Can outrage be truly moral?

Abstract

We critically examine the concept of moral outrage within the context of contemporary psychological and moral theory. The prevailing assumption that moral outrage should be a form of disinterested anger—where the individual is uninvolved—presents significant challenges for understanding its moral character. We propose a socioconstructionist approach to moral emotions, arguing that outrage is not inherently disinterested but rather deeply embedded in social norms. By examining moral anger and outrage through this lens, we show that outrage, though often seen as subjective or selfish, can still serve prosocial functions and uphold moral standards. We discuss how norms shape the expression of moral emotions, suggesting that outrage not only responds to norm violations but also plays a role in shaping these norms. We conclude by proposing a more nuanced understanding of moral outrage that integrates personal involvement and normativity to reconcile its subjective nature with its social and moral significance. This approach provides a framework for more robust empirical studies and a deeper understanding of the relationship between emotions, morality, and social norms.

It might not be an overstatement to say that there is an explosion of research on moral outrage. This might just be a reflection of our times, where social injustice is more visible and moral outrage readily voiced. Ranging from studies intending to shed light on the role of outrage in activism (Thomas & McGarty, 2009) to research on the neural underpinnings of outrage and morality in general (Pascual et al., 2013), there is an intense and genuine interest in understanding how outrage is shaping our lives. This increased attention on the topic has not, however, brought about a wide and shared understanding of the nature of the phenomenon at hand and, on the contrary, has yielded some contradictory findings.

In this paper, we aim to clarify in what sense outrage can be considered moral, and to what extent the lack of consensus on outrage’s nature and moral relevance is preventing extant research from reaching, if not a unified view, a less fragmented one. One assumption that, we believe, particularly hinders progress is the idea that outrage is disinterested anger. This view comes from the wider assumption that, in order to be truly moral, a moral emotion requires a disinterested experiencer. Thus, we first review theories on moral emotions to better understand what is at stake in considering outrage moral. Then, we critically discuss research on outrage that assumes that for it to be moral, it must be disinterested. We argue that this approach not only leads to paradoxes (i.e., to the extent to which every emotion involves the self, either all emotions are somehow moral or none of them are), but faces the impossible task of parsing out, empirically, the selfish and selfless elements of outrage. Finally, we propose a socioconstructionist approach, emphasizing social norms' crucial role in defining outrage’s moral qualities, offering a new understanding of this complex emotion and, crucially, one that makes its moral character compatible with the involvement of the self.

What is a moral emotion anyway?

The concept of moral emotion in psychology was initially put forward and popularized by Haidt (2003) and it has since been further developed (Cova et al., 2015; Tangney et al., 2006). Haidt (2003) notes that philosophers have generally taken one of two approaches to define morality: one based on the formal features of moral statements (universality, prescriptivity, categoricalness), and the other on the “material conditions of a moral issue” (e.g., to be in the interest of others; p. 853). Considering that a linkage between morality and language would not allow him to account for the morality of prelinguistic moral emotions, Haidt (2003) adopts the second approach and proposes that moral emotions are those generally “linked to interests or welfare of either society as a whole or at least of persons other than the judge or agent” (p. 853). Since emotions share certain prototypical components, such as elicitors and action tendencies, Haidt unpacks the material conditions for the moral character of moral emotions among these two components, arguing that the more disinterested the elicitor, and the more prosocial the action tendency, the more likely the emotion will be deemed moral.

While the second feature, prosociality, seems readily at hand when checking our intuitions on what a moral emotion would be, disinterestedness does entail further assumptions. Traditional normative ethical theories assume some version of a divided-self moral psychology, where acting morally means to defer to reason over desires and emotions. In deontology, most notably, to act on our inclinations is the very definition of failing to be moral: even if they motivate us to do what is right, our action must come from our recognition of its necessity, and never from its expediency. In utilitarianism, the imperative to maximize well-being for the many often involves acting against our own interests. In Aristotelian virtue theory, phronesis dictates the mean and guides the formation, through our pleasures and pains, of our character. Thus, if moral emotions are to be moral, Haidt (2003) thinks, they should encompass this very feature of disinterestedness: they should be brought about by, or be about, something above and beyond one’s own self.

The idea of moral emotion espoused by Haidt (2003) presupposes a basic notion of morality (e.g., the kind of considerations associated with disinterest and societal well-being) that is tremendously useful as a research program, allows for operationalizable variables, and avoids begging the question (that is, to assume that a certain emotion is moral to then discover that it is). By identifying the extent of the “disinterestedness of the elicitors” (which we read as the extent to which one’s selfish concerns are involved), and the extent to which the action tendency is prosocial, we can assert and even measure its moral “stature.” Besides, identifying different combinations of elicitors and action tendencies can potentially yield a full taxonomy of moral emotions, while also allowing us to distinguish them from nonmoral emotions. Notice, however, a couple of key aspects in Haidt’s definition.

