Shame, shaming, and their role in advocacy

The Oxford English Dictionary says shame is ‘a painful feeling of humiliation or distress caused by the consciousness of wrong or foolish behavior’. Psychologists tend to emphasize the internal psychological state associated with the emotion of shame, highlighting the perceived ‘discrepancy between ideal and actual self’. Sociologists and some social psychologists tend to emphasize the external social implications of shame, saying for example that ‘in common parlance, shame is a negative, crisis emotion closely connected with disgrace’ which gives rise to ‘the feeling of a threat to the social bond’ (Scheff Reference Scheff2000, 97). The element of disgrace, defined by the OED as ‘loss of reputation or respect as the result of a dishonourable action’, emphasizes shame's social consequences. Even social psychologists, however, acknowledge that shame is a psychological emotion, not just a social situation. Comparable to other emotional states, shame produces involuntary stereotypical physiological effects such as slumped shoulders, an evasive gaze, and sometimes blushing.

Adapting a definition from the current international relations literature, I define shaming in the context of human rights advocacy as emotionally charged public criticism that denounces or humiliates human rights violators and their abettors ‘in a way that targets the essence of an individual's identity’ (Markwica Reference Markwica2018, 82). This narrow definition is consistent with the social psychological literature that distinguishes between guilt, which focuses on responsibility for a bad action, and shame, which implies a defective personal trait that may be difficult to remediate. The parallel distinction at the group level distinguishes between routine social practices with low cultural salience as opposed to expressions of culture that are salient to the group's fundamental identity (Boddy Reference Boddy1982; Cloward Reference Cloward2016, 19–20, 228–51).

Shaming normally involves verbal characterizations of behavior as ‘shameful’ or ‘inhumane’, but simply naming violations for which amnesty is legally forbidden (genocide, war crimes, and crimes against humanity) can be considered inherently shaming. Some human rights advocates and scholars routinely apply the term ‘naming and shaming’ to a much broader spectrum of sins and pecadilloes, ranging from mass killing to the failure to purchase fair trade coffee, but here I am not referring to the more mundane applications of this terminology. While Neier and Brysk are correct about the centrality of shaming to much rights advocacy, contemporary human rights work includes activities that may not include shaming of any kind, such as the provision of health, humanitarian, or economic development services using a method that conforms to the standards of the ‘human rights approach’ (Barnett Reference Barnett2005).

Shaming, in my narrow definition, exists on the harsh end of a rhetorical continuum that includes, on the less harsh end, factual reporting, recommending constructive remedies, and numerical rating and ranking. The middle of the spectrum is occupied by pointedly criticizing policy, identifying legal and normative violations (‘naming’), and pressuring states and other actors to impose punishment for specific misbehavior (‘ending impunity’). To illustrate the various kinds of naming, shaming, and other criticism that characterize contemporary human rights discourse, and to assess the prevalence of shaming in that litany, I ‘randomly sampled’ Human Rights Watch's list of its ten most recent reports as of March 2019.Footnote ² Their topics addressed (1) Japan's recent ‘regressive’, ‘harmful’ legislation on transgender status, quoting an interviewee who accused the law of ‘wrecking people's dignity as a human being’; (2) ‘segregation’ and ‘discrimination’ against disabled children in Kazakhstan notwithstanding reforms that were ‘too slow’; (3) torture, stigmatization, and forced confessions of children of ISIS fighters held in Iraq; (4) governmental ‘valorization’ of ‘shameful’ ‘vigilante groups’ of ‘cow protectors’ in India whose ‘cruelty and loathing’, according to a quoted local activist, ‘has penetrated the souls of young people’; (5) Polish ‘government raids’ of pro-abortion women's organizations and associated policies of the Catholic church that ‘demonize women’ in a way that ‘fosters a climate of fear’; (6) unprosecuted acid attacks to disfigure Cambodian women; (7) ‘abusive laws’ passed by Myanmar's elected Parliament targeting free speech and journalism; (8) ‘abusive’ labor practices leading to ‘tragic’ deaths in the context of Pakistan's ‘conservative society’; (9) an European Union anti-migration policy leading to ‘cruel, inhuman, degrading treatment’ of detainees in Libya; and (10) UK government cost-cutting pressures that led to a failure to fund legal entitlements to health care for the elderly.

All of these reports reflected a high level of professionalism, and there is no reason doubt their accuracy. All 10 of these reports featured factual reporting; naming of violations of law, ethical norms, or professional standards of best practice; criticism of government actions or of the failure of government to take steps against abuses; statements clearly intended to exert pressure on responsible parties; and concrete recommendations for remediation. They interviewed locals, and whenever possible expressed the harshest criticism in the words of local sources. They tried not to lead with international law and global moral standards, but these were sometimes invoked.

Criticism was aimed at top-level policy makers in every case. In seven instances criticism also targeted private actors, religious authorities, or working-level government functionaries. In two or three instances, the targets were ordinary private individuals engaging in abnormal but socially tolerated and generally unpunished abusive behavior, such as acid attacks on women, vigilante attacks on minorities, and neglect of the disabled. In six to eight of these reports, criticism of the government rose to the level of shaming. These included attempts to humiliate explicitly (‘shameful’, ‘inhuman’) or implicitly given the nature of the charge (ordering or condoning torture). In two cases private individuals or mass groups were targeted for denunciation and humiliation (the acid attackers and the vigilantes).

Especially noteworthy is that in all 10 cases the abusive policies or negligence of the government reflected underlying widespread mass attitudes (gender-based bias, neglect of the disabled, prejudice toward an outgroup, religion-based opposition to women's health rights, toleration of child labor practices, opposition to immigration, and popular support for policies enacted by elected legislators). Although it is the policy of Human Rights Watch to name and shame state authorities and individual violators rather than entire nations, peoples, religions, or cultures, in these cases individuals who see their self-esteem and status as linked to that of their national culture may have reason to react as if they themselves were shamed. The strict definitional standard of shaming that targets culturally salient features of a group's fundamental identity is arguably met in the reports on ‘soul-penetrating cruelty’ of Indian cow vigilantes, Polish Catholic church policies that demonize women, and inhuman behavior of European opponents of immigration, and possibly the reports on Japanese transgender law and labor abuses in conservative Pakistan.

