What are blame and shame? “Blame” is ambiguous; it refers sometimes to an attitude and sometimes to an action.3 You might blame someone for doing something but have reason to keep this entirely to yourself—in that case, you have the attitude but don’t perform the action. Or you might have sufficient practical reason to blame someone for something even though you wholeheartedly believe they are not blameworthy—in that case, you perform the action without having the attitude. “Shame” is also ambiguous as between an attitude and an action, but in a different way. It is one thing to be ashamed of or to feel shame for something and another to shame someone for something. My concern in this paper is with the attitude of shame, and my first order of business is to distinguish between the attitude of blame and the attitude of shame. So “blame” and “shame” will refer to attitudes in the first instance, although we will later have occasion to consider the actions of blaming and shaming (§6).

I don’t mean to imply anything substantial by speaking of blame and shame as “attitudes”; the point of this is just to indicate the contrast with the actions of blaming and shaming. I intend to stay neutral on the question of whether and to what extent blame is an emotion. I will speak freely of “feeling” shame (and, in a moment, of “feeling” guilt), although in doing so I again don’t mean to imply anything substantial, e.g. that shame (or guilt) is an emotion.

I won’t draw any distinction between feeling shame and being ashamed. Moreover, although I will assume (below) that shame is always “for” something, grammatically we speak just as easily, and often more naturally, of being ashamed to φ (e.g. to be a member of the American Philosophical Association), ashamed of x (e.g. of your big ears), ashamed about x (e.g. about having overslept on your wedding day), or ashamed that p (e.g. that your clothes are lame). I will use these various grammatical forms without implying any substantive distinctions.

I propose to distinguish blame and shame in terms of their fittingness conditions. “Fittingness,” in this context, is equivalent to accuracy: an attitude is fitting if and only if it accurately represents its object. Belief has a fittingness condition, namely, truth: your belief that p is fitting if and only if it is true that p, and unfitting otherwise. But emotions also can have fittingness conditions. The fittingness condition for fear, for example, is (roughly) dangerousness: your fear of x is fitting if and only if x is dangerous, and unfitting otherwise. Note well that the present concept of a fitting attitude is distinct from the concept of a reasonable, rational, or justified attitude, distinct from the concept of an understandable attitude or an attitude that it makes sense for you to have, distinct from the concept of an attitude that manifests a properly functioning cognitive or emotional system, and in general distinct from the concept of an attitude that you ought to have.4

Here’s my proposal:

As this formulation suggests, shame, but not blame, is essentially self-directed. It is, as Gabrielle Taylor (1985) puts it, a species of “self-assessment.” When you feel shame you represent your φing as shameful. In the same way, when you blame someone you represent their φing as blameworthy. However, I do not mean to suggest that such assessment is sufficient for blame or shame, and I want to leave open the possibility that blame and shame require, in addition to such assessment, some affective or emotional response.

The claim that shame is essentially self-directed is consistent with the fact that you can feel ashamed of something someone else did. For example, you might be ashamed that your ancestors enslaved people. On my view, this involves you representing your having ancestors who enslaved people as shameful. Your shame is a response to what this fact says about you. From the perspective of your shame, so to speak, it is shameful to be descended from slaveholders. Shame requires some such connection—some association or relationship—to the self. Of course, it is consistent with all this that your ancestor’s conduct was, itself, shameful. However, your shame represents your having ancestors who enslaved people as shameful, rather than representing their enslaving of people as shameful.5^,6

You can, of course, blame yourself for something—and that is guilt. Thus, we can add:

Guilt, in other words, is self-blame.7 When you feel guilt you assess your φing as blameworthy.8

Blame and shame, on the present view, are always for something. If you blame someone, there must be something for which you blame them; if you feel shame, then there must be something for which you feel shame. You cannot simply blame someone, but not for anything, and you cannot simply feel shame, but not for anything. If you feel shame, there must be some φ such that you represent your φing as shameful. You might, for example, feel shame for eating an entire bag of cookies or for feeling amused during a funeral. But shame need not be for some something so specific and concrete; you might feel shame for something general and abstract. You might, for example, feel shame for being a bad person or for being a loser. But you cannot merely feel shame, where there is nothing for which you feel shame. Now, you might feel shame for φing without knowing what that φ is. You may find yourself wondering why you feel shame. But there is always something for which you feel shame. (And the same, mutatis mutandis, for guilt.)

We should not assume that φing, in the present formulation, is an action or anything that we would ordinarily describe as something you do. It is possible to blame someone and to feel shame for being a member of some group (e.g. for being Catholic) or for having some property (e.g. for being cowardly). (That is not yet to claim that blame and shame in such cases is fitting—perhaps being a member of a group or having some property is never blameworthy or shameful.)

It follows from the present view that it is unfitting to blame someone for doing something that they did not do and unfitting to feel shame for something you did not do.9 You cannot fittingly blame me for not doing the dishes if I actually did the dishes; you cannot fittingly be ashamed that your ancestors enslaved people if they actually did no such thing. And the same, mutatis mutandis, when it comes to blame and shame for things other than actions.

At least one important difference between blame and shame corresponds to the difference between blameworthiness and shamefulness (§1). So, what are blameworthiness and shamefulness, and how do they differ? I am not going to propose accounts of blameworthiness and shamefulness here. Instead, in this section and the following, I will articulate two central differences between blameworthiness and shamefulness, which will be sufficient to characterize them for our present purpose—namely, distinguishing between blame and shame.

The first central difference between blameworthiness and shamefulness is that blameworthiness is essentially deontic while shamefulness is essentially axiological.10 More precisely, I am going to assume that the blameworthy is a species of wrongdoing and that the shameful is a species of legitimate expectation violation.11 Your φing is blameworthy only if (and partially in virtue of the fact that) it is wrong for you to φ; and your φing is shameful only if (and partially in virtue of the fact that) you are legitimately expected not to φ. “Expectation” in this formulation refers to a normative expectation rather than a predictive expectation: I (normatively) expect my flaky friend to pick me up from the airport, because he promised to pick me up, even as I (predictively) expect that he will not show up, because of his lousy track record. (When I say that the shameful is a species of legitimate expectation violation, I do not mean to suggest that all legitimate expectation violations are shameful.12 In the same way, when I say that the blameworthy is a species of wrongdoing, I do not mean to suggest that all wrongdoing is blameworthy.)