First, seemingly obvious moral emotions like shame or embarrassment occupy an uncomfortable position, for whereas their action tendencies are clearly prosocial, their elicitors involve the self. The same happens with moral emotions of victims of transgressions (e.g., where it is ourselves, or those close to us, who are unjustly wronged), or emotions associated with purely prudential considerations (e.g., affective responses implied in self-care, such as sympathy for one’s own misfortunes, pride for one’s achievements, etc.; Haidt, 2003).

Second, according to Haidt (2003), elicitors that are morally relevant are those where the self has no stake in the triggering event, where elicitors are understood as a conflation of events and their interpretation (Haidt, 2003, p. 866). This makes it difficult to flesh out the link between the relevant components of the emotion, since it requires clarity on when, and in what respect, the elicitor does or does not place the self “at stake.” This seems to be the case with guilt: in a certain respect the triggering event (the harm done to other) does not involve my self, but it clearly does to the extent that it is the fact that I caused the harm that triggers the guilt. It also requires clarity on when, and to what extent, the action tendency initiated by the emotion is genuinely prosocial. This could be the case with embarrassment, which prompts me to hide, serving both prosocial and prudential ends. It is worth noting that Haidt (2003, fig. 45.1, p. 853) believes shame and embarrassment to be less disinterested than guilt, presumably because guilt is all about the other’s harm, whereas shame and embarrassment are about the extent to which one’s self fails to meet others’ standards. While these points are problematic (we come back to them below), they have not been a hurdle to the recent vigorous research on moral emotions.

Moral emotions are traditionally categorized into four families depending on whether the focus is on the transgressor (other-condemning: contempt, anger, disgust), on the victim (other-suffering: sympathy), on meeting a moral standard (other-praising: gratitude, elevation), or on evaluating the self (self-conscious emotions: shame, guilt, embarrassment; see also Rozin et al., 1999).

Note that outrage is mostly operationalized as moral anger (see Batson et al., 2007; Dubreuil, 2015; Hechler & Kessler, 2018; Lomas, 2019; Rothschild & Keefer, 2017; Russell & Giner-Sorolla, 2011), though not uniformly. ¹ Moral anger is one of the other-condemning emotions (Rozin et al., 1999), triggered by an unfair event and with punishment as its action tendency (Haidt, 2003). In consequence, according to Haidt’s approach, moral anger can be differentiated from less moral forms of anger by assessing both the disinterestedness of the elicitor and the prosociality of the action tendency. As to the first, we could say that anger at unfairness where the self is directly involved would be less moral (that is, less distinguishable from a self-interested response to the inability to achieve one’s goal) than one where the self is not, and where we can more safely assert that it is unfairness that triggers it. As to the second, anger that prompts the direct removal of a threat would be less moral (less distinguishable from the advancement of a personal agenda) than anger that motivates punishing a thief, which ultimately helps uphold sociomoral norms.

Refinements and alternative analyses on the notion of moral emotion in psychology ² have focused on the functions of these emotions as a way to individualize them. Moral emotions, according to Tangney et al. (2006), confer to their bearers some form of long-term benefit in their interactions with other members of society, by motivating moral behavior and inhibiting immoral behavior. Emotions in general may have several functions, and only some of them can be morally relevant. For example, emotions may have evaluative (e.g., blame), commitment, motivational, and communicative/signaling functions (Noon, 2019), and their moral character depends on whether these functions themselves serve a moral purpose.

In this vein, Giner-Sorolla’s (2012) Integrative Functional Theory (IFT) of emotion offers a more nuanced analysis. According to him, a functional analysis expands on Haidt’s (2003) basic intuition that disinterest and prosociality account for the morality of moral emotions. However, they do so by describing not the process of the emotion but its outcome (Giner-Sorolla, 2023). An emotion process could be self-interested, but if it adapts to a goal that benefits others it is functionally disinterested and, therefore, moral. On the other hand, an appraisal-to-action process, an action tendency, is only one of the four ways in which these emotions could functionally support prosociality. The other three ways include means of (a) association, (b) self-regulation, or, crucially, (c) communication. Gratitude, for example, self-interested though it may be, benefits others; furthermore, it prompts us to give back to those who have helped us, it conditions us to be grateful in the presence of similar stimuli, it allows us to regulate our behavior vis-à-vis those who help us, and it conveys to others honest signals of our moral worth. Combined, all these function to promote and uphold important social norms related to helping others. Applied to moral emotions, the IFT offers a helpful guide to evaluate whether an emotion is actually moral (by examining the functions it fulfills). However, this focus on the outcome, or the adaptation to a moral outcome, entails that moral emotions may not be moral in and of themselves. Or, at least, that we still have to explain how these moral functions reflect back on our subjective experience. After all, there seems to be a distinctive feature present in emotions like shame or outrage that is not present in emotions like sadness or joy, even though these emotions may benefit others, accomplish moral functions, and support moral norms and practices. In other words, could there be something in the process of an emotion, and not only in its outcome, that makes it moral?

We have thus far identified that the psychological approaches to the definition of moral emotions face two sets of problems: (a) first, the problem of whether, and to what extent, disinterested elicitors and prosocial action tendencies account for the morality of moral emotions, and (b) second, whether it is at the level of the process or at the level of the outcome that these features occur. Be what it may, it is clear that we appraise the relevant action as (im)moral, that we act somewhat prosocially as a result, and that these moral emotions work as commitment devices. The commitment, in this case, is to behave in a way that is costly in the short term (a moral way) and to signal its content appropriately as deriving from a moral interpretation of the situation.

