Disidentifcation and emotions in individualistic versus collectivistic societies

Disidentification is the active distancing from a group that is relevant to the self-concept [4, 5, 9–12] and it consists of three components: disengagement (a negative motivational state that ranges from a rather passive and cognitive alienation to an active separation from one´s ingroup), dissimilarity (the degree to which individuals perceive themselves as different from the ingroup prototype) and dissatisfaction (a negative evaluation of one´s group membership; [5]). Individuals who disidentify from their groups continue to formally belong to the groups but have internally distanced themselves from them; they see themselves to be different from other group members and are dissatisfied with their group membership (e.g., [5, 12]). Disidentification is more than low identification or a lack of identification (nonidentification), because it represents an active and unique psychological state (e.g. [5, 9–11]. This means, disidentification involves disconnecting (typically negative) aspects of the group from the self, separating one’s identity from the group, feeling dissatisfied about belonging to the group and perceiving oneself as different from other group members. In contrast, nonidentification with a group is a rather neutral process which is characterized by a lack of both identification and disidentification [13]. Up until now, different antecedents of disidentification have been studied: For example, it is especially induced by negative experiences with a group, such as rejection [12], a bad reputation [4], a lack of ingroup-support [14], a perceived conflict of values between oneself and the group [6] or when an identity is under threat [15–17].

However, there have only been a few studies on emotional predictors of disidentification. One study conducted with German students found that negative emotions in general (an average score including anger and shame) mediated the link between un-preferred group assignment (being in a group participant did not feel they should belong to) and disidentification [18]. It remains unclear, however, whether there were any differential effects of anger and shame, because the analyses were conducted on a score that collapsed across particular kinds of emotions. Becker and Tausch illustrated that dissatisfaction was associated with shame and shame predicted intentions to conceal one´s group membership [5]. However, past research on emotional predictors of disidentification has been limited to studies of people in individualistic cultures. Recently, Bierle, Becker & Ikegami, explored whether emotional correlates of disidentification varied between individualistic and collectivistic cultures [19]. They found that disidentification exists in both individualistic and collectivistic countries but appears to serve different psychological functions: In more collectivistic cultures disidentification was used as a coping strategy to deal with intra- and intergroup conflicts. For example, when Japanese participants experienced anger because of a group conflict, they were more likely to disidentify from their ingroup. Thus, in more collectivistic societies, such as Japan, disidentification serves as a useful coping strategy to deal with culturally undesired emotions (i.e., anger) while still maintaining ingroup harmony. In contrast, in more individualistic societies, such as Germany, direct confrontation with others was used as a coping strategy to deal with intra- and intergroup conflicts. For example, when German participants experienced anger because of a group conflict, they were more likely to openly confront members of their group [19]. These findings raise the question of why these different coping strategies are used across cultures. Why would Japanese be more likely to disidentify from their ingroup when they experienced anger, when Germans tend to openly express their anger towards their ingroup?

Anger and shame in collectivist versus individualist societies

Anger and shame are both implicated in relationships as they can be elicited by other people, by an ingroup or outgroup. Individuals feel angry when they are affronted by others, their ingroup or outgroup, and they experience shame when they fail in the eyes of others. However, cultures differ in some of the ways that relationships are formed. In more individualistic cultures, maintaining individual autonomy in relationships tends to be emphasized, meaning that people have the tendency to see themselves as independent from others, aim to pursue their individual goals actively, are motivated to feel good about themselves and focus on individuals’ needs [20–22]. In contrast, in more collectivistic cultures, achieving a sense of belongingness is emphasized [23], meaning that social norms tend to focus on group needs, interdependence, and social harmony [24]. As a consequence, it is likely that people from collectivistic and individualistic cultures experience anger and shame in interpersonal and intergroup situations somewhat differently. Indeed, much research finds that the social desirability of experiencing and expressing anger and shame varies across cultures.