A consequence of the present formulation is that the blameworthy and the shameful are not exclusive. Paradigm cases of wrongdoing involve the violation of legitimate expectations, and are cases of wrongdoing at least in part because they are cases of legitimate expectation violation. A burglar who sneaks into your house violates your legitimate expectation to be left alone, and that, at least in part, is why their conduct is wrong. For this reason, the same conduct will often be both blameworthy and shameful. When Joseph McCarthy accused Fred Fisher, assistant to Joseph Welch during the 1954 Army-McCarthy hearings, of being a Communist, he did something wrong. But he also did something shameful—as Welch famously pointed out when he asked after McCarthy’s sense of decency. As Flanagan (2021) argues, “[s]hame about breaking promises, or about being rude to service workers, or for being a malicious gossip … are all fine, because one ought to feel ashamed of these things” (134). These are examples of shameful conduct, but they are at the same time examples of wrongdoing. Indeed, I want to leave open the possibility that all cases of wrongdoing are cases of legitimate expectation violation. R. J. Wallace (1994: ch. 2) argues that reactive emotions, including blame, essentially involve holding someone to an expectation, and elsewhere (2013) defends a “relational conception of morality,” on which it is a conceptual truth about wrongdoing that wrongdoing provides other people with a reason to blame the wrongdoer (234). This suggests that blameworthiness, no less than shamefulness, is essentially connected to expectations.

What then distinguishes the blameworthy and the shameful? First, there is a normative difference: blameworthiness requires the violation of moral expectations, but shamefulness can be grounded in the violation of non-moral expectations. As Williams (1993/2008) notes, “we … can be mortified or disgraced by a failure in prowess or cunning as by a failure of generosity or loyalty” (92). Imagine that you are a wealthy banker who also manages your family finances. However, in a surprising display of naiveté, you are taken in by a phishing scam and relieved of several thousand dollars. The harm to your family is trivial (you are extremely wealthy) and you violated no professional obligation (it was not company money that you lost). What you did was not wrong. But it was absurdly stupid, and you feel ashamed for having been fooled so easily. Is your shame fitting? I think we should grant that it is, and thus that your conduct—your giving the money to the scammers—was shameful. You made, as you might explain it, a shameful failure of judgment, given your professional expertise, the simplicity of the scam, and the fact that you lost money your family had entrusted to you.

I said that the shameful is a species of legitimate expectation violation. In the present case, whose expectations are violated? Your family expected you to manage their money—that was one of your familial responsibilities—and your shame will be acute when you have to confess your blunder to them. But your shame would increase were your colleagues and others who admire you professionally to find out about what happened. They also expected you not to be taken in by a crude financial scam, not because they were relying on you (as your family was), but because of their professional esteem for you, which the discovery of your shameful conduct would tend to undermine.

The fact that shamefulness can be grounded in the violation of non-moral expectations explains why shame is associated with seeing yourself as inadequate or defective. There is something right about the ideas that “[g]uilt … normally involves a consciousness of having done wrong, and shame a consciousness of some personal inadequacy” (Gibbard 1990: 137), that shame involves a perceived “inadequacy” or “failing” (Williams 1993/2008: 221), that “[w]e feel ashamed in situations in which we are not up to scratch” (Blackburn 1998: 17), and that “shame focuses on defect or imperfection” (Nussbaum 2004: 207). On my view, what these observations get right is the fact that the expectations violated in cases of the shameful need not be moral expectations—shame can involve merely seeing yourself as inadequate or defective, without seeing yourself as having done anything morally wrong. However, on my view, shame does not require seeing yourself as inadequate or defective: shameful wrongdoing need not evoke any such feelings.

Second, there is an ontological difference between the blameworthy and the shameful: only actions can be blameworthy, but things other than actions can be shameful. The blameworthy is a species of wrongdoing, but only actions can be wrong, and so only actions can be blameworthy.13 By contrast, because the shameful is a species of legitimate expectation violation, things other than actions can be shameful. You can do wrong only by doing something—in the sense of “doing” on which doing something entails performing an action—but you can also violate legitimate expectations by being a certain way. Imagine that you are a prominent historian and have been invited to give a lecture on your area of professional expertise. When it comes time to deliver the lecture, however, you realize that you have forgotten a number of basic facts about the historical events you are meant to discuss. Naturally, you might be ashamed of this. However, what you are ashamed of, in this case, is not an action, but a state—the state of being ignorant of the relevant historical facts. Granted, you might be retrospectively ashamed of having bungled the discussion or prospectively ashamed of the fact that you are about to bungle it, but you are also, I would say ultimately, ashamed of not knowing what you were expected to know. Could it be argued that you are really ashamed of some action, in this case, perhaps of not having memorized the relevant historical facts better or of not having reviewed them before the lecture? We can imagine the case like that, but we can also imagine a case in which your ignorance is not your fault—there is nothing you did—no action you performed or failed to perform—about which you feel ashamed (or, for that matter, guilty)—but simply a matter of non-culpable forgetting. Now, is your shame fitting? It seems to me that we should allow that it is. Given your professional stature, it is shameful not to know the basic facts in your field. Your audience expected you to speak from knowledge, and you violated that expectation. This illustrates, again, the connection between shame and expectations.