This further suggests that, if we want to build an empirically robust and coherent research program, a more precise (and perhaps emotion-specific) description of the way moral outrage is construed is required. Drawing from Haidt’s (2003) and the IFT approaches, we can go a step forward in this direction by analyzing both the process and the functions performed by moral outrage.

Can outrage ever be moral? Impartial morals and biased anger

Psychology tries to understand expressions of moral outrage in the context of our day-to-day experiences. Some research has explored how online media has reshaped moral outrage’s expression and implications, examining its role in social behaviors but also questioning its moral basis (Crockett, 2017; R. Fan et al., 2019; Sawaoka & Monin, 2018). Studies that call outrage’s motives into question suggest that it sometimes serves self-interests rather than moral principles (Jordan & Rand, 2020; Rothschild & Keefer, 2017; Rothschild et al., 2013). These raise questions about outrage’s morality and its capacity to drive social change, highlighting a need to reassess its emotional and moral identity within psychology.

In his systematic review, Lomas (2019) identifies a multiplicity of triggers of moral anger, ranging from betrayal and injustice to unfairness and, more generally, witnessing violence. It would seem that moral anger is the default emotional response to any moral violation. If so, what is particularly moral about outrage?

Let us assess the underlying appraisal of outrage. Anger may involve several different appraisals: goal blockage, group antagonism, threat from a conscious being, threat to the self (Lazarus, 1991), personal goal frustration, and injustice and unfairness (Giner-Sorolla, 2012). Prima facie, only the latter seems a genuinely moral appraisal and, thus, an elicitor of outrage. This is what Haidt (2003) probably had in mind when he posited the disinterestedness criterion: elicitors where the self has no stake tend to trigger moral emotions whereas elicitors involving the self tend to trigger nonmoral emotions. Haidt’s description in terms of elicitors allows him to talk about events and their interpretations as triggers of emotional episodes, in a way that recalls the core relational themes (Lazarus, 1991) of classical appraisal theory. Thus, these elicitors/appraisals are considered moral only when “the self has no stake in the triggering event” (Haidt, 2003, p. 853). Only situations interpreted as unfair or unjust, and where one is not personally involved, would be elicitors/appraisals of moral anger.

A virtue of this analysis is that it clearly precludes the possibility of outrage when the triggering event is a mere goal blockage or frustration (e.g., being angry at my computer for not saving the last paragraph), or even a goal blockage resulting from someone’s intentional action that cannot be considered unfair (e.g., being angry at a teacher for asking me a question in class). An undesirable consequence, though, is that it also precludes self-interested, legitimate moral claims (e.g., outrage at being victim of an injustice) or claims where a moral wrongdoing concerns somehow the self (e.g., your brother is a victim of an injustice) as elicitors/appraisals of outrage.

Batson et al. systematically explore these ideas (Batson, 2011; Batson et al., 2007, 2009; O’Mara et al., 2011) and take them to their logical conclusion: there are not really any moral emotions, outrage the least of them, since personal involvement is almost invariably present in moral assessments and their subsequent emotional responses. In their experimental work, Batson et al. show that empathy is the best predictor of self-reported outrage. Thus, outrage is not a “disinterested emotion,” since it appraises situations and motivate behaviors that are ultimately related to the self (helping a friend, considering in-group injustices as worse than equivalent out-group injustices, etc.). In other words, moral anger is indistinguishable from personal anger and, as such, it is either an unnecessary construct or, at best, a framing strategy to legitimize selfish claims.

If the trigger of outrage is the violation of a norm where the self has no involvement whatsoever, then learning through the news about injustices in distant communities must be equivalent to witnessing them directly, since the triggering event (the norm violation) remains the same. The cases Batson examines are perfectly fitted to test this claim. Batson et al. (2007) and O’Mara et al. (2011) examine unfair exclusion from a group and unfair distribution of goods, where the level of empathy (Batson et al., 2007), or the degree of proximity and fairness (O’Mara et al., 2011), are manipulated. In the first case, they find that empathy is the best predictor of anger (personal anger is indistinguishable from moral anger) and that experiencing an exclusion is associated with higher levels of self-reported anger—suggesting anger at harm to the self but not at unfairness per se. Furthermore, Batson et al. (2009) consider the case of torture, a morally unambiguous situation, and find that self-reported moral outrage by Americans is higher when the victim is an American soldier, despite participants judging torture as similarly immoral regardless of the victim’s nationality, and regardless of the potential instrumental gains that could be achieved through torture. Notice that these findings are consistent with research showing that fairness considerations are modulated by group membership (Abbink & Harris, 2019; Mendoza et al., 2014). All the cases analyzed by Batson et al. show that outrage is never truly disinterested and that, if anything, disinterestedness is a predictor of the possible absence of outrage (Batson, 2011, p. 234).

