism. The person who attains philosophical insight knows what that is like. The happily married person knows the goods that come from a loving marriage but has no real grasp on what those other romantic pursuits might have been like. Epistemic value pluralism claims that through different experiences, we come to know different values in the world. I will argue later that this is why some disagreements can be informative. Other people come to know values we can barely imagine. These disagreements may be impossible to settle because we cannot share the same evidence. Nevertheless, there could be a metaphysical fact of the matter, but we will not settle that here. I will argue that, in the right conditions, these sorts of disagreements can provide higher-order evidence about our capacity for normative judgment.

One might think that moral realism requires convergence, so it cannot be open to these sorts of disagreements.Footnote ⁷ Against this, I will try to show that the philosopher most concerned with making moral disagreement disappear, Derek Parfit, was only committed to a very mild form of convergence. This is Parfit’s Convergence Claim (or CC):

I have italicized all the words that weaken this claim about convergence. The most obvious is that we would have similar normative beliefs. Similarity is not identity. What counts as enough similarity? I read Parfit as only intending to rule out a disagreement over claims like “torturing children merely for fun is wrong.” That is the most common sort of claim he uses against subjectivism, which he takes to be his main opponent. We can accept that sort of claim and accept epistemic pluralism. When we turn our attention to how ideals change our normative concepts and what counts as a distorting influence, we will see that Parfit’s CC is compatible with a lot of disagreement. We can embrace epistemic value pluralism without fear of losing realism.Footnote ⁸

One might contend that Parfit believes he has discovered the correct moral theory. Surely, he thinks that if people understood his arguments, then they would all agree. Here, we make the familiar move that the act utilitarian makes regarding the distinction between the good and the praiseworthy. The act utilitarian does not need to claim that we always just apply the principle of utility in good deliberation. Likewise, Parfit should agree that not all good deliberation requires directly applying his theory. To borrow his metaphor, we can climb the mountain from different sides. There are different ways to lead a good life, and different lives involve different types of deliberation. The way a soldier deliberates will be different from the way an artist deliberates. Not only because they face different circumstances, but also because they are oriented toward different values. We will say more about this when we come to Buss’s work, and we will revisit this worry about realism as necessary.

With this understanding of epistemic pluralism, we can now turn to how ideals lead to more complicated disagreements. Consider the ideals that P.F. Strawson describes:

Here, we find disagreement. However, differing ideals are not taken as signs of mistakes, though. Different ideals embrace a way of life we reject, but we have a positive evaluation of it nevertheless. Strawson appears to be making psychological claims about ideals. I find this plausible, but I am interested in defending an epistemic claim about normativity. Different kinds of lives deliver different insights into value. Engaging with these different lives can be crafted to form higher-order evidence about our normative competence. Such interactions can give us higher-order evidence about the quality of our judgment. Disagreement might allow for vindication by finding our values to be worthwhile from a perspective different from our own.

Strawson thinks ideals not only conflict, but they could also never be reconciled with one another. He writes:

Strawson talks of truth here, so he is advocating for metaphysical value pluralism. I do not want to defend that claim, so I will ignore it. If we focus on the epistemic thesis, then we just need to defend the claim that we cannot “see life steadily” while also seeing it as a whole. I take “see” here to mean a way of knowing. The claim is that we can know values in different ways, and no one can know all values in the same way. Taking this on, my epistemic value pluralism is the claim that there are a variety of ways to come to know values, and some of these ways of knowing are incompatible with others. To take one striking example, both Churchill and Gandhi knew something about living a good life. They disagreed over important moral issues. Echoing Strawson, I think it is futile to think we can systematize these men’s ideals. In denying the systematization of these ideals, we are denying that we can harmonize the values of Churchill and Gandhi without losing something essential. There is more than one way to lead a good life, and these lived experiences generate different moral knowledge. I defend this claim in the next section.