The fact that things other than actions can be shameful, on my view, is the truth in the idea that guilt is concerned with actions and shame is concerned with the self.14 Taylor (1985) writes that “a person feeling shame becomes conscious not merely of what he is doing, but becomes conscious also of his self” (59) and that such a person “feels herself degraded, not the sort of person she believed, assumed, or hoped she was or anyway should be” (64; see also 89–92), Sandra Bartky (1990) that “[s]hame is a distressed apprehension of the self as inadequate or diminished” (86), Bernard Williams (1993/2008) that “[s]hame looks to what I am” (93; see also 89–90), Martha Nussbaum (2004) that whereas shame focuses “on some aspect of the very being of the person who feels it, guilt focuses on an action” (207; see also 212), and Owen Flanagan (2021) that “[w]hereas guilt typically focuses on acts, deeds, or doings[,] shame typically focuses on aspects of the person that are personal or characterological weaknesses” (135; see also 149, 161). However, I see no reason to say that actions cannot be shameful15, nor that when we are ashamed of doing something we are in some sense really ashamed of what our so doing says about who we are. An athlete or a musician could be ashamed of a disastrously bad performance—a shameful performance, in their view—without thinking it was anything more than a bad night. When you are ashamed of something you represent it as a violation of legitimate expectations, but actions, along with traits and other properties of persons, can violate expectations. For the same reason, I see no reason to say that only whole persons—rather than aspects of them—can be shameful.16 Guilt is concerned with actions inasmuch as blameworthiness is a species of wrongdoing, and shame is concerned with the self inasmuch as things other than actions, including traits and other properties of persons, can be shameful. However, it seems fair to say that feeling shame for φing typically or in paradigm cases involves an awareness of how your φing reflects on who you are.

That the shameful is a species of expectation violation explains why shame has an essential connection to the idea of being seen by others.17 Taylor (1985) writes that “in feeling shame the actor thinks of himself as having become an object of detached observation” (60), and that for this reason the concept of shame “centrally relies on the concept of another” (67), and Williams (1993/2008) that “[t]he basic experience connected with shame is that of being seen, inappropriately, by the wrong people, in the wrong condition” (78; see also 220–222). The fact that shame has some such connection to the idea of being seen by others explains why it paradigmatically involves a desire to hide, disappear, withdraw, or conceal yourself.18 I have characterized the shameful as a species of expectation violation, and the expectations violated by the shameful will at least typically and in paradigm cases be those of other people. This explains why shame so often involves a feeling of being seen, a feeling of exposure or publicity—because the shameful at least typically and in paradigm cases concerns what other people expect of you. Shame is for this reason a “social emotion,” connected to the possibility of exclusion or ostracism.19

When you feel fitting shame for φing, and thus when your φing is shameful (§1), we can get a grip on whose expectations are violated by thinking about whose awareness of your φing would exacerbate, intensify, or make acute your shame. Consider Flanagan’s example of malicious gossip. If you are ashamed of having engaged in malicious gossip, whose awareness of your having engaged in malicious gossip would exacerbate, intensify, or make acute your shame? Surely, in the first instance, the people about whom you gossiped. Your shame would be enhanced by their discovery of what you had said or heard about them. You violated their expectation that you not maliciously gossip about them. But presumably the awareness of other people—more discreet members of your family or peer group or community who were not involved—would also enhance your shame. You violated their more generic expectation that you not engage in malicious gossip.

I am going to assume that some things are shameful. Guilt and shame are popularly perceived to be harmful and unhealthy emotions, symptoms of pernicious cultures of blaming and shaming. Many of the things that people are often made to feel shame about are not shameful. Shame—prescribed by shaming and social stigma—plays an important role in unjust social hierarchies and systems of oppression and domination, as its critics have emphasized.20 However, the fact that these hierarchies and systems exploit our capacity for shame does not mean that shame is never fitting.21 Those same hierarchies and systems exploit our capacities for love and belief, to name just two others, but this does not mean that love and belief are never fitting. It is important to keep in mind that when you feel shame for φing you represent your φing as shameful (§1). Shame is not merely an unpleasant feeling or state of irritation occasioned by other people expressing contempt or disdain for you. Granted, you might feel shame because you have internalized certain alien or external values, by contrast with values that you autonomously or reflectively endorse. (The distinction between those two kinds of values is somewhat obscure.) Either way, the values that ground your shame are your values.22 However, you might think that there is always something wrong with internalizing alien or external values. Nevertheless, that does not speak against the possibility of fitting shame. It merely speaks to a problem with shame grounded in alien or external values.

Defenders of the moral value of shame, when explaining its nature, often give examples of people ashamed of things that do not really seem, upon reflection, to be shameful: a man carrying a stack of packages who stumbles and knocks off his own hat (Williams 1993/2008: 221), someone who merely cannot play the piano very well (Blackburn 1998: 18), a painter’s model who becomes aware that the painter sees her as an object of sexual interest (Taylor 1985: 61). These things seem embarrassing, perhaps, but hardly shameful.23 I have given two examples of things that are shameful, above, and I will give more in the following section. It is on the basis of examples like these that I assume that some things are shameful.

The blameworthy is a species of wrongdoing and the shameful is a species of legitimate expectation violation (§2). However, there is a second central difference between blameworthiness and shamefulness: blameworthiness, but not shamefulness, entails responsibility. If your φing is blameworthy, then you are responsible for φing.24 Note well, however, that this entailment is neutral on the matter of explanatory priority. It is consistent with the view that you are responsible for φing because you are blameworthy for φing—and, more broadly, with the idea that responsibility is explained by the propriety of blame.25 But it is also consistent with the view that you are blameworthy for φing because you are responsible for φing—and, more broadly, with the idea that the propriety of blame is explained by responsibility.

Articulating the sense of “responsible” on which blameworthiness entails responsibility is one of the central tasks for a theory of moral responsibility. This is why, as I mentioned at the outset, defenders of blaming people for their beliefs have appealed to various senses in which we are responsible for our beliefs. But there is one sense of “responsible” that we can immediately set aside. There is an important sense of “responsible” on which being responsible for doing something means that it is your duty, job, or obligation to do it—in other words, that it is your responsibility.26 In my department, the first-year grad students are responsible for setting up the reception after a colloquium talk. But this is not the sense of “responsible” on which blameworthiness entails responsibility. In wondering whether someone is responsible for doing something wrong, we are not wondering whether they have an obligation to do what they did!

By contrast, shamefulness does not entail responsibility.27 Your φing can be shameful even if you are not responsible for φing. On the present account, this is explained by the fact that you can fail to meet others’ legitimate expectations without being responsible for so failing, although we need not assume that account to appreciate the fact that shamefulness does not entail responsibility. Consider three kinds of cases of shamefulness without responsibility.