Other research on outrage explores a similar motive. Hechler and Kessler (2018) propose that moral anger is triggered by the intention of causing harmful effects, independently from the actual action and its outcomes. Since the negative outcomes of an unfair action result in threats to the self or empathy with the victims, outrage is not really about the self, although the self ends up involved. This dissociation saves nicely the disinterestedness gap. As far as our analysis is concerned, a key point is their assumption that the self must not be involved in the elicitor/appraisal (for other work with a similar premise, see Van de Vyver & Abrams, 2015).

Although it is possible to criticize some of this work based on methodological concerns, ³ what we find more intriguing about it is the extent to which the assumption that experiencing and identifying genuine moral anger requires no involvement of the self goes unquestioned. Batson (2011) even suggests, anticipating later discussions on virtue signaling (Levy, 2020), that widespread talk of outrage might be symptomatic of moral hypocrisy, for it invariably disguises one’s own disinterest.

From Batson’s point of view, it is not difficult to call into question the moral usage of other emotion functions. Regarding action tendencies, it follows that “to the degree that these emotions are not moral emotions, any motivation produced by them is not likely to be moral motivation—i.e., directed toward the ultimate goal of upholding moral standards” (Batson et al., 2009, p. 159). The same idea guides some of the challenges to altruistic punishment, a behavior motivated by emotions like outrage (Pedersen et al., 2013, 2018), and to emotionally motivated behavior more generally (McAuliffe, 2019). Regarding the self-regulation function (which partly overlaps with the action tendency), some problems follow as well. For example, it has been found that outrage modulates threats against one’s moral identity, that is, that outrage can alleviate guilt and help restore a sense of morality. Rothschild and Keefer (2017) dubbed this phenomenon “defensive outrage” and suggest that it can be differentiated from real outrage by individual justice sensitivity (Rothschild & Keefer, 2018). Protecting one’s moral identity allows one to reduce the incongruence between their objectives and their actions, thus helping them to provide a coherent narrative for their actions, and therefore does not constitute a genuine moral emotion. As far as the communicative function is concerned, outrage can also be self-serving (Giner-Sorolla, 2012) and fulfill only instrumental purposes, as in the case of virtue signaling (Jordan & Rand, 2020; Levy, 2020). In fact, Kupfer and Giner-Sorolla (2017) show that people routinely interpret expressions of anger as being self-interested. Jordan and Rand (2020) show that reputation cues automatically activate expressions of outrage, because reputation is normally at stake in situations when a norm is violated and action is possible. The implication is, then, that expressions of moral outrage are always intertwined with reputational concerns.

A clear conclusion follows from this approach: given the abundant evidence of self-involvement in its appraisal, expression, and self-regulation, moral outrage is at most very rare. Even if we encountered a real case of moral outrage, we would not be able to recognize it, for it would be nearly impossible to tease apart its constitutive moral appraisal, action tendency, or communicative function, its selfish from selfless elements. There is a “natural kind” of events, namely, norm violations, that trigger outrage, but evidence shows they rarely, if ever, do so without personal involvement.

As we suggested earlier, this conclusion runs counter to our intuitions and daily experiences. Presumably, most people experience outrage regarding a matter that is personally relevant to them, such as an unjustified increase in taxes or police brutality in their communities. For most people, moral appraisals might only be possible on matters relevant to the groups they belong to: from families to schools, and more complex organizations like cities and countries. Arising from this consideration, let us look into three points that call into question the impossibility of moral outrage: (a) reactions to injustices toward the in-group, (b) moral disagreement on (un)fairness, and (c) co-occurrence of emotions when facing injustice.

Quintessential interested moral outrage occurs when reacting to injustices against the in-group. In-group moral emotions underlie reactions to unfair social policies (Giner-Sorolla, 2012; Wakslak et al., 2007) and even help explain the rise of social movements (Jasper, 2011; Pagano & Huo, 2007). Clearly, characterizing social protest as motivated by selfish motives does not adequately capture the idea of resisting an injustice. That in-group biases guiding our judgments also shape our discussions and disagreements on moral matters surely does not deprive them of their moral weight, nor transform them automatically into prudential issues. As Giner-Sorolla (2012) points out, when a situation is interpreted as both unfair and a threat to one’s group, a legitimate self-interested, moral claim might occur.

There is evidence suggesting that our own ideological commitments determine how injustice actually triggers moral outrage. Research on system justification theory (Chapleau & Oswald, 2014; Krauth-Gruber & Bonnot, 2020; Wakslak et al., 2007) shows how ideologies and beliefs shape our perception of what counts as unjust, and thus highlights the deeply group-relative nature of outrage. Wakslak et al. (2007) show that endorsing a system-justifying ideology reduces outrage about social and economic inequalities, thus lessening aid to the disadvantaged. This shows that moral outrage cannot be sufficiently described without considering one’s point of view, in this case one’s moral beliefs and ideologies. Labeling as mere “anger” the emotions experienced by individuals and groups about issues that affect them directly does not allow us to acknowledge their calls for justice and thus their status as possible victims.

