The Positivity of Anger: Non-Expression of Anger Causes Deterioration in Relationships ()
1. Introduction
Researchers focusing on relationships have been interested in anger management, and some have argued that not expressing anger is important for anger management (Baker & McNulty, 2010; Mattson, Frame, & Johnson, 2011) . For instance, their argument reveals that unexpressed anger has affiliative impression for close partners and often helps manage relationships successfully. However, recent researchers imply that unexpressed anger might prompt dissatisfaction rather than maintaining or strengthening relationships (Baker, McNulty, & Overall, 2014) . Assuming that unexpressed anger causes deterioration in close relationships, we examined the effect of perception of unexpressed anger on relationships.
Most evidence has indicated that unexpressed anger has positive interpersonal consequences. Using an autobiographical approach, Wallace, Exline, and Baumeister (2008) examined the effect of the perception of displaying forgiveness (rather than feeling forgiveness) on people’s subsequent behavior and their motivation to avoid repeating offenses: They found that participants were not likely to report repeat transgressions when victims had forgiven them. Moreover, Fincham and Beach (2002) conducted a longitudinal study of marriage, revealing that spouses’ reported tendencies to forgive their partners reduced their partners’ reported psychological aggression. These findings imply that suppressing or regulating anger is key to maintaining relationships. However, whether such anger management contributes to maintaining or strengthening relationships continues to be a question even today. In fact, some researchers suggest that unexpressed anger increases depression (Page, Stevens, & Galvin, 1996; Thompson, 1995) , perception of the partner’s resigned attitude for maintenance of relationships (Mikula, 1986; Rusbult, 1987) , or dissatisfaction. These findings suggest that unexpressed anger might prompt dissatisfaction rather than maintaining or strengthening relationships, implying that for anger management in relationships, it is important to not only conceal one’s anger but also express one’s anger. On the basis of the discussion, we assumed that expressing anger is also important in anger management for maintaining or strengthening relationships.
Nevertheless, compared with positive implications of not expressing anger, scholars have paid very little attention to potential negative implications of not expressing anger. But, some findings have such implications. Graham, Huang, Clark, and Helgeson (2008, Study 4) examined the longitudinal effect of expressed negative emotion, such as angry feelings, indicating that participants’ self-reported willingness to express negative emotions to their roommates was associated with increased intimacy. This result implies that unexpressed anger is detrimental to relationship development and intimacy. In a sample of married couples, Yoo, Clark, Lemay, Salovey, and Monin (2011, Study 1) asked participants to report their tendency to express anger toward their spouses and marital satisfaction. Their research revealed that unexpressed anger served to prolong marital satisfaction only when couples’ communal strength (the degree to which individuals wish to take responsibility for a relationship partner’s welfare) was weak, implying that unexpressed anger does not always benefit relationships. Furthermore, in a longitudinal study of newlywed couples, McNulty (2011) confirmed that among spouses reporting their tendency to express forgiveness (not to feel forgiveness) to their partners, those partners’ reports of physical aggression increased more over the first 4 years of marriage. These findings suggest that non-expression of anger should be more associated with deteriorating relationships.
Why should non-expression of anger prompt relationships to deteriorate? One reason is that expression of anger is a self-disclosure behavior. Emotion scholars have asserted that whereas any internally experienced emotion signals information about one’s personal needs, outwardly expressed emotion conveys that information to others (Frijda, 1993) . According to this perspective, expressing anger indicates that something unjustifiably wrong has happened, and the person is needy (Clark & Finkel, 2005; Graham et al., 2008) , thereby suggesting to partners or potential partners that the needy person trusts the partner and is willing to be openly vulnerable. Therefore, expression of anger is perceived as self-disclosure revealing personal information. Conversely, concealing behavior would be perceived as inappropriate in close relationships. Since people expect partners’ open expression of personal needs (Parks & Floyd, 1996) , they regard unexpressed anger as rejection of disclosure, which harms relationships. For this reason, we assume that people often perceive unexpressed anger as dislike, disfavor, or dissatisfaction in intimate relationships.
Although scholars have emphasized that not expressing anger is a significant strategy for maintaining and/or strengthening relationships, we assume that those who do not express anger can be perceived as refusing self-disclosure, thus prompting negative feelings in those perceiving unexpressed anger. To test this possibility, we prepared a hypothetical scenario depicting the participants’ friend as clearly experiencing anger. In particular, the scenario included three conditions, and participants were randomly assigned to one of them: In the Expression condition, the friend expressed relevant anger; in the Non-Expression condition, the friend did not; in the Control condition, the scenario did not mention whether the friend expressed anger or not. After reading this scenario, participants were asked to assess their emotional experience of the scenario they would experience. We predicted that participants would report more dissatisfaction when their friend was described as experiencing, but not expressing, anger than when the friend was described as experiencing and expressing anger, or as simply experiencing anger with no information about whether the friend expressed it. Overall, the current study examines effects of not expressing anger on relationships’ longevity.