First, inherited personality or character traits, such as selfishness or narcissism, can be shameful when they exceed what is reasonable or normal, even when you are not blameworthy for having them. In such cases it is fitting to feel shame. Imagine that you find out that your children have covid-19 and the first thought you have is that their quarantine is going to interfere with your plans to watch Wimbledon. And imagine that this selfish thought fits a pattern of selfish thinking, which evidentially manifests an abnormal degree of concern for your own interests as against those of others. You are ashamed to be so selfish, and you are right to be ashamed, which is just to say your selfishness is shameful. But it is easy to imagine that you are not blameworthy for being as selfish as you are. It is easy to imagine that there is simply nothing you can do about it, and equally easy to imagine that you have exhausted yourself trying every reasonable strategy to mitigate your selfishness, without any change. Nevertheless, your selfishness is shameful.

Second, involuntary behavioral responses, such as flinching or averting your eyes, can be shameful when they manifest implicit biases that sustain unjust social hierarchies and systems of oppression and domination, even when they are not blameworthy. Imagine that you are an elderly man, raised in a conservative midwestern family around the middle of the 20^th century, but eventually converted to enlightened liberalism. Nevertheless, public displays of romantic affection between men still make you uncomfortable. You suffer not from any anti-gay ideology nor from anything that would normally be called “hatred” of gay people, but from a residual form of what used to be called “homophobia.” Suppose now that you are relaxing on the beach when the two men under the next umbrella start smooching. You inadvertently flinch and turn away, before realizing how ridiculous and awful you are being and hoping that the couple didn’t see. You are ashamed of flinching and turning away, and you are right to be ashamed, which is just to say that your flinching and turning away is shameful. But it is easy to imagine that you are not blameworthy for flinching and turning away. It is easy to imagine that these were involuntary behavioral responses, that neither they nor your underlying “homophobia” can be controlled or changed, and that you simply cannot rid yourself of your implicit bias.28 Nevertheless, these responses are shameful.

Third, group membership, such as being a citizen of a country or being a member of a family, can be shameful in virtue of the conduct of other members of the group, even when you are not responsible for their conduct. Imagine that you are a member of the Australian delegation at an important international summit on climate change. After a long flight, you arrive at the hotel and quickly retire to your room to get some sleep before the summit begins in the morning. However, the other members of your delegation decide to have a nightcap in the hotel bar. When you wake up, you discover that your compatriots have humiliated themselves by going on a sloppy and disgraceful spree, partying all night, vandalizing the hotel, singing offensive drinking songs, and passing out in the lobby. You are ashamed to be a member of their delegation, and you are right to be ashamed, which is just to say that being a member of their delegation is shameful. But it is easy to imagine that you are not blameworthy for being a member of their delegation. Granted, you might think that your colleagues’ conduct was so awful that it requires you to resign your commission, thereby ending your membership of their delegation, and in that sense you might be blameworthy for remaining a member. However, I think what you are ashamed of, in the present case, is not merely remaining a member of their delegation, but the fact that you were a member to begin with—these are your people, who you arrived with, with whom you belong, and resigning, even if it is required, cannot change that. And it is easy to imagine that you are in no way responsible for what happened—it is not that you knew this might happen and could have prevented it, or that this kind of thing is typical of the Department of Foreign Affairs and Trade and so you should never have applied for the job in the first place. Nevertheless, it is shameful to be a member the delegation that behaved so badly.

My thesis is that it is sometimes fitting to feel shame for believing something. Given what I have said about shame (§1), it will suffice to establish this thesis if we can establish that it is sometimes shameful to believe something. I have motivated the assumption that some things are shameful (§2) and argued that shamefulness does not entail responsibility (§3), so neither general doubts about shamefulness nor concerns about whether we are responsible for our beliefs stand in the way of the idea that beliefs can be shameful. Let us see if we can construct a compelling case of a shameful belief.

It will help to begin by considering the ways in which beliefs can be embarrassing.29 First, beliefs are sometimes embarrassing because of their content. In an episode of the sitcom Seinfeld, the hapless George Constanza reveals that he thinks the umbrella stand at his coffee shop is for communal umbrellas that customers are meant to take. It is a silly and ridiculous thing to believe, and when George realizes the error, he is embarrassed, if only slightly. People sometimes share online the most embarrassing thing they believed as a child, e.g. that breaking a graham cracker into quarters yields more cracker than eating it whole.30 Second, beliefs are sometimes embarrassing because they manifest our intellectual weaknesses. It is embarrassing to believe something when your so believing reveals or suggests that you are foolish, stupid, or gullible. It is embarrassing if people find out that you were deceived by an April Fool’s joke, a fake news meme, or an obvious phishing scam. At least when you reach a certain age, it is embarrassing to believe in Santa Claus or the Tooth Fairy. Third, beliefs are sometimes embarrassing because they are also held by people with whom it is undesirable to be associated. Imagine that you believe, just on the basis of your own speculations, that the condensation trails left behind in the sky by jet aircraft consist of noxious chemicals produced by airplane engines. (They are actually just clouds made of water vapor.) Unbeknownst to you, the belief that condensation trails are composed of noxious chemicals happens to be part of the influential and outrageous “chemtrail” conspiracy theory. Your belief is embarrassing, in this case, because it represents an association between you and those who embrace that conspiracy theory, which suggests broader similarities or affinities between you and them. Their craziness and irrationality now seem to be yours as well.

These three ways in which beliefs can be embarrassing correspond, I propose, to three ways in which beliefs can be shameful. Here’s the kind of case I have in mind. Consider the beliefs constitutive of what is known as “racial resentment” (or “symbolic racism”), such as the belief that White Americans are subject to more discrimination than non-White Americans. Imagine that you believe this.31 Your belief is shameful; it is shameful for you to believe this. There are at least three reasons why this is so.