An increasingly prominent form of human rights criticism is the numerical rating and ranking of government performance, such as Freedom House's ratings of media freedom and civil liberties or Transparency International's ratings of corruption (Merry et al. Reference Merry, Davis and Kingsbury2015; Morse Reference Morse2016). While numerical ratings and rankings may be seen as less emotionally charged than verbally shaming reports, and therefore less likely to trigger defensive backlash, political scientists Judith Kelley and Beth Simmons note that these indicators are not mainly intended for diagnostic purposes: ‘Recent GPIs are relentlessly comparative, suggesting an intention to pressure, shame or provoke competition among states’ (Kelley and Simmons Reference Kelley, Simmons, Kelley and Simmons2020, 4). Compressing complex causality in order to create simple category labels that have shaming power, rankings produce leverage through invidious comparisons with low-status or rival states. The empirical chapters of their volume find strong evidence that state actors are successfully shamed or at least highly annoyed by the fact that they are being called out. Their evidence is less clear that rankings per se succeed in stimulating strong social pressure on the noncompliant state.Footnote ³ When improved compliance is observed, ranking interacts with necessary facilitating conditions – for example, the country is a democracy, the message resonates with the values of the country's citizens, and the state has the resources to comply (Kelley and Simmons Reference Kelley, Simmons, Kelley and Simmons2020, 16).

These qualifications echo the well-established finding in mainstream human rights scholarship acknowledging that a long list of structural scope conditions constrains the effectiveness of standard advocacy tactics, including shaming.Footnote ⁴ Many studies in the human rights and democracy promotion fields present evidence that shaming works best in easy cases where civil society resistance to the repressive state is already strong (Murdie and Davis Reference Murdie and Davis2012), where the state is less able to hide compensatory repressive moves (Hafner-Burton Reference Hafner-Burton2008), where the state has signed a treaty consenting to an obligation (Clark Reference Clark, Risse, Ropp and Sikkink2013), or where the state has fewer opportunities to engage in ‘counter-norming’ – for example, invoking sovereignty and illiberal cultural traditions; denouncing the decadence of liberal sex and gender norms (Cooley and Schaaf Reference Cooley, Schaaf, Hopgood, Snyder and Vinjamuri2017).Footnote ⁵ A major shortcoming of the theoretical anchors of shaming tactics, especially evident in the present climate of illiberal backlash, is the lack of a mechanism explaining why liberal human rights advocacy should prevail in shaming contests with illiberal social movements that oppose abortion, refugees, international criminal accountability, rights for women and gays, and the principle of universalism. These illiberal movements argue with considerable impact that it is the liberal view that is shameful when assessed from the standpoint of religious, communal, or national values.

Psychological and sociological theories of the emotion of shame

Some social psychological theorizing and empirical research make big claims for shame as the central social emotion, the crucial glue that holds social relations together (Scheff and Retzinger Reference Scheff and Retzinger1991; Scheff Reference Scheff2000). If so, this would suggest that human rights shamers may be on the right track in their quest for a powerful lever to change attitudes and behavior. However, precisely because of shame's apparent power, many – perhaps most – social psychologists focus on shame's destructive potential and the likelihood that shaming, if carried out ineptly, will produce effects that are the opposite of those intended.

A prominent theme of this literature is that shame and shaming play a necessary role in deterring violations of social norms and in the formation of an individuals' conscience, yet shaming can also backfire. Depending on the circumstances, the target, and the technique, shaming can lead to a self-reinforcing cycle of humiliation, anger, hatred, social withdrawal, and attachment to a counterculture of proud deviance. An alternative undesirable path may lead from shame to despondency and evasion rather than compliance. Shaming is least likely to misfire when it targets discrete behaviors that can be altered rather than inherent character traits, when it comes from inside a social identity group rather than from outsiders, when it avoids denouncing mass cultural attitudes and practices, and when it deftly pressures the abettors and associates of norm violators, not just perpetrators themselves.Footnote ⁶

Most scholars agree that shame and shaming need to be understood at the individual and the social levels simultaneously. That said, it makes sense to begin at the individual level, in part because a predominant group level theory, SIT, is anchored to a substantial degree in individual-level issues of self-esteem.

Some early psychological theories, including offshoots from orthodox Freudianism, placed shame-related concepts at the center of their theories. Alfred Adler based his theory of the ‘inferiority complex’ on prestige and self-esteem, with long-term low self-esteem being tantamount to ‘chronic shame’ (Adler Reference Adler1956; Scheff Reference Scheff2000, 86). Karen Horney, in a move that is closely echoed by several contemporary scholars, posited a ‘pride system’ featuring sequences of shame and anger that when sustained lead to a ‘vindictive personality’ (Horney Reference Horney1950). More positively, Abraham Kardiner argued that shame was central to the workings of the ‘superego’ or conscience. Erik Erikson saw shame and guilt as one of the developmental stages characteristic of youth. Charles Cooley wrote about ‘the looking glass self’, implying that the dimension of pride and shame constituted the basic social emotions (Cooley Reference Cooley1902, 183–84; Scheff Reference Scheff2000, 88). An updated version of this theory, recast as ‘self-monitoring’, posited a stable personality trait that captures the individual's level of concern with the impression being made on others. This characteristic has recently been shown to successfully predict which US Presidents have been willing to fight for reputation (Yarhi-Milo Reference Yarhi-Milo2018).

Contemporary social psychologists agree that a productive turn came with Helen Lewis's Shame and Guilt in Neurosis (Reference Lewis1971), which simultaneously addressed the social context of shame, the feeling of shame, and people's cognitive understanding of those feelings. She wrote about ‘feeling traps’ in which people not only feel shame for their shortcomings, but may also feel shame for feeling ashamed and for feeling angry about being shamed by others, leading to a self-perpetuating cycle. Her empirical work with transcripts of therapeutic interviews fleshed out earlier theorists' conjectures about shame/anger sequences and placed them in a social context (Scheff Reference Scheff2000, 95). She argued that shame can either provoke anger toward the source of the shaming, leading to resentment, or anger can turn back on the self, leading to feelings of guilt. In both of these variants, Lewis highlighted the pathologies that result when the feeling of shame is unacknowledged, repressed, and redirected into festering guilt, ‘resentment’, ‘hostility’, ‘righteous indignation’, and ‘humiliated fury’ (Lewis Reference Lewis1971, 37, 44–45, 248–49, 323; Scheff Reference Scheff2000, 94–95).