3. Disagreement and higher-order evidence

One could argue that when we state the moral truths sufficiently abstractly, we will find that these disagreements disappear. It might be thought that there is still one set of moral truths that can be systematized. Again, this metaphysical claim may well be true. However, to take this abstract point of view greatly distorts how we engage with value and is not how good people deliberate. Buss (Reference Buss2023) deftly explains this fact. She notes that “all of us have personal ideals” that we come to internalize. That is, they play an essential role in how we see the world and determine our “apprehension of certain requirements” (233). By internalizing these ideals, we allow ourselves to respond directly to facts as reasons. If we fail to do this, then we succumb to having one thought too many or become a normative fetishist (235–238). There is something deficient in an agent who does not directly respond as their ideals dictate. If one has to consult whether acting in line with one’s ideal is permissible, then that is one thought too many. You have not internalized the role of being a good spouse unless you can non-derivatively act in a way a good spouse does. If, instead of acting as a good spouse does, you consult what one has the most reason to do, then you have a normative fetish. You care not about your spouse, but only insofar as they have normative significance to you. Buss argues that while we cannot reject thin concepts, we cannot live by them alone. We need deep concepts to guide our lives. Now there may be times when we need to be more reflective. Evaluation of our ideals is possible, but this comes with some cost. Acting under substantive ideals is necessary for “acquiring and maintaining a substantial identity” (234). To lose this would be a great loss.

I contend this is true because some moral judgments are formative. Smith (Reference Smith2002) argues that moral judgments can vary along three different dimensions. There is certitude, which is a measure of confidence at a given moment. There is robustness, which is a measure of stability over time. And there is importance, which is a measure of strength. Moral judgments can also be formative in that they change how we relate to reasons. Adopting an ideal is formative. That is a measure of how moral judgment develops and shapes other values. The features that Smith notes vary independently of it. Consider the happily married person noted earlier. That person could lack certitude at any point in the marriage. The robustness will likely wax and wane throughout the marriage. The person could certainly recognize that other moral considerations are more important. Nevertheless, it could still play a powerful formative role in the adoption of other values. It would be natural for the spouse to care more about what is happening in the spouse’s profession than others do. A non-philosopher spouse married to a philosopher may care about the trends in philosophy even when they do not directly impact the spouse. Note that a judgment that wanton cruelty is wrong could have certitude, robustness, and importance, while not being particularly formative.

By calling it a dimension of moral judgment, I am saying it varies by degrees. It is not like other purported features of moral judgments, like being overriding or normative. Those are binary features. A decision to be more attentive to the beauty of classical music is formative but limited in scope. One gains access to new aesthetic values, but that only touches one aspect of life. Being a spouse covers much more ground and introduces a wider spectrum of values. Some ideals, like being a monk, might be all-encompassing.Footnote ⁹

Here is an example of how ideals as formative judgments change our relation to reasons. We all should believe the loss of innocent lives in war is terrible. We all have a very strong reason to prevent it. However, the reason against collateral damage in a just war will appear differently to a general than it does to a war journalist. They both find it to be a contemptible fact of war that counts strongly against various actions. However, a general has to develop a measure of insensitivity to effectively lead the war effort. Dwelling on innocent deaths would be a distorting influence that leads to depression and failure. The war journalist has to capture the collateral damage in all its gruesomeness to accurately depict the great horrors of warfare. To not depict it because it is horrifying would be a distorting influence that lets the politicians and the citizens who have sanctioned the war effort off the hook. The ideals of a general and a war journalist change what counts as a distorting influence and how we attend to different normative concepts. The reasons we come to know intimately, rather than merely being aware of them, will depend on the ideals we adopt.

These differences will lead to a lack of convergence, but that does not undermine realism. We can expect the general and the war journalist to disagree, and it is good that they do so. They have internalized values that are good for both of them to have. Winning a just war and depicting the moral costs of that war may be mind-independent goods. We can also render this disagreement in a sentimentalism-friendly way. We can accept this sort of pluralism even if we disagree with the truth-makers of these claims.

Buss (Reference Buss2023) also argues that these ideals can lead to moral dilemmas where legitimate values conflict with one another. Buss defends the possibility of intrapersonal moral dilemmas. Since we have many different ideals, a rational agent can face cases where “apparently irresolvable conflicts among moral requirements really cannot be resolved” (244). It should be noted that the word “pluralism” never appears in her argument. She is skeptical about the need for unity in agency and thinks most people have more than one ideal. But she seems to want to avoid any epistemic claims. She is interested in “normative seemings” and never considers when an ideal “is justified” (234). I am summarizing her argument to adapt it to my goals. I will take her phenomenological claims about “normative seemings” and turn them into claims about ways of knowing.