First, your belief is shameful because of its content: that White Americans are subject to more discrimination than non-White Americans is an absurd thing to believe. It’s silly, dumb, moronic, ridiculous, crazy, idiotic.32 It is a preposterous reversal of the truth, not just a common mistake, but what used to be called a “howler.”

Second, your belief is shameful because it manifests your intellectual weaknesses—indeed, your worst intellectual tendencies, your most vicious dispositions. There are a few compatible possibilities to consider here. Your belief might manifest intellectual laziness, e.g. if you attended to an unrepresentative sample in forming it. Myopically focusing all your attention on the plight of relatively disadvantaged White Americans or on the particular disadvantages White Americans are subject to may have made it look like White Americans are subject to a lot of discrimination. Your belief might manifest gullibility, e.g. if you formed it under the influence of some media narrative or piece of propaganda. Uncritical consumption of a poorly curated diet of journalism may have given you the distinct impression that the greatest crisis in American life is an epidemic of White professors getting canceled for microaggressions and political incorrectness. Your belief might manifest wishful thinking or rationalization, if you happen to be a White American, e.g. if the reason you believe that White Americans are subject to more discrimination than non-White Americans is that you very much do not want it to be the case that your own success and comfort depends crucially on race-based discrimination against other people. The sheer inconvenience of believing otherwise may have made it appear compelling that White Americans are subject to more discrimination than non-White Americans. Finally, your belief might manifest the specific forms of irrationality constitutive of the epistemology of ignorance that sustains the global system of White supremacy.33 Indeed, your belief seems most shameful when seen in this light: if your laziness and gullibility were not merely intellectually vicious, but vicious with a purpose or function, then you are revealed not merely to be a fool, but a “useful idiot.”

Third, your belief is shameful because you share it with the worst of the worst: White supremacists, neo-Nazis, terrorist militias, you name it—on this, you are in agreement with them, and thus, in that respect at least, you are one of them. Even if you have nothing else in common, this salient similarity suggests that you are more alike than you seem, and their noxious ideologies and vicious hatred now seem to reflect badly on you.

I said that the shameful is a species of legitimate expectation violation (§2). In cases of shameful belief, whose expectations are violated? In the case of the professor mistaken for a waiter, it seems fair to say that the professor himself expected not to be mistaken for a waiter, at least not on the basis of such evidence as you possessed. We might say something similar about the case of racial resentment, e.g. that non-White Americans expected not to have their situation misunderstood in such a way. However, it seems to me that shameful beliefs, at least in general, violate more generic social expectations, expectations that we all have of each other.34 We expect each other to form and sustain our beliefs with a suitable degree of care and conscientiousness and to inquire in a manner appropriate to the question we are asking, where suitability and appropriateness are determined, perhaps among other things, by the situation and the subject matter.

However, I do not mean to suggest that unjustified or irrational beliefs are always shameful. Even if it were the case that you are always legitimately expected to form only justified and rational beliefs—which seems dubious in cases like that of casually counting the number of motes of dust on your desk—it hardly seems shameful to irrationally and without justification believe that there are 32 motes of dust on your desk, on the basis of a hasty and sloppy count.

Because shamefulness does not entail responsibility, that your belief is shameful does not entail that you are responsible for believing that White Americans are subject to more discrimination than non-White Americans. Now, to be clear, even if your believing as you do is not blameworthy, other people might be to blame (e.g. political leaders and journalists who disseminate fake news and propaganda), and impersonal structural, institutional, and environmental factors might deserve scrutiny and critical intervention (e.g. social media algorithms, echo chambers and media bubbles, information cascades, political sectarianism, racial segregation, geographic sorting, etc.) for their roles in causing you to believe as you do. That your belief is shameful is consistent with the sentiment expressed by Bob Dylan when he sang, of a White racist, “it ain’t him to blame; he’s only a pawn in their game.”35 One advantage of focusing our attention on shame for beliefs, as opposed to blame for beliefs, is that it allows us to bracket the question of whether the believer is responsible for believing as they do, and attend to the various causes of their belief, both those involving the agency of individuals as well as those involving impersonal structures of social and political power. Granted, the present approach to racial resentment does involve attending to the beliefs of individual people. This is consistent with the fact that racial resentment is primarily objectionable because it is an essential part of the aforementioned epistemology of ignorance. “Racism,” on this understanding, is a political system, rather than an individual pathology of hatred, and “racist” can only derivatively be applied to policies, practices, actions, people, and attitudes, in so far as they sustain that political system. By attending to the beliefs of individual people, as part of our approach to racial resentment, we should not forget the essentially systemic nature of racism.

I have argued that beliefs can be shameful, and, therefore, that it is sometimes fitting to feel shame for believing something (§4). However, with the case of racial resentment in mind, you might at this point worry: so what? Even if we were to conclude that racial resentment, in general, is shameful, what will that do to combat racism? Of course, not every truth about racism is useful for combatting racism. But the alleged fact that racial resentment is shameful seems like cold comfort without some suggestion of how we might proceed to address the socio-political problem it presents. Does this conclusion have any ameliorative value?

Both guilt and shame plausibly have proleptic functions: they are not merely backward-looking evaluations of what you have done, of who you have been, or of what you have been like, but at the same time forward-looking motivators of future conduct. Both guilt and shame are intrinsically unpleasant: when you feel guilt or shame, you have a reason to do what you can to rid yourself of said attitude. Moreover, both guilt and shame are species of negative self-assessment, the content of which will provide practical reasons. Suppose you feel guilt or shame for doing something that you are presently doing. If what you are doing is blameworthy or shameful, you have a reason to stop doing it. If your failure to donate to a local charity is blameworthy or shameful, you have a reason to start donating. Suppose you feel guilt or shame for doing something in the past. If what you did is of a blameworthy or shameful type, then you have a reason not to do things of that type in the future. If eating all your kids’ Halloween candy was blameworthy or shameful, you have a reason not to do the same next year. Finally, suppose you feel guilt or shame for being a member of some group or for having some property. If your being a member of that group or having that property is blameworthy or shameful, then you have a reason to do what you can to end your membership of that group or make it the case that you no longer have that property. If your membership of the Scottish National Party is blameworthy or shameful, you have a reason to quit; if your inability to ride a bicycle is blameworthy or shameful, you have a reason to learn how to ride one.