Later, a distinctive line of research based on evolutionary theory came to conclusions similar to Lewis's. It proceeded from the assumption that ‘rank’ and ‘status’ are crucial for the ability to attract material investments and sexual partners. In this context, shame and stigma are attached to failures in performing four key social roles: sexual behavior (based on deviance, exploitation, and unattractiveness), prosocial behavior (based on fulfilling obligations), conformity (to rules, fashion, and tradition), and resource competition (Greenwald and Harder Reference Greenwald, Harder and Gilbert1998). According to this theory, ‘external’ manifestations of shame and shaming are important for status competition, and this can also lead to ‘internal shame’. Although in this view the feeling of shame may have net evolutionary benefits, specific manifestations of shame can be highly dysfunctional (Sznycer et al. Reference Sznycer, Tooby, Cosmides, Porat, Shalvi and Halperin2016). Shame and shaming may lead to destructive defensive emotions, debilitating anxiety, concealment, stifling conformity, and resort to malicious accusations to take competitors down a peg in the status order. Thus, ‘in a shame system people can behave very immorally’ (Gilbert Reference Gilbert2003, 1213, 1225). In short, social psychologists, whether devotees of harmonious social integration or partisans of Darwinian cutthroat social competition, have converged on the finding that shame and shaming can easily degenerate into dysfunction.

Effects of shame and guilt

Whereas early academic writing on shame was, like vernacular discourse, loose in distinguishing it from guilt, more recent social psychological literature typically sees guilt as focused on a bad action, whereas shame is focused on a flaw in the person (Scheff and Retzinger Reference Scheff and Retzinger1991, 65). Thus, the OED defines guilt as ‘a feeling of having committed wrong or failed in an obligation’. This literature (as well as a big stack of self-help books) argues that feeling or instilling guilt is better for the person and for social relationships than feeling or instilling shame. With guilt, you fix your behavior, pay your debt for past mistakes, and you are good to go. With shame, you (and observers) may feel there is something inherently wrong with you, something that may be impossible to fix (Lewis Reference Lewis1971, 30, 37, 40). For that reason, shaming can easily degenerate into stigmatization by society and lead either to demoralization or backlash on the part of the shamed (Tangney and Dearing Reference Tangney and Dearing2002; Markwica Reference Markwica2018, 19). The prescription for the activist seems clear: when possible, instill a sense of remediable guilt for the behavior, not irremediable shame.

What can be done when the problem isn't just an isolated bad action, and thus amenable to management in the less volatile guilt frame, but an embedded practice, outlook, or character trait, which gives rise to feelings of shame? In principle, good outcomes from shaming can result either from external deterrent constraints, as when authorities or collaborators are shamed into imposing costs on those who misbehave, or from internalized socialization to a prevailing norm, as when a young person or new group member is shamed into conforming appropriately to an established social role (Braithwaite Reference Braithwaite1989, 69–77). But much social psychological research confirms Lewis's findings about shame mechanisms that produce negative effects (Every Reference Every2013). Since ‘what counts as shameful is … contestable’, shaming can become a focal point for indignant resistance that justifies counter-normative behavior (Every Reference Every2013, 670). Psychologically suppressing or evading acknowledgement of shame may ‘displace’ shame into such forms of hostility as ‘bullying’ (Harris Reference Harris and Drahos2017, 69–70). Even when shame and shaming do not provoke violent backlash or scapegoating, they still might not induce conformity with desirable norms. Instead, acceptance and internalization of the legitimacy of the shaming can lead to feelings of hopelessness. For example, interviews report that undereducated, impoverished, rural Wisconsin voters realize that Trump won't really help them, but at least he acknowledges their frustrations with declining status and their resentment toward privileged Madison urbanites who they believe hold them in contempt (Kramer Reference Kramer2016; Wallis Reference Wallis2016).

Feeling ashamed and being targeted for shaming have been found to correlate with low empathy, social isolation, non-constructive violence, and self-destructive behaviors (Tangney and Dearing Reference Tangney and Dearing2002). Indeed, sometimes all of these pathologies may emerge and feed on each other. As a result, shaming can be a powerful tool of oppression, as in the shaming of gays and of women whose bodies fail to meet ideal standards, yet also a goad to destructive backlash (Brown Reference Brown2008).

So how can the potentially beneficial effects of shame and shaming be tapped without triggering these negative consequences? An important strand of the literature emphasizes the possibility of ‘shame management’, avoiding unintended outcomes from shaming through practices that lead to the reintegration of the shamed person into society (Braithwaite Reference Braithwaite1989; Ahmed et al. Reference Ahmed, Braithwaite, Harris and Braithwaite2001). This school of thought, pioneered by John Braithwaite, argues that shaming should be ‘respectful’, distinguish the person from the person's behavior (as with guilt), and culminate in a ceremony that reintegrates the person or group into society. When possible, he argues, shaming should consist of generic reminders of behavioral standards, which often can be more effective if the violators are not singled out for humiliation (Braithwaite Reference Braithwaite1989, 83). At all costs, shaming should avoid the kind of stigmatization that makes ‘deviance’ a ‘master status trait’ of the shamed, which leads to ‘greater offending’, blocking of opportunities to participate in legitimate social activities, perceptions of injustice, and subcultures supporting deviance (Braithwaite Reference Braithwaite1989, 65–68; Harris Reference Harris and Drahos2017, 59–61).

In one of the few substantial studies applying social psychological theories of shame and guilt to human rights concerns, Mark Drumbl draws extensively on Braithwaite's concept of reintegrative shaming to argue in favor of the traditional Rwandan local community justice process of gacaca as a more effective alternative to the formal post-genocide criminal trials mounted by the International Criminal Tribunal for Rwanda, centered on guilt and punishment (Drumbl Reference Drumbl2002, 1256). This conceptually sophisticated argument, however, runs counter to a substantial body of empirical research that documents the manipulation of the gacaca process by Paul Kagame's authoritarian regime to selectively control potential Hutu opposition (Ingelaere Reference Ingelaere, Huyse and Salter2008; Clark Reference Clark2010).