For example, past research in some collectivistic cultures finds that expressing anger is less socially accepted [25, 26] and is experienced to a lesser extent than in individualistic cultures [27–29]. In fact, anger poses a serious threat to the central collectivistic goal of relational harmony and embeddedness, not at least because in collectivistic cultures, people’s close relationships are less likely to change, even those with significant conflict [30–32]. Consequently, self-assertion or expressing anger in group situations is perceived as undesirable by fellow group members as it makes it problematic for fitting in well with others. In order to comply with social norms, such as being polite, maintaining harmonious relationships, and showing emotional restraint, people in collectivistic societies avoid expressing their anger, and often rely on subtle and passive strategies instead [28, 29, 33–36]. Therefore, disidentification with one’s group may provide a way to psychologically escape the anger-eliciting group, and may function as a regulatory response to contain the disruptive emotion and limit its potentially harmful consequences [37]. We thus hypothesize that in more collectivistic societies, anger elicited by a group conflict should predict more disidentification with one’s group, because it allows to psychologically detach their selves from the ingroup so that the anger becomes less intense.

In contrast, from the perspective of individualistic values, anger supports goal pursuit and self-assertion and is perceived to be constructive [38–42]. In individualistic societies, experiencing and expressing anger is understood as a desirable sign of healthy and mature self-expression [37, 43–45], whereas not expressing one’s anger is perceived as potentially harmful in the long run [46]. In contrast, in collectivistic societies, individual self-expression is not as important, but instead, the relational aspect, namely managing the relationship between the self and the group is given primary importance. Supporting this argument, research has found that Chinese-Canadians experienced anger less intensely than European-Canadians and were more comfortable with strategies that served to reduce their experience of anger, whereas the European-Canadians suffered from physiological consequences if they did not openly express anger [28, 29, 47]. Thus, people in more individualistic cultures tend to be more likely to directly express their anger rather than disguising their anger by coping methods such as disidentifying with their groups. Therefore, we hypothesize that anger and disidentification should be less strongly related in individualistic cultural contexts compared with collectivistic ones.

On the other hand, shame is less connected to the goal of actively combating injustice through action intentions or of confronting responsible actors [38]. Instead, shame is often experienced when an individual attributes some failure or wrongdoing to internal, global, and stable aspects of the self [48], which should be particularly unpleasant in more individualistic cultures [43]. As the experience of shame highlights both the shortcomings of the individual and their dependence on other people’s judgments [49, 50], it should interfere with cultural goals of standing out and achieving high self-esteem. Hence, in individualistic cultures, shame should be expected to motivate people to hide, withdraw, or escape from the situation or public awareness [51]. Given its particular undesirability in individualistic cultures, people who experience shame need strategies to cope such as down-regulating their shame [51–53], or transforming it into a more acceptable experience [54, 55]. Hence, in more individualistic cultures, feeling ashamed because of a group conflict should lead to disidentification, because disidentification could help individuals to psychologically distance themselves from the group, thereby reducing their feelings of shame.

In contrast, in collectivistic contexts, shame is a more constructive and valued emotion that helps to ensure social control, and facilitates harmonious social relationships by making people aware of their shortcomings, which allows them to adjust their behaviors [50, 56, 57]. In such contexts, shame may be up-regulated and promoted, such that it is more commonly experienced. Moreover, if shame plays a role in restoring or repairing relationships in collectivistic contexts, there may be less of an urge to suppress the emotion. The typical response may instead be to yield to feelings of shame and to capitalize on their relationship-restoring potential by seeking closeness with others [58]. Hence, we expect that people in collectivistic cultures would be less likely to try to reduce their feelings of shame, nor use disidentification as coping strategy.

The present research

We conducted 2 studies to examine whether anger and shame differentially predicted disidentification among people in more individualistic cultures (Study 1: Canada and Germany, Study 2: Germany) and in a more collectivistic culture (Study 1 and 2: Japan). Our selection of countries was guided by Hofstede’s individualism-collectivism dimension [59]. This bipolar dimension ranges from 0 to 100; higher scores indicate stronger manifestations of individualism. Germany and Canada represent individualistic countries (scores 67 and 80), whereas Japan characterizes a mid-range collectivistic country (score 47). Although Japan does not represent an extreme collectivistic country based on this dimension (and becomes more individualistic) central cultural norms are still collectivist in Japan [60]. There is a wide range of research showing the persistence of collectivism in Japan, especially the importance of collectivistic living (increased importance in social obligation, social harmony, and social contribution and decreased importance in individual rights [60–62], and low levels of relational mobility [32].