Moral disagreement would also be impossible if moral outrage were impersonal and triggered by the impartial assessment of unfairness. Only concerning distributive justice, it is possible to identify judgments guided by deservingness, equality, or necessity (Walzer, 2008) that are not necessarily codified as laws. Some in a community might believe resources must be assigned based on merit (to those who deserve them) while others believe resources must be assigned based on need. This would be a genuine case of moral disagreement, with competing moral claims and competing moral emotional reactions. If a moral emotion requires a “view from nowhere,” to borrow Nagel’s (1986) expression, moral disagreement would be very difficult to account for. After all, outrage triggered by justice-as-deservingness infractions is just as moral as outrage triggered by need-based-justice infractions. The shape this kind of disagreement takes in real life crucially depends on the fact that the competing claims are socially and politically situated, that is, that they express a certain point of view.

Lastly, in the case of injustice against the in-group, moral outrage is only one of several emotions we can pin to the situation—fear, hate, and contempt may also occur (Jasper, 2011). Moral outrage is not all there is when we consider the emotional distress caused by being subject to an injustice. Does this mean, then, that outrage is less moral when it occurs in tandem with more self-centered emotions? What about being a victim of intentional harm? Describing emotions as successive is generally an analytical simplification that does not do justice to the complexity of the subjective experience in a highly charged emotional episode. Co-occurring and mixed emotions (Berrios et al., 2015) are a robust experience, one very likely involved in cases of violation of moral standards, where the same situation can be appraised in several, even contradictory ways (Larsen & McGraw, 2014). Recognizing that unfair treatment elicits not only outrage but also sadness or fear does not contradict the fact that the appraisal of this outrage consists of a breach of a norm.

In order to offer a description of moral outrage that, on the one hand, guarantees its status as a coherent emotional response to injustice and, on the other, allows empirical research that distinguishes moral from nonmoral components, two points need clarification. First, how exactly the involvement of the self, far from debilitating the moral status of moral outrage, coexists with the injustice and unfairness that triggers the moral outrage. Second, how different functions of moral outrage, and different co-occurring emotions, interact with one another. The first point requires constructing a notion of morality that brings to balance, at the level of our emotional response, concern for others and also for oneself, co-operation, and pro-social behavior but also the expression of conflicting social and political claims. The second point requires spelling out in more detail the components of the outrage appraisal so that we can adequately isolate the interplay between this and other functions (see IFT above), and between outrage and other emotions.

Norms and emotions

To unpack the tension between involvement of the self and the moral character of outrage, let us rehash the classical Euthyphro dilemma: “Is the pious loved by the gods because it is pious, or is it pious because it is loved by the gods?” (Plato, 380 B.C.E./1995, Euthyphro 10a). For our purposes: Does what is outrageous prompt that emotional response because it is outrageous or is it outrageous because it prompts that emotional response? Given that the outrageous involves an assessment of injustice, and such assessment appeals to an allegedly objective, impartial right or good, the question may be put this way: Does the appraisal prompt outrage because it is allegedly objective/impartial, or is it allegedly objective/impartial because it prompts outrage? If the first, how can an objective, impartial moral judgment (the appraisal/trigger) be constitutive of, cause, or co-occur with in a meaningful way the always-subjective, partial emotion where the self is inevitably involved? If the second, how can outrage depart from a plausible, even if incomplete, interpretation of the situation? How can it be informative of a situation, group-bonding, and, therefore, adaptive?

Discussions between objectivist and subjectivist accounts of emotions abound in philosophy (Sousa, 2001) and have direct echoes in contemporary psychology. Kohlberg’s (1981) objectivist model assumes, as the goal of sociomoral development, an ideal observer who transcends the limits of parochial and conventional norms to ponder universal issues of autonomy and justice. Given that violations of fairness and autonomy are self-evident, taking any point of view would imply a partial view and thus a nonmoral (at most conventional?) judgment and decision. The objective nature of unfairness and injustice fits well with the impossibility of moral outrage, since emotions are by definition experienced by someone, from a point of view, and thus are partial. It is no coincidence that objectivist accounts, such as this one, rely on a strict separation of reason and emotion, and depict moral development as the transcendence of the latter, not as something to be achieved within or along with our emotions.

Haidt’s (2003) subjectivism suffers from the opposite problem. If emotional, intuitive reactions are enough to moralize a situation (e.g., outrage projects injustice onto the world), then it is not possible to have agreements on injustice, nor can emotions have any normative value. ⁴ The subjective nature of unfairness and injustice makes the possibility of moral outrage tenable, but it blurs the difference between moral and nonmoral concerns: as far as the appraisal is concerned, either our subjective experience makes situations fair or unfair, and then all emotions are somehow moral, or it can never make them fair or unfair, and then no emotion can be moral in a significant way.

Framing the problem of moral outrage as a tug of war between objectivist and subjectivist approaches irremediably leaves us with either a lofty and distant observer, to whom we cannot attribute emotion, or a situated actor, driven exclusively by selfish interests, of whom we cannot predicate moral concerns. If moral outrage is to be found somewhere, it is where subjective and objective concerns meet: the values we endorse, the norms that express them and their scope.