In devising strategies to reduce the risk that shaming and guilting will reinforce backlash and deviance, one issue is whether to put a bright spotlight on deviant behavior or to allow the violator a fig leaf to ease social reintegration. Lewis contended that openly confronting shame is needed to avoid dire social and psychological pathology. Erving Goffman's The Presentation of Self in Everyday Life (Reference Goffman1959) reflects the opposite sensibility. Goffman and his followers analyze embarrassing situations in which social role expectations involving ‘rights and duties attached to a given status’ are disrupted. This occurs when a person faces mixed audiences of different status or degrees of intimacy, placing the person under incompatible role expectations for appropriate behavior (Goffman Reference Goffman1959, 13, 16, 49; Lizardo and Collett Reference Lizardo and Collett2013). As Cooley had noted, ‘a man will boast to one person of an action – say some sharp transaction in trade – which he would be ashamed to own to another’ (Cooley Reference Cooley1902, 184–85). Goffman explores the subtle role of tact in navigating such contexts of disrupted expectations and cross-pressures, including tacit collusion by all parties to manage the resulting damage to participants' social images (Goffman Reference Goffman1959, 14, 234). For Goffman the tactful agreement not to acknowledge the elephant in the room preserves the actors' ‘performance’ of their roles and the social order.

Thinking about government performance indicators as levers of shame, a follower of Goffman might appreciate the tactful subtlety of rating a state's performance relative only to its own past without the embarrassment of an explicit ranking relative to other states, especially compared to intra-regional rivals. In contrast, a follower of Lewis might see public comparative rankings of rights compliance as a much-needed reckoning with the truth (Cooley Reference Cooley, Cooley and Snyder2015, 13–14). Perhaps their different emphasis reflects the fact that Lewis was considering cases of clinical neurosis whereas Goffman was concerned with normal people navigating awkward social circumstances. How to anticipate and manage the tradeoff between tact and forthrightness in varied social settings is a challenge for shaming strategies.

Shame and shaming in the context of group psychology

Scholarship on the social psychology of shame and pride generally accepts that these emotions should be studied at both the individual and group levels, which are seen as interactive. Exactly how to merge insights at these two levels has been a longstanding subject of scholarly discussion. Tajfel's SIT, widely invoked in social psychology as well as political science, has been at the center of many of the debates around this issue (Tajfel, Reference Tajfel1981, 41–53). Neither SIT nor its theoretical variants support the idea that group psychology is likely to be conducive to shaming strategies by outgroup norm innovators.

SIT, says Tajfel's chief collaborator, John C. Turner, ‘began as a way of making sense of discrimination between groups and its fundamental psychological idea was that where people make comparisons between groups, they seek positive distinctiveness for their ingroups compared to outgroups in order to achieve a positive social identity’ (Tajfel and Turner Reference Tajfel, Turner, Austin and Worchel1979, 40–41; Turner Reference Turner and Haslam2004, xix). In other words, individuals derive self-esteem from the status of their group. As Steven Ward puts it in an application to international relations, members of a respected group feel ‘pride’, whereas members of disrespected groups feel ‘shame’, possibly leading to anger and frustration.Footnote ⁷ In light of what individual-level research suggests about adverse reactions to shame, this basic hypothesis from SIT would not seem promising grounds for outsiders to pursue a strategy of human rights shaming.

Tajfel was reacting to Muzafer Sherif's earlier landmark study of group discrimination and conflict which was grounded in ‘realistic conflict theory’ (Sherif et al. Reference Sherif, Harvey, White, Hood and Sherif1954, Reference Sherif, Harvey, White, Hood and Sherif1961). Sherif's Robbers Cave Experiment randomly assigned 12-year-old, middle-class, well-behaved boys to two groups, which were pitted against each other in sports and contests, the outcome of which determined the allocation of points and prizes. Competition quickly escalated to negative stereotyping, discrimination, badmouthing, theft, and destruction of property.

Tajfel's experiments showed that discrimination and ingroup bias could emerge even without stacking the deck with built-in resource competition or task-related reasons for ingroup solidarity. People were randomly assigned to groups that were given no task and no reason to cooperate. These ‘minimal group experiments’ showed that people would allocate resources preferentially to ingroup members even when this discrimination would reduce the overall pool of resources for their own group (Tajfel et al. Reference Tajfel, Billig, Bundy and Flament1971).

But there is an irremediable problem in this research design. Although the artificially formed groups were new and meaningless, the participants were not new to being in groups in general. Like humans everywhere, the participants had spent their whole lives being socialized to the norms and functional benefits of group loyalty. It is impossible to rule out that these habits, rather than some innate need to derive self-esteem from group status, explain ingroup bias in these minimal groups (Tajfel Reference Tajfel1981, 236). This loophole opens up SIT to alternative interpretations.

Without renouncing SIT's basic self-esteem account for ingroup bias, Turner (Reference Turner, Ellemers, Spears and Doosje1999, 7–8) complains that many applications of SIT have treated self-esteem-driven discrimination in minimal groups as ‘the end of the story’. He worries about the implication that there is nothing that can be done to counter the universal, ‘automatic’ tendency of group members to bias against outgroups as a way of shoring up their personal pride. He argues that the minimal group condition is not representative of the situation of real groups, but a decontextualized circumstance in which group members can achieve ‘positive distinctiveness’ only through denigrating the outgroup (Mols and Weber Reference Mols and Weber2013, 507). In real situations, according to Tajfel and Turner (Reference Tajfel, Turner, Austin and Worchel1979, 41), the extent of bias in intergroup comparisons depends on variable factors, such as the extent to which individuals identify with an ingroup and internalize group membership as part of their self-concept. Also important is whether a particular outgroup and a particular dimension of comparison is considered relevant to the group's status. Following this line of conjecture, other scholars have suggested that social identity and self-esteem need not rest on a zero-sum status competition if different groups value different accomplishments (Brewer Reference Brewer and Ashmore2001; Larson and Shevchenko Reference Larson and Shevchenko2019).