Study 1 explored the relationships between anger, shame and disidentification by exposing participants to 4 different conflict scenarios and asking them to report the degree that they would feel anger, shame, and disidentification. When selecting suitable scenarios, we were interested in situations which occur in both Japan and Germany and are thus realistic in both countries and likely to trigger authentic emotions in Japanese and German participants. In addition to that we selected scenarios which differed in group size and group type, because disidentification occurs when the group is disliked but still important to the self-concept. If one is indifferent to a negatively rated group, there should be a different emotional reaction to it. Cross cultural research has shown that the importance of groups to ones’ self is also influenced by the culture in which we grow up: specifically, people in rather individualistic cultures are by tendency intergroup oriented, that is they identify themselves with the ingroup as an abstract social category and look at the ingroup as a monolithic social category comprised of members who share similar attributes in the comparative context with outgroups. On the other hand, people in rather collectivistic cultures are by tendency intragroup oriented, that is they perceive their self to be related to other ingroup members via relational ties and the ingroup to be a bounded network of such ties [26].

Therefore, given the cultural differences regarding group and conflict types, we selected four different scenarios in Study 1 that vary in group and conflict type instead of confining us to one single scenario that might have more relevance in one specific cultural context. Study 2 was designed to replicate the finding of Study 1 by systematically testing the causal effect of anger and shame as emotional predictors of disidentification.

Method

Results and discussion

Measurement invariance was not established. Given the lack of invariance, we used a careful approach, analyzed data within countries only and avoided comparisons between countries (Results can be obtained from the first author).

We tested predictors of disidentification using anger and shame as predictors in regression models for each country and conflict situation separately. An initial analysis revealed that results were not moderated by gender (see S2 Table in S1 File). Corresponding descriptive statistics are provided in Table 1 and intercorrelations among all study variables are provided in Table 2 (intercorrelations among all study variables for each scenario separately are provided in S3 Table in S1 File). The variance inflation factor suggests that we do not have to deal with multi-collinearity.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Means and standard deviations of anger, shame and disidentification (Study 1).

https://doi.org/10.1371/journal.pone.0289918.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Intercorrelations among all study variables (Study 1).

https://doi.org/10.1371/journal.pone.0289918.t002

Among Japanese, anger was positively related to disidentification in all 4 conflict situations (see Table 3). In contrast, shame was not related to disidentification for the INTER_small, INTRA_large and INTER_large group scenarios. In the INTRA_small scenario, however, shame did positively predict disidentification, which was counter to our hypotheses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Main effect model for prediction of disidentification for each conflict situation (Study 1).

https://doi.org/10.1371/journal.pone.0289918.t003

Among Germans, shame positively predicted disidentification in the INTRA_small, INTER_large and INTRA_large scenarios, and was in the same direction for the INTER_small scenario, although this effect only approached significance. In contrast, anger did not predict disidentification in any of the scenarios, and showed a marginal negative relation with disidentification in the INTER_small scenario.

Among Canadians, shame positively predicted disidentification across all 4 scenarios. In contrast, anger did not positively predict disidentification in any of the scenarios, and was negatively related to disidentification in the INTER_large scenario.

In sum, for the most part, the results were consistent with our hypotheses. Given that we found comparable effects across the four heterogenous scenarios, we are confident that the results can be generalized to different situations involving small and large groups, intra- and intergroup conflicts. While anger predicted disidentification among Japanese participants for all scenarios, shame was unrelated to disidentification across scenarios with the exception of the INTRA_small scenario. In stark contrast, among both the German and Canadian samples, anger did not positively predict disidentification for any of the scenarios, however, shame positively predicted disidentification for all scenarios. These findings provide preliminary support for our hypotheses and suggest that anger but not shame is related to psychological ingroup detachment for individuals in collectivistic cultures, whereas shame but not anger is related to psychological ingroup detachment for individuals in individualistic cultures.

One limitation of these findings is that we measured anger, shame and disidentification and can only speculate in terms of the causal directions. While we have reasoned that it makes sense to expect that people’s emotional reactions will lead them to disidentify it is also possible that it is the experience of disidentification that triggers anger in more collectivistic and shame in more individualistic cultures. Study 2 was conducted to address this limitation by testing the causal impact of emotions on disidentification in an experiment.

Methods

Results and discussion

Measurement invariance was not established. Given the lack of invariance, we made again only within country comparisons. As before, results were not moderated by gender (see S4a and S4b Table in S1 File).