Moreover, prospective guilt and shame plausibly have proleptic functions. That you would feel guilt or shame for doing something is a reason to refrain from doing it—as is the closely related fact that doing so would be blameworthy or shameful. And that you would feel guilt or shame for becoming a member of some group or for acquiring some property is a reason to do what you can to avoid those outcomes—as is the closely related fact that those outcomes would be blameworthy or shameful for you. Indeed, merely (epistemically) possible guilt and shame need to be taken into account. That you might feel guilt or shame were you to do something is a reason not to do it—as is the closely related fact that doing so might be blameworthy or shameful or might result in blameworthy or shameful outcomes for you. If you are disposed to guilt and shame, the blameworthy and the shameful may well place limits on your practical options: at least in the first instance, you may well only consider actions that you are confident enough are both (i) neither blameworthy nor shameful and (ii) would result neither in blameworthy nor in shameful outcomes for you. In this way we are motivated not only be guilt and shame, but by the fear of guilt and shame.

The proleptic function of shame is a theme in the work of those who defend shame as morally valuable.36 Similarly, many defenders of blame—despite other substantial disagreements about the nature of blame—argue that blame has a proleptic function.37 Just as guilt and shame can affect the future conduct of the person who feels guilt or shame, blame can affect the future conduct of the person blamed.38 This is because being blamed by someone can affect how they treat you in three related and compatible ways. First, they may reveal, whether intentionally or not, that they blame you, and being blamed is inherently undesirable. Second, blame may dispose them to various kinds of undesirable treatment, ranging from avoidance and unfriendliness to ostracism and hostility. Third, the person who blames you may choose to actively blame you—that is, they may, given that they have the attitude of blame, choose to perform the action of blaming (§1). How all this can affect your future conduct—how the proleptic mechanism of blame works—is a topic considered by the aforementioned defenders of blame.39

It is initially tempting to think that, just as feeling shame for doing something can cause you not to do so in the future, feeling shame for believing that p can cause you not to believe that p in the future. However, on closer examination this is implausible. It seems that to represent your belief that p as shameful is already to doubt that p. It does not seem coherent to believe that p and at the same time feel shame for believing that p. Someone who comes to feel shame for believing that White Americans are subject to more discrimination than non-White Americans is already well on the way to not believing that White Americans are subject to more discrimination than non-White Americans—it is more accurate to say that they feel shame for having believed that White Americans are subject to more discrimination than non-White Americans. It does not seem possible to stop believing something because so believing is shameful, in the way that it is possible to stop doing something because doing so is shameful.

What explains this? You might think that a belief is shameful only if it is false. Certainly our paradigm case—the belief that White Americans are subject to more discrimination than non-White Americans—is a case of false belief. Can true beliefs be shameful? Perhaps, but I think it would be difficult to come up with an intuitive case of a shameful belief that was not irrational, unjustified, or intellectually vicious in some way. It does not seem coherent, for example, to think that you know full well that p, and yet to feel shame for believing that p. In an alternative reality in which you knew full well that White Americans are subject to more discrimination than non-White Americans, your belief would not plausibly be shameful.

In any event, I think it is implausible that shame can bring about changes in your beliefs. However, it seems to me that fear of shame can bring about changes in your beliefs, which is how shame performs its proleptic function in the case of belief. You can coherently believe that p and at the same time worry that your belief might be shameful. In general, that you are worried that your φing might be shameful gives you a reason to inquire into whether your φing is shameful. In the case of belief, that you are worried that your belief that p might be shameful gives you a reason to inquire into whether your belief that p is shameful. However, such inquiry will inevitably incorporate inquiry into whether your belief that p is true—which is simply inquiry into whether p. And that is exactly the kind of inquiry that can result in your changing your mind about whether p.

Why will inquiry into whether a belief is shameful “inevitably” incorporate inquiry into whether that belief is true? Again, it may be that only false beliefs can be shameful, in which case the question of truth is essential to any inquiry into whether a belief is shameful. However, even if true beliefs can be shameful, it is hard to imagine a case in which the truth of the belief in question is not important and relevant to whether it is shameful. In our paradigm case, for example, the fact that White Americans are not subject to more discrimination than non-White Americans is obviously an important part of the explanation of why your belief is shameful.

Inquiry into whether p can involve at least two things. First, inquiry into whether p can involve reflecting on the evidence you possess relevant to the question of whether p. This can involve a number of things, including attempting to determine what evidence is relevant, attending carefully to said evidence, and attempting to determine what conclusion, if any, that evidence supports. Diligence in this last particular can require the consideration of various interpretations and explanatory hypotheses and attention to possible defeaters, including undercutting defeaters involving your own biases and blind spots. How might this go in the present case of racial resentment? I think you might more or less immediately realize that you lack sufficient evidence to believe that White Americans are subject to more discrimination than non-White Americans—for example, that your merely anecdotal evidence of anti-White discrimination was of the wrong kind to support a generalization, or, for another example, that your evidential basis is misleading because you lack evidence relevant to the extent of race-based discrimination against non-White Americans. Selective attention to our evidence often leads us to make mistaken generalizations, which reflection can correct. For example, if all you think about are various affirmative action programs, it might seem to you that there is something like “discrimination” against White people, and if you ignore all the existing forms of discrimination against non-White people, then it perhaps seems that there is more discrimination against White people than against non-White people. This conclusion is the result of an arbitrary and myopic focus on anti-White “discrimination.” Reflecting on all your evidence by “stepping back” or “zooming out” can correct your mistake.