Buss argues that people often have many ideals. One aims to be a good philosopher and a good parent. These ideals are not completely determinate, but there are some actions that one cannot perform while being committed to the ideal. Because of this, she “wants to consider circumstances in which a person takes herself to have an overriding reason to do something even though she believes this will involve violating a requirement not to behave in this way under these circumstances” (Reference Buss2023: 245). Her key example is someone who has internalized the ideal of being kind but faces a situation where the only way to save a child’s life is by heaping psychological abuse upon her. She concludes the person is obligated to act in this way to save the child’s life. In doing so, she violates a requirement to be kind to children that is required by her ideal.

The skeptic of moral dilemmas may respond that this was not a case of failing to be kind. That answer is attractive. It seems incoherent to say you are required to do something that you are not permitted to do. However, Buss questions “whether we always have a good reason to attribute this incoherence to a normative mistake” (245). She thinks not “because there is more to the substantive identity of each substantive ideal than its identity as an ideal” (246). Kindness as an ideal entails not being cruel, even when it serves the greater good. She claims to reject this is to endorse “an ideal of rational agency according to which being a good enough responder-to-reasons requires manifesting a measure of normative fetishism, which is less than ideal” (246). We want a unity of the self that is thick in the way ideals provide, not a thin unity of simply responding to whatever reasons carry the day. We want to be a good parent or a good philosopher, not a mere good-doer. Normative fetishism leads us to live a life where we are only superficially attached to what matters to us.

Buss grants that this is not a knockdown argument, and the skeptic of moral dilemmas has reasonable ways to avoid it. Again, we do not need to settle this issue to draw an epistemic lesson from it. Once we switch to interpersonal conflict, the most pressing objection to Buss’s argument falls away. Now the conflict is not within a single agent trying to form a unified conception of how to live but between two different people with conflicting moral views. Accepting epistemic value pluralism, we note that internalizing ideals is a matter of learning different values. We can take Buss’s claims about seeing a reason or requirement as a way of coming to know a reason and not a normative seeming.

Return to our general and war journalist. The general internalizes an ideal that makes her apprehend and appreciate reasons distinctly. She sees enemy troop movements not as mere threats to her soldiers, but as particularly dangerous ones, given their weapons and location. She plots her strategy. She sees the facts in a normative significant way due to her training and experience. The war journalist sees enemy troop movement, but he looks to capture the soldiers’ and civilians’ faces as they see it. He looks to where he can best capture the impending battle and scrambles to that position. He sees the facts as presenting different normative significance given his training and experience. You would need a general’s or a war journalist’s knowledge to describe the differences more completely.

After the battle, the general and the war journalist talk. They know the same non-normative facts about a key moment in the battle. The war journalist thinks the general acted wrongly in advancing instead of protecting the civilians. The general reminds the war journalist she knows full well that she risked innocent lives by advancing on the enemy but says it was a risk worth taking. They bitterly disagree. They both reflect on it later. The war journalist knows the general knows more than he does about strategy. The general is not a fool, and she does care about civilians. She had to make a hard choice with deadly consequences. The war journalist still disagrees but admires the general. The general grieves over the dead, but she still believes she is right. A general who takes no chances loses opportunities to end wars. The war journalist has a compassionate heart, and the general knows that it is good, but it is not something she can allow herself to feel. She admires the war journalist but does not want to hear his advice on battle tactics.

The war journalist and the general do not have to live a single life, so they do not need a single answer to the question of how to fight the battle. There is no threat of alienation as in Buss’s case. They do not need to even agree on what is right. And this disagreement is not higher-order evidence that neither is justified in their belief. In fact, at a later time, they may acknowledge to each other that neither one believes the other is vicious or a fool. They might admit how they each admire the other because of their commitment to their respective roles. This seems to me to be higher-order evidence that each one is not a moral evil demon. A person who has internalized a different ideal that allows them to see the facts as normatively significant in another way judges that another is a competent judge. In the next section, I will argue that we should all see it that way.