Think of the case of witnessing someone cutting in line: this might provoke a reaction ranging from mere annoyance to anger. If you are informed that this person did not know about this norm or similar kinds of norms (because they come from a very remote place), or could not recognize the situation as a situation where the norm applies (because they have a vision impairment), it is almost certain you would excuse their behavior and your anger would recede. No matter how general a norm is, the point stands: injustice always involves a norm and its scope of application. If I recognize that a norm does not apply to a particular situation, I cannot interpret the situation as a norm violation and, therefore, outrage would not occur. But, very importantly, if I do not recognize the norm as applicable, or at least related, to me, I cannot interpret the situation as a norm violation either and, therefore, outrage would not occur. For example, if I come as well from a very remote place and do not recognize queuing as applicable to me, I would not be outraged by a person cutting in. Something similar happens, we believe, when I consider a norm as applicable to the other person and not to me, but where the norm is somehow related to me. For example, when a teacher gets angry at a student for talking in class. The norm applies to the student, does not apply to the teacher, but it clearly relates to the teacher and their activity in class. In his Rhetoric, Aristotle recognizes that often it is rank that inspires anger, for instance when a superior feels anger when an inferior wrongs them (Aristotle, 330BCE/2001, Rhetoric, 1379a5).

The scope of norms always concerns myself and at least one “other”: this contrast between myself and others, between my group and others, is at the core of what counts as fair, and obliges us to some sort of moral reasoning: Why does this norm apply to you? Why does it apply to you, vis-à-vis me? What’s moral about it is not necessarily its abstractness and objectivity, let alone its validity and truth, but the fact that it looks to the questions why does a norm apply to someone else, and how is the norm related to me—the answers to which may range from the most parochial, subjective considerations to the most universal, objective ones.

This sorts out the Euthyphro dilemma: the outrageous prompts a particular emotional response because it is outrageous, that is, because there is recognition of a norm violation. However, this recognition, although about others, involves the self. And this is the case whether dealing with subjective or objective considerations. This explains outrage’s enormous plasticity to framing strategies: we can manage to keep ourselves outraged by making norms applicable to others, by relating them to ourselves, and by invoking in the process the whole range of available reasons, from the most parochial to the most universal. Similarly, we can manage to not feel outrage even when we apparently should: by questioning either the applicability of the norm or its relation to the self.

Emotions related to norms can be better understood as modes of perception (Sousa, 2001): they give us access to certain sorts of knowledge but also reveal value. Moral outrage (felt or observed) gives access to the norms that are deemed to be violated, as do disgust and contempt (when moral), and also reveals their respective value qua norms. Thus, one can only be angered at injustice when there is some shared standard of what it is. This standard, however, is not monolithic, but challenged by different claims in a permanent, complex process of negotiation, imposition, resistance, and acceptance. Outrage reveals a system’s values but also helps construct them. Describing an experience as moral outrage implies understanding the relevant normative aspects of the sociocultural system in which the emotion takes part.

Socioconstructionist theories of emotion (Averill, 1984; Barrett, 2017; Weber, 2004) allow the possibility of having shared norms as constitutive elements of the emotions. In common with appraisal analyses, socioconstructionist theories posit that moral anger is elicited by a situation appraised as unfair. Unlike appraisal theories, the socioconstructionist perspective emphasizes the social nature of these appraisals by directly appealing to the social norms that constitute and regulate emotions. Thus, emotions are both an expression of and are subject to social control. Identifying the norms that govern moral anger is then indispensable to its proper explanation.

A constructionist analysis

Given the inherently social nature of outrage, we rely on Averill’s (1984) constructionist analysis of anger to make sense of the fine mechanisms through which outrage occurs. Let us briefly outline in what sense emotions are socially constructed. According to Averill (1985), emotions are “socially constituted syndromes . . . that include a person’s appraisal of the situation and that are interpreted as passions rather than as actions” (p. 98). In this definition, the word syndrome refers to a series of processes (physiological, cognitive, and behavioral) that “run together” and which, only when organized and interpreted socially, can properly be called an emotional response. Thus, an emotion is socially constructed insofar as its underlying processes are clustered together in, and by virtue of, the social context we are in. For this reason, no specific process, or set of processes, constitutes a necessary and sufficient condition for an emotion to happen, although a specific emotion, say, anger, entails a rather typical set of processes. One of these is the appraisal, which does not antecede nor properly cause the emotion but rather comes upon us, as it were, in the package. Averill’s (1984, 1985) understanding of emotions can be framed in Gergen’s broader socioconstructionist approach. For Gergen, our knowledge of the social world is not, and cannot be, knowledge of a completely independent, taken-for-granted world. It rather is mediated by terms that crucially depend on the “vicissitudes of social processes” (Gergen, 1985, p. 6).