2. Method
2.1. Participants
In this study, 154 Japanese university students (50 males and 104 females) participated voluntarily; their mean age was 20.79 years and the SD was 1.80. They were asked to participate to obtain partial course credit in a psychology class.
2.2. Scenario
The students of a psychology class who agreed to participate in our research were provided a questionnaire titled “Psychological Survey of Interpersonal Relationships” and asked to complete it immediately. The questionnaire consisted of one scenario and items requiring participants’ responses to it. We developed three versions of the scenario by manipulating whether the friend who was clearly experiencing anger expressed that anger and provided participants with a scenario in each expression condition: Expression condition, Non-Expression condition, or Control condition. The target person whose friend was experiencing feelings of anger was labeled “you,” and participants were instructed to read as if they were the target person.
More specifically, participants were asked to picture themselves in the following scenario: “Imagine that you are having lunch with a same-sex friend1. During lunch, you notice that s/he is silent and has a look of displeasure.” In the Expression condition, participants then read, “When you ask her/him what is matter, s/he expresses to you her/his anger because s/he is having a fight with another friend.” In the Non-Expression condition, the sentence above was replaced by the following: “When you ask her/him what the matter is, s/he does not express her/his anger over interpersonal conflict with others to you.” In the Control condition, participants read the scenario without any mention of whether the friend expressed anger. In our research, these three conditions were between-participants variables.
2.3. Materials
After reading the scenario, participants completed two emotional response questionnaires.
The first emotional response was the measurement of the dissatisfaction by listing nine adjectives related to anger (see Batson et al., 2007 ). Participants were asked to indicate how strongly they felt irritated, angry, upset, annoyed, offended, outraged, mad, perturbed, or frustrated. The second response measured positive feelings through seven adjectives. Participants were asked to report how strongly they felt pleased, satisfied, lucky, content, joyful, delighted, or comfortable. These two types of adjectives were intermixed and rated on a six-point scale ranging from 0 (not at all) to 5 (very strongly) through the question “What would you feel if you encountered this scenario?” Figure 1 displays sample of the questionnaire used.
If unexpressed anger were perceived as negative behavior, what is the consequence to relationships of expressed anger? We speculated that people who perceived expressed anger would be more satisfied in close relationships, while people who perceived unexpressed anger would be less satisfied. Although this issue is not our main focus, we included the scale measuring positive feelings to elaborate the question of whether it is better to express anger or not.
3. Results
1We kept the gender of the participant and the protagonist in the scenario the same to minimize cross-gender self-presentation concerns.
Participants were randomly assigned to one of three conditions: Expression (n = 52), Non-Expression (n = 52), or Control (n = 50). Gender difference was not a main effect and did not qualify effects of experimental manipulations either on
![]()
Figure 1. Sample of the questionnaire used.
reported dissatisfaction, F(1, 146) = 0.01, p = .76, and F(2, 146) = 1.66, p = .19, or on reported positive feelings, F(1, 146) = 0.81, p = .37, and F(2, 146) = 0.05, p = .24, respectively. Given that there was no reliable effect of gender, all reported analyses are collapsed across both genders.
3.1. Reported Dissatisfaction after Reading the Scenario
To assess dissatisfaction after participants read the scenario, we averaged scores for nine dissatisfaction items to create an index of dissatisfaction (Cronbach’s α = .95). We then performed a one-way analysis of variance (ANOVA) with each of the three expression conditions as a between-participants variable, revealing that a main effect of anger expression was significant, F(2, 149) = 8.35, p < .001, η2 = .10. As can be seen in the first row of Table 1, participants reported more dissatisfaction when the friend did not express anger (M = 2.05) than when the friend did express anger (M = 1.18, p < .01) and when no information about expressing anger was provided (M = 1.44, p < .05). No difference between Non-Expression and Control conditions was significant, p = .35.
3.2. Reported Positive Feelings after Reading the Scenario
To assess participants’ positive feelings, we created an index of positive feelings
![]()
Table 1. Means and standard deviations of measures in each experimental condition of expressed or suppressed anger.