Note how my argument differs from more familiar arguments about moral disagreement. Skeptics usually point to disagreement over first-order beliefs as higher-order evidence against the justification of first-order beliefs. My argument does not concern any first-order belief. Rather, an admirable disagreement can serve as higher-order evidence about one’s capacity for making normative judgments. It is evidence that vindicates one’s self-conception as a competent normative judge. The admiration is not aimed at any one belief but at a person’s capacity for normative judgment based on the internalization of a different ideal. It provides evidence about normative competence, but normative competence is not normative perfection. Since my argument is focused on our capacity for normative judgment, it is somewhat like Street (Reference Street2006). She argues that what we know about evolution gives us reason to think our capacity to make normative judgments does not track normative truth. I will not address her argument here. But for my part, in questions about value, I believe I can learn more from those I admire and disagree with than I can from evolution. My aims are also more modest than Street’s. She thinks her argument undermines all realist-sounding moral knowledge. I am just claiming there is a way to acquire positive higher-order evidence, not that this sort of evidence outweighs all other evidence.

4. The authority of admiration

Recall our worry is that our moral reasoning might be compromised. Others have been moral evil demons. How do we know we are not like that? This is not a worry we should normally entertain. That would be a lack of conviction. It is not clear when it is called for, but we want to avoid dogmatism while ruling out some viewpoints. My strategy is to find higher-order evidence by examining admirable disagreement. However, disagreeing admirably is just one possible attitude we have towards others. What gives it epistemic authority? This section answers that question.

We should first note the social nature of much moral knowledge. The social nature of the development of our normative capacities means that we need to rely on our communities to shape these capacities. McGrath (Reference McGrath2019) makes this point well. She claims that we can attain moral knowledge by moral inheritance. She notes in most cases by the time anyone achieves the ability to engage in moral reasoning, “she will have absorbed some such knowledge from her social environment” (59). Just as with historical knowledge, most of us have some moral knowledge “in virtue of having acquired the views from reliable sources” (60). Now, you might think that at some point the person must engage in some deliberations to make the knowledge her own. However, we do not think that historical knowledge requires one to personally research each issue. McGrath argues there is no reason to treat moral knowledge differently. She claims this is possible because not all evidence is transparent. That is, there can be evidence that a claim is true, but not evidence that provides insight into why it is true. McGrath claims that “opaque evidence for a proposition suggests that the proposition is true without providing any indication as to why it is or would be true” (89). In these cases, we would not be able to provide reasons why the proposition is true. We can lack moral understanding while having moral knowledge.

Moral knowledge is even more dependent on social interaction than historical knowledge or other inherited knowledge. Consider this argument from Gibbard (Reference Gibbard1990: 179–181). Gibbard claims we must give some weight to other agents’ normative judgments to have any faith in our own. That is, if we do not grant some fundamental authority to others, then we will fall into complete normative skepticism. Gibbard notes that to trust our judgments, we must trust the past influences of others on ourselves. Each one of us has been shaped in countless ways in our judgments by the judgments of others. We judge as we do, in part, because of the causal influence of others on our ability to make judgments. To trust ourselves, we must have some trust in these past influences. If we do not trust ourselves, then we cannot make any normative judgments. Hence, skepticism looms if we fail to trust others. Since, in most cases, there is no explanation of why past influences deserve weight and future ones do not, we must accord authority to some of the judgments of others we interact with if we are to make any judgments at all. Given the way that others’ judgments shape our capacity to judge, we cannot simply ignore them. Trusting others is essential to moral knowledge.

I understand Gibbard’s argument to be empirical in that it rests on how we, as natural creatures, develop our normative capacity via social interaction. It is also transcendental in that if you trust yourself, then you must trust others. Our conclusion is conditional. If we take ourselves to be justified, then we must take others to be justified as well. This is not in any way a cheat. Moral skepticism and nihilism were always off the table. The paper is aimed at anyone who thinks there are ways to be better or worse in moral thinking. Vindicating morality is beyond the scope of this paper.

All of us have good reason to accept that there is an important social element to moral knowledge. If we were born into cruel and vicious societies, we would have far less moral knowledge. However, when we are born into just and caring societies, we will have more moral knowledge. Trusting others, at least in some cases, is unavoidable for developing our capacities effectively.