Second, inquiry into whether p can involve gathering evidence relevant to the question of whether p. There are several possible strategies here, including attempting to acquire more evidence, seeking out additional and multiple kinds of evidence, gathering evidence from additional and multiple sources, and seeking out defeating evidence. Again, doing this well will require sensitivity to alternative positions and a concern with undercutting defeaters. How might this go in the present case of racial resentment? It would be natural for you to address your worries about the shamefulness of your belief by seeking out more or higher quality information, such as empirical research, as opposed to anecdotes or speculations. Sometimes, we simply did not have all that much evidence to begin with, such that we believe things on a relatively thin evidential basis. Other times, prior inquiry has been tainted by some combination of intellectual carelessness, unreliable heuristics, biases, misleading evidence, and personal or systematic manipulation. For example, if you are inclined to click on and read news stories about anti-White discrimination and to ignore news stories about discrimination against non-White people, you may end up with sufficient evidence for believing that White people are subject to more discrimination than non-White people. Further inquiry, if conducted with even a modicum of fairness and caution, will result in your coming to possess a completely different body of evidence, indicating a different conclusion.

Both of these species of inquiry are familiar ways of responding to situations in which we become concerned about the correctness of our beliefs. Asserting something in front of an audience often prompts both reflection on your evidence and further evidence gathering, especially when the audience is skeptical or unconvinced, and even more so when the audience is offended or upset. Adam Smith said that it was “always mortifying not to be believed,”40 and at least this much seems right: public disagreement sometimes causes us to reconsider our beliefs, in the hope of having been right, but with the possibility of having been wrong, and thus with the possibility of having to revise our beliefs. When we are aware that we may have publicly made a mistake, especially an offensive or upsetting mistake, we are motivated to find out whether we indeed did make a mistake. Concern that a belief of yours is shameful can motivate inquiry in exactly the same way.

That shame has a proleptic function goes some way towards addressing our concern about the ameliorative value of the conclusion that racial resentment is shameful. However, you might think it does not get us very far. I have described someone motivated to inquire into whether White Americans are subject to more discrimination than non-White Americans by fear that their belief that White Americans are subject to more discrimination than non-White Americans is shameful. However, you might ask, where is this fear supposed to come from? Why would someone suffering from racial resentment ever worry that their relevant beliefs are shameful?

Where does the fear of shame come from in general? One important source is being shamed by others—here, we obviously mean the action of shaming (§1)—if someone shames you for φing, it is hard not to at least wonder whether your φing might be shameful. To shame someone for φing is, roughly, to communicate to them the shamefulness of their φing with the intention of causing them to feel shame for φing.41 Sometimes, we invoke shame explicitly: “You ought to be ashamed of yourself!,” “Shame on you!,” and, anachronistically, “For shame!” However, we just as often invoke other emotions. Sometimes, we invoke the emotions of the person shamed, as when we describe their conduct as embarrassing or humiliating. Other times, we invoke our own emotions, such as disappointment, as when we say things like “I’m so disappointed in you” or “I expected more of you.”42 And ridicule is an aspect of some but not all cases of shaming, as when we describe someone’s conduct as “idiotic” or “stupid.” Finally, we should keep in mind that shaming need not involve language at all: sometimes, merely shaking your head or rolling your eyes is enough to shame someone. It is beyond the scope of the present discussion to explore the nature and scope of appropriate shaming, but it certainly seems possible to be prompted to worry that your belief that p is shameful by being shamed for believing that p. Imagine that you have confessed to a friend that, although you know it is politically incorrect, it seems to you that White Americans are subject to more discrimination than non-White Americans. “That’s ridiculous!,” they exclaim, “You know better than to believe something like that! I’m shocked you could be so naïve.” It is easy to imagine how this might cause you to reconsider the question of whether White Americans are subject to more discrimination than non-White Americans.

However, the fear of shame has important sources other than being shamed by others. Some also involve the presence of others. Shame or the fear of shame can be prompted merely by our awareness that someone knows that we did something—as when you realize that someone has seen you do something shameful or something that might be shameful.43 But shame or the fear of shame can also be prompted merely by our imagining what others might think or say if they did know that we had done something. And I think it is possible to be prompted to worry that your belief that p is shameful by merely imagining what others would think or say if they knew that you believed that p. This explains, at least in part, why we are motivated to “do our homework” and perform various kinds of “double checks” before seriously presenting something to an audience. Above, I asked you to imagine being shamed by a friend. But it is also easy to imagine reconsidering the question of whether White Americans are subject to more discrimination than non-White Americans merely because you know that your friend—or your friends in general—think it is an absurd and silly thing to believe.

I have been defending the proleptic value of shame with the aim of showing that doxastic shame might play an ameliorative role vis-à-vis pernicious beliefs like those constitutive of racial resentment. However, the proleptic value of shame, in the relevant cases, should not be overstated. If someone shames you for believing something, for example, you are unlikely to moved to reconsider your belief if you are convinced that they are your epistemic inferior—e.g. if you think that you are significantly more reliable than them when it comes to the relevant topic or that you have significantly better relevant evidence than they do. This is just an instance of a more general phenomenon: we will generally be indifferent to being shamed by someone whom we hold in low regard. What this means is that the proleptic dynamics of shame and shaming will be affected by phenomena like deep disagreement, political polarization, and conspiracy thinking. Shame and shaming cannot be expected to have beneficial consequences when parties involved regard one another as bigoted, uneducated, deeply confused, misinformed, or deranged. It seems to me that this is a feature of the proleptic value of reactive attitudes in general. Blame, shame, anger, resentment, and offense can only play a proleptic role when the relevant parties enjoy some basic level of mutual respect or esteem. If all this is right, the proleptic value of doxastic shame will be limited, but not non-existent.

Above, I assumed that blameworthiness entails responsibility (§3). Several recent accounts of “epistemic blame” imply that epistemic blameworthiness does not entail responsibility. Let us consider them, by way of comparison with my account of doxastic shame.

