Behavioral responses to observed ostracism
Punishment of sources and targets
After making a judgment about whether ostracism is morally acceptable in a situation or not, an important question is how observers will react to the observed episode. One possibility is that observers act in line with their moral judgment. For instance, if individuals attribute ostracism to a malicious motive of the sources and devaluate the sources for it, they may aim to restore fairness either by compensating the target for being ostracized, or by punishing the sources for their non-normative behavior. Wesselmann and colleagues (2013) demonstrated in a number of studies that when individuals observed another person being ostracized in Cyberball without an apparent reason, they tried to compensate the target by directing more ball-throws towards that person (thereby also punishing the sources by directing fewer ball-throws towards them). To be able to differentiate between the motive to compensate the target and the motive to compensate the sources, Will, Crone, van den Bos, and Giiroglu (2013) let their participants play economic games after having observed exclusion in Cyberball and showed that observers both compensated the target and punished the sources, even when participants had to invest their own resources to do so (see also Giiroglu et al., 2013; Over & Uskul, 2016).
In contrast, if observers feel that the targets are to blame for their own exclusion and thus attribute ostracism to a punitive motive, observers should support the sources instead of the targets. Participants in Wesselmann and colleagues’ (2013) study stopped compensating the target when s/he was throwing the virtual ball more slowly than the other players in Cyberball. But observers not only stop compensating, they are even willing to additionally punish the targets for their misbehavior: In a series of studies, we presented participants with social group interactions, either as imagined scenarios or as a part of an alleged previous experiment (Rudert, Ruf, & Greifeneder, 2019). In these settings, the target either had or had not violated a social norm, and was subsequently excluded or included by the sources in Cyberball or an alleged group task. Participants then had the option to give a certain amount of money to both the target and the sources, or alternatively, to allegedly subtract money from the target’s and the sources’ bonuses. Whereas observers consistently punished the sources for ostracizing without a socially accepted reason (by giving them less money or even subtracting money from their bonus) we found no evidence for a monetary compensation of the targets.
However, when observers knew that the target had violated a social norm (e.g., lying to appear superior, moving in on one of the sources’ boyfriends, acting greedy and uncooperative in a dilemma of the commons game), they decided to punish the target instead. Additionally, punishment of the target was also influenced by the sources’ behavior towards the target: Observers punished a norm-violating target most strongly if they did not know how the sources had reacted to the target’s norm violation. If the target had already been excluded by the sources, evidence suggests that observers punished the target less severely, possibly assuming that the target has already atoned for her/his misdeeds. Interestingly, when the observers watched the sources include the target following a misstep of the target, the degree of punishment was diminished as well. Observers appeared to have a reduced urge to punish the target after seeing that the target had apparently been forgiven by the same people that had actually been hurt by the target’s actions. Thus, whereas observers seem concerned with restoring fairness following an ostracism episode, they also take into account which actions the sources took and adjust their own reactions accordingly (see also Pryor, Reeder, Wesselmann, Williams, & Wirth, 2013).
Even if observers and sources agree that the target had violated a social norm and thus deserves punishment, they might still disagree about what a “fair punishment” looks like. Interestingly, the direction of this disagreement may go several ways. On the one hand, given that observers are not directly affected by the target’s norm violation—and are thus psychologically more remote to the situation—observers might opt for a milder or more reversible punishment than the sources do. On the other hand, for the same reason of being psychologically more remote to the actual situation, observers might place stronger emphasis on values and norms (Ledgerwood & Callahan, 2012; Liberman & Trope, 1998) and weigh practical thoughts to a lesser extent. Thus, observers might opt for a harder punishment than the sources, who may consider other factors such as the target’s usefulness or their relation to the target. A classic example might be a person who gave his unfaithful partner the silent treatment for cheating on him but eventually forgives her, even though his friends may still advocate against it. In line with these considerations, Riva and Andrighetto (2012) have shown that participants feel that outgroup members (who are socially more distant) are less affected by social exclusion compared to ingroup members, which might further decrease observers’ empathy and make them favor a harder punishment.
Observers’ decisions on whether to punish or help the target might have consequences for them, too. After all, observers are observed by the target, the sources, and potentially other non-involved individuals. If, for instance, the sources see that the observer supports the ostracized target, the observer might run into the risk of becoming the next target. In doubt about who is to blame, observers might therefore choose not to act or even derogate and thus distance oneself from the target, just to be on the safe side. Accordingly, it has been demonstrated that even though observers might sympathize with the target, they will also perceive her/him as less human, more vulnerable for exploitation, and have less interest in interacting with him/her compared to the sources or neutral persons (Arpin et al., 2017; Park & Park, 2015). Especially if observers have to fear direct negative consequences from siding with the target, the motive to protect oneself from becoming a victim might become dominant. As a consequence, observers might eventually choose to support or even join the sources, even though they are not morally convinced that it is the right thing to do (Klauke & Williams, 2015). Williams (1997) called this behavior defensive ostracism, stating that it may be used to protect oneself from threats and/or to obtain control over a situation. For instance, whistleblowers in an organization are often defensively ostracized by their colleagues who are afraid of being associated with the whistleblower and thus becoming a target as well (Williams, 2002).