The response scale for each measure was 0 to 5. The numbers within parentheses are standard deviations.
by averaging the score for seven positive adjectives (Cronbach’s α = .85). A one-way ANOVA revealed a significant main effect of anger expression, F(2, 149) = 7.61, p < .001, η2 = .09. As can be seen in the second row of Table 1, the mean reporting of positive feelings in the Expression condition (M = 0.41) was slightly higher than in the Control condition (M = 0.06, p < .001). However, there was no difference between the Expression and Non-Expression conditions, M = 0.41 vs. 0.19, p = .14, meaning that positive feelings did not increase despite the friend expressing anger.
4. Discussion
Consistent with our prediction, participants who read a scenario describing a friend experiencing anger self-reported more dissatisfaction when the friend did not express the anger. This finding provides unambiguous support for our hypothesis, meaning that participants viewed unexpressed anger with displeasure, dislike, or disfavor. Existing literature on interpersonal consequences of expressing anger has suggested that expression of anger leads to decreased intimacy and then to deterioration in relationships. According to present data, however, we claim that for anger management in relationships, it is important not only to conceal one’s anger but also to express one’s anger.
Paralleling rated dissatisfaction, experimental manipulation affected measurement of positive feelings: Participants self-reported more positive feelings when their friend was described as expressing anger (Expression condition) relative to when information about expressing anger was not provided (Control condition). Still, the difference between Expression and Non-Expression conditions was not statistically reliable; a floor effect was confirmed since mean response to positive feeling items in each of the three experimental conditions was quite low. This suggests that there is no evidence that people who perceive expressed anger are more satisfied. Although anger expression is certainly a self-disclosure behavior, the person perceiving the expressed anger may have difficulty accepting it or responding positively because the visible behavior also includes a hostile, complaining, or destructively aggressive tone.
Interestingly, given the results above, interpretation indicates that a close partner’s unexpressed anger elicits high relational dissatisfaction, but expressed anger does not elicit high satisfaction. This interpretation is relevant to debates in which people react sensitively toward suppressive behaviors, but not toward expressive behaviors. This implies that perceptions of expressed and suppressed anger are not common, but rather independent. What contributes to the asymmetrical effect of these perceptions?
One possible reason for the asymmetrical effect may be, as we assumed in the introduction, that an implicit expectation serves. In close relationships, people may regard concealed anger as transgressing the relational norm since people believe that close individuals should be open about their personal needs. In fact, Clark, Mills, and Corcoran (1989) found that participants checking lights more frequently meant that the partner appeared to be concerned with their needs if the partner was a friend. These researchers’ results show that people are strongly concerned about whether a partner is open about personal needs. The reason people react more sensitively to a partner’s suppressed anger, rather than to expressed anger, may be attributed to strong normative consciousness.
Second, from a social cognitive perspective, possibly an expresser’s behaviors that are incongruent with experienced emotion simply garner more attention. According to cognitive dissonance theory (Festinger, 1957) , behavior corresponds exactly to the emotion that an individual experience internally. Therefore, perceivers may feel dissonance with a person who shows inconsistency between internally experienced and outwardly expressed emotions; dissonance may result in restoring consonance through misinterpretation and so on. Since anger is believed to elicit motivation for outwardly impulsive behavior, non-expression of experienced anger may likely stand out by contrast and actually attract more attention than expression of experienced anger when a perceiver is confronted with inconsistent behavior. This may be why people react sensitively toward suppressed anger.
Although the present research’s main finding implies the association of unexpressed anger with negative relationship outcomes, we must be cautions in generalizing because of methodological limitations. The first limitation is that we did not directly measure how participants judged unexpressed anger in the scenario. The theoretical background that the partner not expressing experienced anger is judged to be less likable, less social, and less favorable is that the behavior is regarded as refusal to self-disclosure. However, whether or not participants perceived the behavior as a refusal to self-disclose is unclear. Another limitation is the type of intimate relationship. Although we focused on friendships, much remains for study in relationships regarded as intimate. In addition, even the same type of intimate relationship does not necessarily have the same level of intimacy. Therefore, to say the least, whether the same results will be found by examining a marital relationship, a romantic relationship, or other relationships in which the type of intimacy differs from that of friendship, remains questionable.
5. Conclusion
We examined whether unexpressed anger can be detrimental to the maintenance of relationships. Although the study had some limitations, we successfully suggested that unexpressed anger leads to increased dissatisfaction in relationships. This finding implies that expression of anger is a significant strategy for maintaining and/or strengthening relationships.
Acknowledgements
Funding for this research is supported by Japan Society for the Promotion of Science KAKENHI (Grant Number 15K17247; awarded to Shunsuke Uehara).