Turner later supplemented SIT with his rather differently grounded ‘self-categorization’ theory (Turner Reference Turner1991, 143–73). Its most distinctive claim was that ingroup bias stems from people's need for socially useful ‘reality testing’, which requires convergence of opinions, attitudes, and factual assumptions of ingroup members for the practical purpose of coordinating views of reality with people in their own social sphere. Turner generalized this insight further as ‘social influence theory’, which invokes such mechanisms as socialization, the drive for conformity, the tendency toward social mimicry, and processes of cognitive framing and priming in coordinating ingroup members' convergent social identity (Turner Reference Turner1991; Johnston Reference Johnston2001, 494). Tajfel, however, warns in passing against overstating the force of socialization to ingroup favoritism, since cultures may also socialize people to a generalized norm of ‘fairness’ (Tajfel Reference Tajfel1981, 270–71).

Apart from Tajfel's offhand remark about generalized fairness, none of these group psychology insights seem at all promising as a basis for human rights shaming, especially by outsiders. Basic SIT seems especially ill-suited, since it combines prejudice against outgroups with a neuralgic sensitivity to issues bearing on group status and individual self-esteem. The self-categorization and social influence approaches are little better insofar as they presume that each community tends to create its own consensus version of the truth, which is bolstered by a raft of reinforcement mechanisms. At best, this could make some ingroups indifferent rather than hostile to assertive social influence attempts by outsiders, but it hardly supports the effectiveness of shaming groups who are listening to their own drummer.

More ambiguous is the case in which a minority is socialized to its own norms and worldview, and faces discrimination and stigmatization by a dominant majority identity group. This might make the oppressed minority highly receptive to a third party's universalizing rhetoric that shames the oppressive majority. However, if the aim is to convince the majority to stop its abusive practices, SIT would seem to recommend persuasion from inside the majority community based not on shaming but on its own vernacular concepts of decency and justice (Levitt and Merry Reference Levitt and Merry2009).

Conditions when shaming might work

While the core arguments of these theories warrant wariness of adopting shaming strategies, some social psychologists have considered the conditions in which the beneficial effects of shaming might exceed its risks (Braithwaite and Drahos Reference Braithwaite and Drahos2002, 269–88). Shaming seems more likely to succeed when it is aimed at individuals, not the group in general; when shaming is respectful and is directed toward the goal of reintegration into the group; when an insider or a highly respected outsider is doing the shaming; and when the targets of influence are weak identifiers with the ingroup and have an aspirational identification with a high-status outgroup.

Shaming works best when it comes from a respected source (Braithwaite and Drahos Reference Braithwaite and Drahos2002, 273). ‘Only groups whose approval an actor values will have this influence’ (Johnston Reference Johnston2008, 80; also Mantilla Reference Mantilla2018, 324). A particularly effective shaming move can occur when the shamer and shamed share a social identity, the shamer points out an inconsistency between their shared identity norms and the shamed actor's deeds, and the shamer can credibly claim that the target's misbehavior is making the ingroup as a whole look bad in the eyes of outsiders. Chinese nationalists were very effective in playing this card against Chinese who continued to practice foot-binding, which made Chinese culture look backward and barbaric in the eyes of the world (Keck and Sikkink Reference Keck and Sikkink1998, 39–40, 59–66, 73–74). Two causal mechanisms made this work. Christian missionaries put pressure for reform on Chinese who weakly identified with traditional practices, and aspiring Chinese nationalists identified with the technological power of the modern West, if not with its other cultural attributes (Appiah Reference Appiah2010, 53–100).

Research that is mainly on guilt rather than shame finds that people who identify more weakly with their ingroup are more able to accept outsiders' criticism that undermines their social identity. In Germany, high and low identifiers differ in how willingly they acknowledge the Holocaust.Footnote ⁸ However, guilting or shaming weak identifiers may mean hitting exactly the wrong targets. Galia Press-Barnathan (Reference Press-Barnathan2018, Reference Press-Barnathan2019) argues that boycotting Israeli events, people, or products is painful for cosmopolitan Israelis, but is shrugged off by high identifying nationalists. Assuming that policymakers are usually high identifiers, this would suggest that it is hard to guilt and perhaps to shame them (Branscombe and Doosje Reference Branscombe and Bertjan2004; Tarrant et al. Reference Tarrant, Branscombe, Warner and Weston2012, 513–18).

Why these conditions are not the norm

Notwithstanding the circumstances in which shaming from an outgroup might succeed, these seem more the exception than the rule. A key goal of counter-norming is to disarm the shaming capacity of outgroups (Schweller and Pu Reference Schweller and Pu2011; Cooley and Schaaf Reference Cooley, Schaaf, Hopgood, Snyder and Vinjamuri2017). When cosmopolitan shamers from outside a culture are trying to mobilize individuals' emotions against rights abuses, it is an uphill battle. Jonathan Mercer argues that ‘group emotion is often stronger than individual emotion’, especially among people who strongly identify with the group, as result of the group's validating ‘emotional consensus’ and its policing of those whose emotional responses fail to conform to the expected group norm (Mercer Reference Mercer2014, 515–26).

Group leaders may exploit the contest between outside shaming and ingroup loyalty to boost their legitimacy in domestic politics (Ward Reference Ward2017, 38, 55–56). Although human rights advocacy normally tries to aim shaming at individual perpetrators or responsible state officials, looser parlance often collectivizes shaming discourse, for example generalizing about ‘lazy Greeks’ and ‘overbearing Germans’ (Adler-Nissen Reference Adler-Nissen2014, 143–76; Adler-Nissen Reference Adler-Nissen and Zarakol2017, 198–218). Especially when such characterizations are seen as unjust or hypocritical, shaming provokes a collective emotion of ‘popular outrage’ (Hall Reference Hall2017, 1–29; Ward Reference Ward2017, 50–51). Whether the target of shaming is an individual or a collectivity, ‘insofar as shaming promotes anger, humiliation and denial rather than empathy, guilt, and responsibility, shaming may harden rather than resolve the problem of human rights violations’ (Lickel et al. Reference Lickel, Schmader, Barquissau, Branscombe and Doosje2004, 52).