How can this help with the problem that our moral judgments might be corrupted? Consider the facts that are before us now:

1. 1. There are a plurality of ways of coming to know values. (Section 2)

2. 2. Each way of knowing values gives the knower privileged epistemic access to those values. (Section 3)

3. 3. On pain of skepticism, we must grant epistemic authority to some others. (Section 4)

Given these facts, we can see how disagreement can be a useful tool. We know only a limited set of values. Others know some other set of values. Moral knowledge requires an element of trust. If we extend this trust to others we disagree with, we increase our access to knowledge of value. This will allow us to compare our moral reasoning against a different standard. What epistemic value pluralism allows is for us to avoid vicious circularity. If our only way to assess our judgments is by appealing to others who agree with us, there is no measure of independence in our assessment. By recognizing pluralism, we can appeal to others who use different values and thick concepts to assess our judgments.

Accepting all points of view will not improve our epistemic position. Arguing with fools or villains will not do us any good. Perhaps consulting those we disagree with will undermine our knowledge. McGrath (Reference McGrath2019) seems to think this. On moral cases, she thinks we could only grant authority to “people whose sense of right and wrong overlap strongly with our own, as measured by agreement with respect to specific moral issues” (100). It is unclear how much agreement is required for our views to “overlap strongly.”Footnote ¹⁰ I advocate only for as much convergence as Parfit’s CC, which I read as just requiring agreement on claims like “torturing children merely for fun is wrong.” That leaves plenty of room for disagreement. We should discount those we judge to be fools or evil. However, it would be irrational to ignore those we admire but disagree with. When we admire MJ, we judge that he is a competent judge. We are impressed by him. If we dismiss our admiration, we need some account of why we are justified in doing so. We look at more cases in the next section. As it stands, the admiration of agents who hold different views or embrace different ideals is an excellent source of higher-order justification.Footnote ¹¹

5. Weighing the evidence

Now, one might think this does not amount to much. It is some measure of evidence that might be swamped by other sorts of evidence. In some cases, it likely will be. In other cases, it will be decisive evidence that one is a competent normative judge and not a moral evil demon. Let us move through some examples to illustrate how we might gather significant amounts of evidence. I introduced my idea with the example of you and MJ. I do not want to stipulate facts about you, so I will switch to Philosophy Professor (PP). MJ and PP disagree over whether it is permissible to demean others to advance their goals. MJ has internalized an ideal of competitive excellence where winning is the ultimate standard. Since he is committed to the ideal, he does not have any thoughts about whether this ideal is permissible when he acts. He is committed to winning and does not have a normative fetish. PP has internalized the ideal of being a philosopher and following wherever the argument leads. Since she is committed to the ideal, she does not have any thoughts about whether this ideal is a permissible ideal when he acts.Footnote ¹² She is also committed to kindness. PP follows where the argument goes but does not express those ideas where she thinks it will demean others. If the argument told her she ought to demean others, then that led her to reevaluate the argument. It might convince her that the previous modus ponens is actually a modus tollens. Our two characters disagree over whether to demean others in the course of pursuing their goals. Now the two come to know each other reasonably well. In doing so, they admire each other while maintaining the disagreement. What is going on here is that each is judging the other as a competent normative judge. This is evidence because one cannot simply dismiss the views of what one takes to be a competent normative judge. A person who employs different thick concepts finds that the other admirably applies their own thick concepts.

Of course, evidence comes in degrees. The better the individuals know each other and the more they admire each other, the stronger the evidence they acquire. Take our general and war journalists. We could imagine this as the beginning of a love story. I will not try to write that story. However, it seems to me that loving another does not require agreeing with them on all moral issues. Their love might depend on the deep respect they have for each other’s commitment to their ideals. The disagreement may be part of what keeps the two deeply engaged in communicating with one another. Of course, this is not limited to romantic partners. Great friendships sometimes involve serious disagreements that endure because of the admiration that the friends have for one another. I think such cases present very strong higher-order evidence in moral epistemology. The fact that my friends and lovers admire me even when we disagree matters a great deal to me. Epistemically, it should matter to all of us.

The number of people who admirably disagree with us would also increase the evidence we have. The ideal upper limit is when everyone we admire admires us in return. That would be decisive evidence of normative competence. Of course, normative competence is not normative perfection. A person in this state could still be mistaken in some ways, but they could not be a moral evil demon. We have good reason to trust that such a person can evaluate the evidence in moral cases very well.