Then, an independent-samples t-test was conducted separately for each country (instead of a 2x2 ANOVA) to compare the effect of condition (anger vs. shame) on disidentification (see Table 4 for descriptive statistics). There was a significant effect of condition on disidentification in both countries (Germany: t(98) = -2.91, p = .004, d = 0.58; Japan: t(69) = 2.57, p = .012, d = 0.61). In Germany, participants assigned to the shame condition reported higher levels of disidentification than participants assigned to the anger condition. In contrast in Japan, participants assigned to the shame condition reported lower levels of disidentification than respondents assigned to the anger condition. In sum, Japanese disidentified more when they felt anger compared to shame, whereas Germans disidentified more when they felt shame compared to anger.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Means (M) and Standard Deviations (SD) of disidentification (Study 2).

https://doi.org/10.1371/journal.pone.0289918.t004

Limitations and future directions

We acknowledge several limitations of our work. First, we did not measure whether people have an independent or interdependent self-construal but used their cultures as a proxy for the degree of individualism/collectivism. However, self-construals can differ within cultures [23]. Therefore, future research should include scales assessing self-construal, which can be used as potential mediators. Second, we focused on individualism versus collectivism in our studies for theoretical reasons, but there are others cultural dimensions that might be relevant such as power distance, uncertainty/avoidance, masculinity/femininity, long-term orientation, and indulgence/restraint [59], in explaining the differences between Japan and Germany. Future research could attempt to replicate the present findings by comparing countries that differ on collectivism and individualism but not on other cultural dimensions.

Second, we selected Japan as collectivistic country. However, recent research illustrates that although central cultural norms are still collectivist in Japan, Japan is getting more individualistic [60], but particularly in the North of Japan [40], where we did not collect our data. Nevertheless, future research could examine emotional predictors of disidentification in a society scoring very low on the individualism scale (e.g., Ghana, Burkina Faso or Indonesia). Furthermore, it would be interesting to involve participants from a collectivistic honor culture—in addition to a collectivistic face culture (like Japan). In contrast to collectivistic face cultures, in unstable and dynamic hierarchies that characterize honor cultures, achieving and protecting honor may require competitive actions to assert and protect one’s own reputation, the reputation of families, friends, or other close social groups [69, 70]. This means, if there is a situation that provides a threat to lose reputation, people in honor cultures can be expected to act competitively or even aggressively [71–73]. This would mean that anger might be more accepted in honor cultures, which could lead to similar results as in individualistic cultures.

Third, in Study 1 we measured emotions using only one item per study. We did this in order to avoid ambiguity. We are not aware of synonym of shame in German language. Moreover, this procedure is in line with several other studies [74–76]. However, future research could try to capture emotions related to anger and shame using additional descriptive items.

Fourth, we have neglected the potential role of social status. Some past research indicates that Japanese people of higher social status are more likely accorded the privilege of expressing anger to maintain hierarchies [77]. Thus, the present findings could be moderated by social status and might be particularly true for people with lower social status in Japan. Therefore, future research would benefit from testing social status as a potential moderator.

Fifth, given that we had specific predictions regarding the emotional predictors in Japan vs. Canada/Germany in Study 1, and focused on the difference between Japan vs. Germany in Study 2, we did not include a neutral baseline group. Future studies that included neutral baseline groups may be informative. Future research should also examine intra-individual differences, that is, to assess identification and disidentification at Time 1, then elicit one emotion (anger vs. shame) and test whether there are intra-individual changes in disidentification and identification based on the emotion that was elicited.

Finally, it is possible, that the emotion participants were asked to imagine did not fit their natural emotional appraisal of the scenario situation. Therefore, it would be helpful if future research could extend the manipulation check by asking participants how strongly they felt the emotion that we induced (and potential other emotions). Furthermore, building on our findings, future studies could use a different emotion manipulation.

Conclusion

In sum, this research extends the literature on emotional predictors of disidentification in a number of ways. First, it is novel to examine anger and shame as specific emotional predictors of disidentification across cultures. Second, research on disidentification has thus far been largely conducted in individualistic cultures (for exceptions see [78–82]), and research has shown a fairly reliable relationship between shame and disidentification [5]. Our research demonstrates that this relationship between shame and disidentification does not generalize to all people, as in collectivistic contexts disidentification is predicted by anger. Moreover, our findings provided similar results across different conflict scenarios (i.e., inter vs. intra group and small vs. large group related conflict scenarios), and across both correlational and experimental designs, suggesting a rather robust tendency. It would be informative in future research to see if these results generalize to other individualistic and collectivistic cultures, and whether similar findings emerge with other related emotion terms.