In nationalist rivalries, the dynamic of shaming, humiliation, and grudge-holding is common and counterproductive. For example, Peter Gries documents the Chinese obsession with past humiliations by the imperial powers dating back to the Opium Wars and the Japanese occupation. Jennifer Lind shows in turn how endless Chinese demands for ever better Japanese apologies serve to fuel mass Japanese nationalism. Dmitri Trenin interprets contemporary Russian grudges against the West as a comparable reaction to the humiliations suffered at the hands of Western states and liberal reformers after the Cold War (Gries Reference Gries2004; Lind Reference Lind2008; Trenin Reference Trenin2016; Gruffydd-Jones Reference Gruffydd-Jones2019).

In an era of populist politics, outgroups can include not only foreigners but also suspect social classes within a society, especially elite cosmopolitans that are seen as traitors to the ‘true people’. Just as higher status groups may engage in class struggle by shaming ‘deplorable’ lower status groups over their violations of elite social norms, conversely the impolite politics of populism can be a weapon of the weak in this struggle to delegitimize elite-dominated politics as usual (Scott Reference Scott1985; Moffitt Reference Moffitt2016; Mantilla Reference Mantilla2018, 326). The shaming tactics of such culture wars feed populist politics, which often link together the struggles against foreign, immigrant, and domestic elite enemies of the ingroup.

Shaming the masses

An especially important distinction addresses the effectiveness and risks of shaming non-elite or disadvantaged people in a society. The social psychological literature focuses heavily on feelings of shame among weak, powerless, downtrodden, and aggrieved groups that see themselves as not being paid their due respect. For example, the literature on the American Christian Right has for the past two decades described the mentality of those suffering deindustrialization and the contempt of cultural elites, which has fueled both social disintegration and populist backlash in the culture wars over race, immigration, gender, and sexual rights (Stein Reference Stein and Goodwin2001). Similar points have been made about the Australian right-wing populist anti-migration movement spearheaded by Pauline Hanson (Every Reference Every2013).

Nor does shaming mass cultural practices have a good track record in developing countries. Keck and Sikkink's seminal Activists Beyond Borders, which explicated the logic of norms-based transnational information politics, includes a telling chapter on the failure of British churches' efforts in the 1920s and ’30s to shame its Kenyan parishioners into abandoning the practice of female genital cutting. These shaming tactics played into the hands of independence leader Jomo Kenyatta, who exploited the issue as a prime example of British cultural imperialism much as his son Uhuru exploits the ICC today (Keck and Sikkink Reference Keck and Sikkink1998, 66–72). In the Kabare region of Kenya where British clergy employed a deliberative approach to the issue, the rate of cutting eventually went down to 35% by the end of the millennium, but in the Kigare region where shaming and excommunication were used, the rate remained around 60% (Boyle Reference Boyle2002, 136–37).

When NGOs initially tried to combat genital cutting in the 1990s with blunt shaming tactics based on legalism and universalistic values, they ran into stiff resistance on grounds of religion and custom (Rajdurai and Igras Reference Rajdurai and Igras2004). Somewhat more effective were arguments offering alternative perspectives from insiders to the locals' faith community, combined with health information, the provision of health services, and in a parallel to a tactic of the anti-footbinding campaign, community pledging not to cut and not to marry girls who were cut (Mackie Reference Mackie1996). Behind these persuasion tactics, however, are structural facilitators. Elizabeth Boyle (Reference Boyle2002, 120, 132) found that the strongest correlates of the intention not to cut were mother's income outside the home and access to world communications media. But widespread awareness of anti-cutting campaigns could produce opposite effects depending on the salience of cutting in the local ethnic identity and on levels of urbanization and education (Cloward Reference Cloward2016; Shell-Duncan et al. Reference Shell-Duncan, Naik and Feldman-Jacobs2016). Ylva Hernlund's field research in Gambia reported that the ‘local and national debate’ was ‘becoming more polarized and acrimonious’, with the reaction to international pressures ranging from ‘relief that outside help is speeding up the elimination of genital cutting to rage at what is perceived as imperialist meddling’ (Hernlund Reference Hernlund, Shell-Duncan and Hernlund2000, 242). More recent field interviews report that ‘in some communities deep resentments over the ‘criminalization of culture’ simmer just below the surface and boiled over when the subject was raised’ (Shell-Duncan et al. Reference Shell-Duncan, Wander, Hernlund and Moreau2013, 831).

The double whammy: outsiders vs. insiders shaming masses

Despite the dangerous potential for popular backlash, shaming mass groups who retain illiberal attitudes and practices such as early marriage, gender and sexual inequality, child labor, and exploitation of migrant labor is part of the routine work of international human rights advocacy. Not surprisingly, shaming works better when the target already accepts the validity of the values being invoked. In statistical research using standard human rights measurements, Amanda Murdie finds, for example, that international activists' shaming of violations of physical integrity rights, such as torture, has a more positive impact in target states than does shaming over women's rights, because there is a greater divergence between shamers and targets over what women's rights should be (Murdie Reference Murdie2014, 193).

It is difficult for cultural outsiders to prevail in normative contests with locals when both are trying to use powerful emotions such as shame and humiliation to mobilize mass social movements (or ‘civil society’) to support their cause. For example, Irena Sargsyan and Andrew Bennett study Muqtada al-Sadr's unexpected success in raising and sustaining an illiberal, militant Shi'a mass movement in post-invasion Iraq. They find that ‘leaders who maintain legitimacy among the local population, connections to indigenous social or religious networks, and a keen understanding of potential supporters’ collective identity and memory are more effective in mobilizing followers than their counterparts who lack one or more of these qualities’, including in this case Western states and humanitarians, foreign fighters, and returning émigré politicians (Sargsyan and Bennett Reference Sargsyan and Bennett2016, 609). They show how local militants like Sadr employ discursive ‘frames that evoke powerful emotions – anger, humiliation, fear, shared experiences of suffering or injustice, and a desire for honor through self-sacrifice … to activate collective emotions and channel them into sustained violence’. ‘Often stressing the threat of impending violence by the outgroup against the ingroup’, such militant locals ‘invoke shared sacred texts or metamorality endorsing violence and self-defense as legitimate in the circumstances; make a plausible claim that collective violence against the outgroup can forestall aggression; are continuously updated to maintain relevance; and are not rebutted by credible counterframes from opponents’ (Sargsyan and Bennett Reference Sargsyan and Bennett2016, 609–10, 618). In contrast, Western efforts to shame Saddam Hussein for gassing Kurds, Sunni rebels for terror attacks, or Shi'a militias for ethnic cleansing got little traction with their core supporters. Rare US successes, such as the recruitment of Anbar province Sunni tribal fighters to give up insurgency and collaborate with the Iraqi army, came from informative dialog and material incentives, not shaming (Lynch, Reference Lynch2011; Biddle et al. Reference Biddle, Friedman and Shapiro2012).