In contrast, siding with the sources and punishing the target might also come with certain risks, especially when the situation changes over time and the target ends up in a powerful position, or when non-involved individuals remotely or retrospectively evaluate the situation. Perhaps it is for these reasons that we found in our studies that observers tend to evaluate even weak norm violations of the target negatively, but choose only to act on behavior that they felt was highly inappropriate. The reasons for this might be twofold: First, observers might not be fully aware of what the specific norm in a group is, and as such, may choose to refrain from action even if they themselves feel that the behavior is wrong. In addition, there might be a higher ambiguity whether the public opinion will shift in favor of the target or the sources, and if observers take the “wrong” side, they might as well become the next target.
But even if observers choose not to act at all in order to be on the safe side, ostracism can still represent a threat for them if they share the same environment with the target and the sources. For instance, think of a situation when a student observes ostracism in his or her own class. After all, if ostracism is a socially accepted behavior within the class, then there is a chance that the observing student might, at some point, become a target themselves. This threat may be even more pronounced when the student shares similarities with the ostracized target: Studies show that observers report stronger feelings of powerlessness and humiliation when they watch a member of their ingroup being ostracized compared to a member of an outgroup (Veldhuis, Gordijn, Veenstra, & Lindenberg, 2014). As a possible response, individuals might attempt to shield themselves against this threat and engage in self-protecting behavior and attributions. For instance, observing ostracism seems to encourage individuals to seek affiliation and increase social attention (similar results have been shown for the targets of ostracism, see Claypool & Bernstein, 2019; Timeo, Kiva, & Paladino, 2019; Williams et al., 2019). In studies with children, it has been shown that priming third-party ostracism increases both imitation of the experimenter (Over & Carpenter, 2009), and increases the likelihood that the children will draw affiliative pictures (Song, Over, & Carpenter, 2015) as well as sit closer to another person (Marinovic, Wahl, & Trauble, 2017).
Methodological considerations when investigating observers of ostracism
Aside from theoretical considerations, there are some important methodological issues that need to be considered when doing research about observers of ostracism. In the majority of observer studies that have been done thus far, participants observed a stand-alone Cyberball game, usually with players that have no common history that the observer is aware of (e.g., Giesen & Echterhoff, 2018; Giiroglu et al., 2013; Masten et al., 2013; Masten et al., 2011; Meyer et al., 2012; Wesselmann et al., 2009; Will et al., 2013). In this kind of situation, observers have no reason to assume that the target is being punished for anything, and so will usually tend to attribute ostracism to a malicious motive of the sources and subsequently empathize with the target, as well as punish the sources. It is important to keep in mind, though, that this finding may represent a distinctive feature of the employed paradigm, whereas ostracism in the wild is embedded in a context. First, in real life, there is usually the possibility that the target had deviated from the norm prior to being ostracized by acting uncooperatively or rudely. Second, in real life, observers usually have more cues they can draw upon, such as characteristics of the context, the target, or the sources (see before). If researchers rely solely on context-free paradigms in which the observed group has never met before, they may erroneously interpret findings that are characteristic for these studies as applicable to all kinds of social exclusion situations.
From this perspective, it is highly important to create or to adapt methodological paradigms so that they allow for more context. To a certain degree, current research paradigms such as Cyberball can be adapted to include more context (Gooley et al., 2015; Wesselmann, Williams, & Wirth, 2014; Wesselmann, Wirth et al., 2013). Moreover, alternative paradigms might be more suitable to investigate observer effects, such as having participants recall autobiographical events in which they observed ostracism (Rudert et al., 2018), or present them with scenarios (Giesen & Echterhoff, 2018; Over & Uskul, 2016; Rudert et al., 2019), chatlogs (Arpin et al., 2017; Rudert et al., 2018), videos (Song et al., 2015), or something else. Within these paradigms, the situational context or characteristics of targets and sources can be systematically varied to subsequently measure the effect on an observer’s cognition and behavior. Meta-analyses could further help identifying paradigms which are particularly sensitive to contextual changes and thus most suitable to investigate potential moderators of observers’ interpretation and reactions to ostracism.