Our admiration should be constrained by the kind of convergence Parfit expects us to find. Recall that there is agreement on claims like “torturing children merely for fun is wrong.” As noted earlier, there is a real worry that people sometimes admire others whom they should not. It manifests a kind of foolish open-mindedness that lacks moral conviction.Footnote ¹³ Consider the allure of a mysgionist MJ (MMJ). MMJ does not express a commitment to excellence in basketball but a commitment to dominating and degrading women. Sadly, millions of young men and teenage boys do admire social media influencers who express just these sorts of toxic commitments. We need to explain why this is a mistake. The key to note here is that we have an overwhelming amount of first-order evidence that commitment is morally objectionable. I will not specify the source of this evidence because I do not want to commit myself to moral metaphysics. However, any theory that cannot vindicate this claim pays a high theoretical cost. We should also note that any admiration these characters generate is typically constrained to echo chambers that exist to shut out dissenting voices. The lack of disagreement that they require to function is evidence of their epistemic viciousness.

The fact that we do not admire others is a reason to find them less reliable. Positive higher-order evidence only comes from those we admiringly disagree with.Footnote ¹⁴ We need not consider those whom we pity or despise. It seems to me that discounting certain people is a good practice. For many of us, knowing that politician X supports policy Y is enough for us to know that we do not support Y. Many of us do this too quickly. I would encourage all of us to be charitable and listen carefully to those we disagree with. One implication of my view is that considering those we disagree with is a key source of higher-order evidence. This explains why diverse societies are more conducive to the production of moral knowledge. However, some agents are beyond the pale. Again, it is hard to know when to draw the line. However, there is very little to be gained in arguing with fools and villains. Nevertheless, diversity of viewpoints is an important tool for moral knowledge.

Again, we can expect there to be hard cases. The musician who writes beautiful music but engages in self-destructive behavior is a case I find hard to form a view on. I admire Kurt Cobain and Elliot Smith. I hope they will admire me. However, both took their own lives. That seems quite strong evidence against their normative competence. However, I cannot bring myself to call them fools or madmen. This is because I am not quite sure how to balance the good of artistic creativity with the good of emotional well-being. Cobain and Smith got it wrong, but I would not know how to advise them to get it right. Part of the explanation of this fact is that I am not an expert in addiction or mental health. But I suspect this is also partly because we have not yet reached the convergence that Parfit thought we would reach. Suffering for your art is a good thing, but how to manage or mitigate it is a question I do not know the answer to.

Finally, we should note how our cases differ from others in the literature on moral disagreement. First, our kind of disagreement is not a disagreement between epistemic peers who have the same evidence. The parties involved here have access to different sets of evidence. One has internalized one set of values, and the other has internalized a different set of values. Each has privileged access to different kinds of lives. It seems to me a general epistemic principle that one gives some measure of deference to others with different life experiences. If you are an experienced painter and I am a novice, it is natural for me to defer to your judgment about painting. Likewise, if I am an expert poker player and you have just learned the rules, then I have an epistemic advantage in playing poker. The fact that it is a moral disagreement plays no role in this deference. Second, even granting this measure of deference, I am not committed to what is often called Sidgwick’s Principle. It has different formulations, but here is how Wedgwood (Reference Wedgwood, Feldman and Warfield2010) puts it:

If you have a belief about a (first-order) question and then acquire the (higher-order) information that another thinker disagrees with you about that question, you are rationally required to suspend judgment about that (first-order) question unless you have independent grounds for thinking that the other thinker is less reliable about that question than you are yourself (224).

This would require a significant measure of deference to everyone. However, Gibbard’s work shows we must trust some others, but not everyone. Admiration is what we use to discriminate. As we saw above, this is not always an easy matter. This principle is likely specific to moral epistemology. However, given the social component in the development of normative judgment, we have good reason to accept it.

What has been presented here is far from a complete moral epistemology. No claim has been made about the nature of first-order evidence. All that has been presented is one kind of higher-order evidence rooted in a type of disagreement. There may be other sorts of higher-order evidence. However, disagreement, all by itself, does not always count as higher-order evidence against what we disagree about. Moral justification and disagreement cannot only coexist, but admirable disagreement can increase justification. Every plausible account of moral metaphysics should be compatible with this claim about moral epistemology.