Shaming elites and their enablers

But what about shaming elites? After all, the mainstream human rights movement and liberal rights ideology more generally prefer to think of powerful elites, especially the oppressive state, as the source of most rights abuse. Even when the abuse is a widespread cultural practice such as child labor or early marriage, rights NGOs prefer to shame state authorities for not doing enough to curtail the practice. Can hard-boiled elites be shamed without worrying about triggering a neurotic backlash?

Some social psychologists argue that elites tend to be too well insulated to be shamed effectively. Politicians are psychologically ‘hardened’ against criticism, which is routine for them. Public relations consultants and corporate image strategists protect economic elites. They are wrapped in a cocoon of free market ideology according to which the invisible hand determines their policies on labor and the environment. Moreover, they are good at segregating their audiences, giving different performances for supportive peers (viz., Mitt Romney's commiseration at the campaign donors' dinner over the shameful 47% of Americans who are freeloading on government handouts) and for the broader public, which sees them going to church on Sunday and mouthing platitudes (Google's ‘don't be evil’).

Nonetheless, there may be chinks in elites' public relations armor of shamelessness. These mechanisms are likely to work not by shaming the elite perpetrators but by shaming their family members, peers, or customers, who can put them under pressure to reform (Braithwaite and Drahos Reference Braithwaite and Drahos2002, 273–74). Perpetrators may fear public reputational consequences (purely external shame) or disapproval from an intimate or a respected peer group (an interaction of external and internalized shame). In a final twist, the peer group (e.g. other potential exploiters of child labor in the same industry) might exert pressure to cave in to a boycott to avoid reputational losses for the whole industry or to prevent the violator from getting a competitive advantage by hiring illegal cheap labor (Ruggie Reference Ruggie2013). Apple, taking advantage of its business model that does not rely on click bait and fake news, has shamed big tech competitors Facebook and Google that do.

Whether the elite targets of shaming are states or businesses, it is not simple to disentangle emotion from cunning in their reactions. For example, a 2001 assessment of the impact of shaming China for its human rights record describes a mix of Chinese ‘indignant white papers’, defensive cultural relativism, and generically evasive, pro forma promises of compliance (Wachman Reference Wachman2001). This litany fits nicely with the typical patterns found in research on the social psychology of the emotion of shame, but it could also fit the logic of a calculating public relations strategy.

Unintended consequences of shaming and their lock-in

Shamers intend to promote compliance with norms they value, but shaming often triggers unintended consequences, including shame/anger/resentment sequences and denial/hiding/deviance syndromes. In the arena of human rights, recent political science literature documents various mechanisms leading to unintended consequences. Jacob Ausderan (Reference Ausderan2014) notes the tendency of shaming by prominent outsiders to rile up in-country constituencies who become newly aware of violations and who overestimate the likelihood of decisive help from outsiders in rectifying these wrongs. He notes that this can heighten government fears of losing control and lead to a crackdown on dissent with an attendant further rise in human rights abuse. For the same reasons, rising press freedom in authoritarian states has been found to increase repressive human rights abuse (Whitten-Woodring Reference Whitten-Woodring2009). In a related pattern, powerful outsiders loudly proclaiming support for abused minorities create a moral hazard when the oppressed are misled into believing that their resistance will be backed by foreign military intervention (Crawford and Kuperman Reference Crawford and Kuperman2006). In accord with the denial/hiding pattern, shaming has been found to lead to ‘whack-a-mole’ shifts in the means and targets of repressive rights abuse (Hafner-Burton Reference Hafner-Burton2008).

An important question is whether adverse consequences of shaming are not only unintended and undesired, but whether they are unanticipated, avoidable, counterproductive, and long lasting (Vinjamuri Reference Vinjamuri, Hopgood, Snyder and Vinjamuri2017; Stein Reference Stein, Huddy, Sears and Levy2003, 382). Human rights activists do not desire or intend pushback from perpetrators, but they do anticipate it under many circumstances. The influential ‘spiral model’ foresees a temporary stage of denial and resistance in response to NGO criticism and foreign economic sanctions targeted on rights-abusing states. In the original version of the theory, this resistance was expected to collapse as the target state became caught in its own self-contradictory rhetoric and in the pincer of domestic mobilization and international pressure (Risse et al. Reference Risse, Ropp and Sikkink1999). A later restatement, however, acknowledged that resistance could be expected to persist under common adverse conditions – in wartime, in autocracies, in states that are too strong or too weak, and in cultures where rights violations are deeply embedded in society (Risse et al. Reference Risse, Ropp and Sikkink2013).

Most concerning is the likelihood that the syndrome of shame, anger, and resistance will spark not only short-run hostility to the rights project, but that the social context of shame will play into the hand of illiberal populist movements and embed self-perpetuating prophecies of deviance. Shaming and anger have the benefit of mobilizing the shamer's own activist base, but the social psychological perspective supports the conjecture that they tend to mobilize and lock in the target's base, too.

Through the self-fulfilling prophecy of shaming and the backlash against it, both sides nurture their ideologies of hostility and contempt. Like teenagers who are labeled juvenile delinquents, pariah states and transnational anti-system groups tend to create subcultures of the ostracized, alienated, and isolationist. By ‘propagating the subcultural ideology’ they gain ‘social recognition of the anti-mainstream’ (Suzuki Reference Suzuki and Zarakol2017, 227–28, 231, 231). As with the bromance among Putin, Trump, and EU right-wing populists, even nativists and nationalists find reasons to seek out membership in an international club of bad-boy sovereignty hawks. In this terrain of emotion-fueled ideology and political mobilization, humiliation hardens into grudge, and human rights can wind up worse off than they might have been with subtler tactics (Markwica Reference Markwica2018, 83).