Averill (1984) proposes several types of rules that constitute and regulate emotions: rules of appraisal (which define the proper object of anger), rules of behavior (relevant to understand the normative action tendencies and expressions of anger), rules of prognostication (on the temporal evolution of anger episodes), and rules of attribution of anger (both on events and the self). Relevant for our current purpose are the rules of appraisal. Averill (1984) distinguishes between the object of an emotion and its proximal cause, where only the former relies on social norms. The objects of anger, which allow us to tell apart different kinds of emotional episodes, comprise instigations, targets, and objectives. Instigations, generally, refer to unjustified harms, norm breaking, and even negligence; targets to the relevant persons or institutions; and objectives to the correction of wrongdoings. In Averill’s scheme, moral anger can be then triggered by appraisals of injustice (representations of relevant norm-breaking events) aimed at a person or institution, with the objective of correcting the wrongdoing. Moral anger so conceived also implies an appraisal of responsibility and blame. The key difference between moral and nonmoral anger is then the social relations involved in both the object of anger and the rules of appraisal.

An interesting empirical consequence of this approach is that nonappropriate cases of outrage—that is, cases where outrage is considered unjustified or excessive—become key to understanding the appraisal of moral anger. Feeling outraged at my computer for breaking down after several years of use borders on the incomprehensible. This is presumably because this situation does not afford the possibility of responsibility attribution, which always requires intention. Online shaming, for instance, is a kind of moral reckoning of an offender frequently considered disproportionate to the point that it may lead to empathize and even justify the offender (Sawaoka & Monin, 2018). In cases like this, complex social norms on the appropriateness of the appraisal, the behavior, the prognostication, and the attribution are at play. These norms are revealed by these failures, as well as the conditions under which they are fine-tuned. There may be violations of norms of appraisal (we take for unjust something just), violations of norms of behavior (we overreact or underreact), violations of norms of prognostications (we tirelessly pedal our feelings of outrage and make everything an instance of the norm violation), or violations of norms of attribution (we take it personal when we shouldn’t, we don’t when we should).

Thus, outrage can also be a form of social struggle, negotiation, and even control (Nugier et al., 2007), and can help explain subsequent conflict when resisting these norms. These prescriptive and proscriptive rules on what is (un)justifiable as an adequate instance of moral outrage are constantly evolving and are always in need of being pinpointed: we see this very clearly in our online interactions. Outrage is not only moral as far as it concerns another’s violation of a norm I am related to; it is moral because the circumstances of its occurrence are heavily socially regulated—outrage implies some grasping of the larger social logic of the emotion.

Outrage’s plasticity (the fact that it is suitable to constant framing and reframing) and sociality (the fact that every aspect of it is socially regulated) account as well for its conflictive character. Anger is said to be a conflictive emotion (Averill, 1984), which means that there are competing social norms shaping it and its expression. These norms involve, on one hand, a duty to retaliate forcefully against perceived injustice and wrongdoing and, on the other hand, to resolve disagreements and instigations in a respectful and forgiving way. That is, although moral anger is many times the morally appropriate response to injustice, it can also lead to morally inappropriate behaviors, such as revenge or disproportionate punishment. This tension is recognized in legal traditions around the world, as Averill rightly points out: anger may backfire when it blocks the possibility of the transgressor’s acknowledging of the wrongdoing and when it creates violence. It is sometimes a duty to be outraged, but outrage is also a risky state.

While outrage is a reaction to a breach of norms, it is also governed by metanorms regarding its expression. This can be likened to a fire alarm, which can certainly be abused—it can be activated to draw attention, interrupt an event, or warn of an actual fire. ⁵ Our intuitive reactions to the abuse of outrage certainly speak about the possibility and nature of these metanorms.

The conflictive nature of outrage highlights the functional character of this analysis. Moral anger functions can be distinguished at three levels: (a) at the physiological level, it prepares us for aggression and action more generally; (b) at the psychological level it prompts the correction of wrongdoings; and (c) at the social level it helps us reassert, uphold, or contest (moral) standards of behavior. The possible conflict between the psychological function (correcting a wrongdoing could entail forcibly removing a threat or obstacle) and the social one (preserving social harmony) might explain the apparently ambiguous moral character of outrage. While this solution is not new and shares its basic elements with both IFT (Giner-Sorolla, 2012) and other functional theories of anger (Nelissen et al., 2013), our emphasis lies on the subject’s representation of the relevant norms and their scope and thus the process of the emotion. Only when one explicitly acknowledges or empirically determines the norm that the experiencer of outrage considers is being violated, and how it relates to the self, can one talk about moral outrage. These norms are not always transparent—they evolve; they are prone to constant negotiation, deliberation, and even contestation; different groups push different perspectives; and sometimes, their consequences are far from moral. But this does not exhaust the role of norms in the constitution of moral anger. Its rules of appraisal, behavior, prognostication, and attribution are also socially constructed, that is, they are clustered in and packaged in a particular way in the social context we are in.