Jessica Brown (2020) argues that “epistemic blame” is a complex propositional attitude comprising a belief that someone believed badly and a desire that they not have believed badly, where believing badly is equivalent to violating an epistemic norm (399), e.g. the norm prohibiting dogmatically believing against your evidence. Her account is modelled on George Sher’s (2005) account of blame, on which blame is a complex propositional attitude comprising a belief that someone is bad or acted badly and a desire that they not have acted badly or not have a bad character (ch. 6). Given this account, Sher argues, blameworthiness does not entail responsibility, given that people are not always responsible for having a bad character (ch. 4). And Brown’s account implies that epistemic blame does not entail responsibility, given that people are not always responsible for believing badly.44

Antti Kauppinen (2018) argues that “epistemic accountability” involves reducing your epistemic trust in the person held accountable and Cameron Boult (2021; 2023) argues that “epistemic blame” essentially comprises a modification of your “epistemic relationship” with the person blamed, which involves a diminished tendency to trust them as a source of information. It seems to me that Kauppinen and Boult agree more than they disagree, so although Kauppinen says that “epistemic accountability does not involve blame” (2), what he has in mind is that epistemic accountability does not involve affective reactions like resentment and indignation (5), which is true of epistemic blame, on Boult’s account. To simplify our discussion, I am going to treat the idea that Kauppinen and Boult agree about as the idea that epistemic blame involves a modification of your epistemic relationship with the person blamed. When you epistemically blame someone, in the relevant sense, you will typically subtract “credibility points” from them (Kauppinen 2018: 6) or “suspend [your] presumption of epistemic trust in that person” (Boult 2021: 525). Boult models his account on T.M. Scanlon’s (2008: ch. 4) account of blame, on which blame essentially comprises a modification of your relationship with the person blamed. On Scanlon’s view:

And to blame a person is simply to modify your relationship with them accordingly (128–129), which typically involves “changes in our readiness to interact with him or her in specific ways,” such as refusing “to make agreements with [them] or to enter into other specific relations that involve trust and reliance” (143). On this view, blameworthiness does not entail responsibility—an involuntary or coerced action might reveal the relevant attitudes, warranting the corresponding relationship modification.45 For the same reason, on Kauppinen and Boult’s view, epistemic blameworthiness does not entail responsibility. Discovering your extreme gullibility might give me cause to treat your testimony with suspicion in the future, even if you are not responsible for being extremely gullible. As Boult (2021) writes, to say that someone is epistemically blameworthy is just to say that they have “done something that indicates they fall short of the normative ideal of some epistemic relationship they stand in with some other person” (524), and there is no reason to think that we are always responsible for doing things that fit that description.46

Obviously, it is reasonable to believe, of those who have believed badly, that they believed badly and to desire that they have believed otherwise. Obviously, you should trust someone less if you find out that they tend to form false or irrational beliefs. Moreover, epistemic blame, as understood by these authors, can plausibly play a proleptic role in ameliorating undesirable intellectual conduct.47 Consider how being epistemically blamed for believing that p, on either of these views, might motivate inquiry into whether p. Brown (2020) argues that epistemic blame entails a disposition to “rebuke the target … in an attempt to get her to appreciate the evidence she previously ignored or flouted” (402; see also 399–400; cf. Sher 2005: 95–96). And Kauppinen (2018) argues that “losing credibility … is bad and undesirable for a person” (7), and therefore:

In effect, the suggestion here is that your epistemically blaming someone can affect them when it is expressed or otherwise revealed, e.g. by your telling them that you think that they made an unappealing mistake (Brown) or by your evident indifference to their testimony (Kauppinen).

All this is consistent with my conclusion that that it is sometimes fitting to be ashamed of your beliefs. My account of doxastic shame and these accounts of epistemic blame are not competing accounts of some phenomenon. There is room for both shame and epistemic blame in our socio-intellectual lives and both may play distinctive proleptic roles.

However, I think those roles are importantly different, because of the distinctively epistemic character of epistemic blame. The proleptic value of epistemic blame depends, given Brown’s account, on your caring about violating epistemic norms48 and, given Kauppinen and Boult’s account, on your caring about preserving your “epistemic relationship” with the person who epistemically blames you. By contrast, the proleptic value of shame (§5) does not depend on your having any such “epistemic” concern, per se, but rather depends on your caring about violating legitimate expectations (§2). It is easy to imagine cases in which you might be more or less indifferent to epistemic blame but nevertheless susceptible to doxastic shame. In one kind of case, violating an epistemic norm strikes you as no big deal, by contrast with believing something shameful. Imagine that you are a high-school basketball coach and believe, as a result of wishful thinking, that your daughter is the best player on your team. It would be one thing to have it pointed out that your belief violates the epistemic norm of impartiality and another to have it pointed out that your belief represents a shameful instance of nepotism. In another kind of case, damaging your “epistemic relationship” with someone strikes you as no big deal, by contrast with believing something shameful. Imagine that you are a pediatrician and believe, as a result of professional negligence, that MMR vaccines cause autism. And imagine that you express this belief to a stranger at an airport lounge, whom you will never meet again. It would be one thing for the stranger to inform you that they are resolved no longer to trust you (at least on medical matters) and another for them to point out that it is shameful for a pediatrician to be so ill-informed about pediatric medicine.

I have argued that it is sometimes fitting to feel shame for believing something, even if you are not responsible for so believing. Even when your believing something is not blameworthy, it might still be shameful. Even if blame would not be fitting, shame might be.

I have been concerned here with the fittingness of blame and shame for beliefs, and have discussed neither the extent to which we blame other people or ourselves for believing things nor the extent to which we feel shame for believing things—nor the extent to which we shame other people for believing things. We should keep in mind that there are negative reactions and critical evaluations of other people’s beliefs, or of our own beliefs, that involve neither blame, shame, nor shaming. We may be disappointed or frustrated that someone believes something; we may lament that they believe something; we may think that their believing something reveals them to be a fool, or a coward, or a jerk. None of this straightforwardly involves blame or shame.

I have neither argued nor suggested that irrational or unjustified beliefs are always shameful. Epistemologists may therefore be unsatisfied by my conclusion, as it lacks the characteristic universality of “epistemic evaluation”—i.e. the practice of attributing justification and knowledge. However, my argument has consequences for epistemology. In particular, given the proleptic function of shame (§5), we need to think about whether and in what way epistemic evaluation should be understood as proleptic, i.e., as a forward-looking practice that aims to affect people’s future intellectual conduct, rather than a backward-looking practice of keeping track of their past intellectual performance.