Implications for a prescriptive theory of shaming

The strategy of human rights promotion that gave a central role to ‘naming and shaming’ was based on a number of social and psychological assumptions that are ripe for reassessment. These include several assumptions that bear directly on persuasion through the mobilization of shame: the innate human instinct to recoil at the exploitation of vulnerable people, the resonance of human rights ideas in the normative systems of most cultures, the persuasive potential of ‘information politics’ to leverage the material power of liberal democracies, the sensitivity of most states to their status in an increasingly liberal international order, and the potency of shaming for isolating past perpetrators and deterring new ones. Key assumptions also include contextual factors bearing on the power of liberal social forces: the rising power of global civil society networks, their tendency to support further waves of democratization based on liberal rights principles, the feasibility of building effective domestic and international institutions that embody these mechanisms of accountability, and the inexorable tendency of all this to induce socialization to and internalization of human rights thinking (Finnemore and Sikkink Reference Finnemore and Sikkink1998; Keck and Sikkink Reference Keck and Sikkink1998; Tarrow Reference Tarrow1998; Risse et al. Reference Risse, Ropp and Sikkink1999; Sikkink Reference Sikkink2011).

Some of the disappointments of shaming tactics are due to the broader political environment in which liberal democracies have struggled to manage challenges of economic stability, immigration, terrorism, domestic governance, and authoritarian rivals. These setbacks present problems for human rights advocacy as well as the credibility of the liberal system more generally. If liberalism were in better working order, human rights shaming might be getting somewhat better results.

While the present moment is therefore a hard case for evaluating shaming strategies, there are also theoretical reasons to believe shaming is in general a weak reed on which to rest human rights activism. At the individual level of emotional psychology, shame and shaming are chronically connected to two mechanisms that seem bound to produce unintended consequences for compliance: the syndrome of shame, resentment, anger, and pride in deviance, and the syndrome of denying and hiding non-compliance. At the level of group psychology bearing on esteem and status, dominant approaches likewise offer little support for shaming by outsiders. The dominant approach, SIT, finds that people in groups base their self-esteem in part on the standing of their group, discriminate against and denigrate outgroups as a way of sustaining personal and group status, and react strongly to humiliation at the hands of outsiders. Other approaches serve mainly to reinforce this tendency. They portray groups as striving toward an internal consensus model of the truth, which is supported by powerful processes of ingroup socialization and a bias for conformity. These individual and social psychological processes provide readily available resources for rallying group members around established norms, practices, and institutions, especially when challenged by shaming from the outside. The overall implication of this research is that shaming is a potentially very powerful weapon that can easily explode in the hands of the wielder.

While sobering, none of this means that liberals should give up on promoting human rights. Notwithstanding current difficulties, liberal democracy based on rule of law and the full panoply of human rights is by far the most successful form of social organization yet invented. It remains true that democracies never fight wars against each other, and no country other than parasitical oil states and Singapore has navigated through the ‘middle-income trap’ (about one-fourth US GDP) without adopting a thoroughgoing liberal order, including human rights (Dollar Reference Dollar2015). Human rights are so important that they need to be promoted effectively, not jeopardized by the unintended consequences of shaming.

How should this be done? One approach might follow from the Braithwaite school of reintegrative shaming. Criticism should be respectful, focused on the deed rather than a possibly irremediable character flaw, and aimed at repairing the social rift. It should come from insiders to the social group, or outsiders who are widely respected and seen as sympathetic. Forceful reminders of principled standards should be directed to everyone, not just those at risk of misbehavior. Braithwaite (Reference Braithwaite1989, 84–97) notes, however, that this works better in communitarian societies, in situations of cultural uniformity, and among people who are socially very interdependent, which can vary by age, gender, and other individual characteristics. With respect to performance indicators, countries that are falling short of standards should be compared with their own prior performance, not shamed by comparison to neighbors and rivals (Cooley Reference Cooley, Cooley and Snyder2015).

Another approach might draw on the literature on vernacularization and localization of norms (Acharya Reference Acharya2004; Levitt and Merry Reference Levitt and Merry2009). Don't lecture; have a two-way conversation about normative standards. Don't insist on using the language of legalism and universalism; acknowledge the validity of local normative systems, and use generic language of respect and fairness that travels across normative systems. Reserve legal talk to subject matter where outsiders have patently legitimate standing, such as respect for legal due process as a condition of doing international business.

A very important strategy is to advance compliance standards not as moral or even legal imperatives but as technical advice for succeeding at a task. Ruling circles in developing countries who are skeptical about human rights are nonetheless keen to gain wealth, technological sophistication, advanced medical services, and other desirable trappings of modernity, many of which flow from advanced liberal democracies and the global capitalist system that liberal states run. States with rights compliance shortfalls tend to be much more enthusiastic about the looser ‘rights-based approach’ of the UN Sustainable Development Goals, which loosely link good governance targets and indicators to tangible development assistance (United Nations 2015). This removes human rights advocacy from the realm of shaming and locates it nearer to management consulting. Chayes and Chayes (Reference Chayes and Chayes1993) argue that most violations of international law stem from incapacity. Sometimes fixing organizational and technical problems can facilitate rights compliance. For example, Indian police with human rights training argue that rule of law might be fine in principle, but they say they have to torture detainees to protect the rights of crime victims because their local court system is so dysfunctional (Wahl Reference Wahl2017). In hard cases that lack a favorable setting for human rights shaming, performance indicators might be more usefully designed as constructive diagnostics for institutional reform than as tools for shaming.

Finally, the credibility of human rights as a standard for social behavior depends on how attractive and dynamic the liberal international order is. It also depends in part on whether people can see themselves and their identity group fitting into that order successfully. This means that a top priority for promoting human rights is restoring the health of the liberal order and tailoring rights initiatives to the prevailing conditions in places where abuses are occurring. The social psychology of emotion suggests that transnational shaming is unlikely to make a constructive contribution to those efforts.