Conclusions

By tracing back and examining the origins and developments around the idea of moral emotion, we have established that the problem lies in the conceptualization of its two main aspects—action tendencies and the moral character of moral emotion itself. While there is a wide agreement on what constitutes the prosocial character of the action tendencies of outrage, the disinterestedness aspect of the elicitor, which captures its moral character, is problematic. This is a fundamental challenge, since the question of what is moral about moral emotions hinges on how we are to understand the involvement of the self and concern for others. This led us to explore in detail how current theories on moral emotion conceptualize outrage. By conceiving outrage as moral anger, that is, anger at unfairness where the self is not involved, these theories do not manage to articulate the disinterested aspect of outrage with the necessary involvement of the self all emotions inherently presuppose. As a result, they end up conceiving an emotion either as triggered by purely objective judgments with no involvement of the self, or as a framing device to mask personal and selfish interests, which makes it a hypocritical emotion with a dubious moral nature. In the first case, outrage would not be an emotion, in the second it would not be moral. This is why some have claimed the impossibility of moral outrage in particular, and moral emotions in general.

In this paper, we have presented a reading of outrage as moral anger—an emotion extensively researched but only partially understood. This reading puts together three elements: (a) Haidt’s (2003) intuition that moral emotions encompass features of disinterest and prosociality, (b) Giner-Sorolla’s (2023) functional turn and focus on outcomes rather than processes, and (c) Averill’s (1985) view of emotions as socially constructed clusters of emotional processes. Moral anger is then an evaluative reaction to the breach of a norm, where this norm is recognized as applicable to others and also applicable, or at least somehow related, to me. Functionally, it is disinterested, for it benefits others (by deterring transgressors, by protecting possible victims, by sanctioning the importance of certain norms, by guaranteeing social stability and conforming a relatively fixed set of expectations, by establishing and maintaining hierarchies). It also supports prosociality through its action-tendencies (punishing, denouncing, signaling wrongdoings), its associations (by promoting anticipated shame: we can relate certain potential stimuli to my, or others’, future outrage), its self-regulation (it intensifies or abides in response to the transgressor’s further actions and their consequences), and its communication (it conveys clear signals of moral worth, like a general concern for justice even when one is not directly affected by it). It is, thus, an emotion clearly adapted to a moral outcome. Now, from the point of view of the process, it is an interested emotion, an emotion where the self is clearly at stake, and to no detriment of its moral character. In order to feel outraged, we need to relate the breached norm to ourselves in a manner that is constitutive of its appraisal: the other’s moral transgressions are in a nontrivial sense about me, otherwise I could not manage to feel outraged. Phenomenologically speaking, this is a crucial aspect of outrage’s lived experience: I become interested, I make a social matter directly my own. Finally, outrage shows itself as socially constructed in two interconnected ways. First, its rules of appraisal, behavior, prognostication, and attribution are the result of socially organized and interpreted processes. But, secondly, it is all about norms: outrage is an integral part of the process whereby social norms are constantly created, upheld, and supported, but also negotiated, tested, and contested.

This socioconstructionist analysis of moral outrage is not the only way to incorporate norms in the explanation of moral emotions. Nelissen and Zeelenberg (2009) do a similar thing by distinguishing proximal and ultimate functions. Fitouchi et al. (2022) put the notion of obligation, a normative concept, at the center of what a moral emotion is, making anger and guilt more prototypical as moral emotions. Even more fine-grained appraisal analyses (Scherer & Moors, 2019) incorporate these concerns, as do alternative versions of social constructionism (especially Barrett, 2017). We put forward an approach to the study of moral outrage (extendable to other moral emotions and compatible with the aforementioned analyses) that puts it against the backdrop of a normative net, where norms are constantly upheld, challenged, and transformed. This approach allows us to (a) make sense of the involvement of the self, (b) articulate its disinterestedness and prosociality aspects while maintaining the possibility of moral disagreement, (c) account for the interplay between different functions (psychological and social) and between different levels of analysis (appraisal, attribution, behavior, prognostication), (d) research more accurately the mechanisms of its plasticity and conflictive character, and (e) identify inappropriate outrage, how it falls outside standards, and which standards it falls outside. Crucially, to study moral emotions, we do not need a fully fledged notion of morality or a hard divide between moral and social norms. Outrage brings forth a whole array of moral claims (partial or impartial, subjective or objective, particular or universal), configuring a changing, contested, evolving landscape where moral reasoning progresses individually and collectively.

Conceiving moral outrage under this lens also brings some exciting challenges in seeing it as an evolving emotion. Establishing new social norms related to fair behavior is surely associated with outrage at their violation. Is outrage key for understanding how a social norm gains its force? This also entails specifying how outrage, conceptualized from a socioconstructionist viewpoint, can be coherent with a causal explanation of appraisals analyses (Scherer & Moors, 2019), where the elements we bring about to interpret a situation as outrageous can be adequately distinguished and, hopefully, controlled in experimental research. All in all, a better understanding of outrage and its social norms might be key to understanding its moral character and causal role, and, even more, to understanding how fairness is born and evolves from our emotions. As far as the morality of moral emotions is concerned, the case of outrage confirms, we believe, Haidt’s (2003) fundamental intuition and Giner-Sorolla’s (2023) take on functions. However, it goes further to show how disinterest and prosociality at the level of outcomes can reflect back to the subjective constitution of an appraisal that is particularly moral: the personal appropriation of a certain norm or standard that makes their compliances, transgressions, imbalances, restorations, and so forth a matter of my own.