Sources of Ostracism

Investigating the Consequences of Ostracizing

Ostracism research was initially inspired by real-world observations that being excluded and ignored had profound negative consequences for the target (e.g., depression, aversive effects on health and well-being; Gruter & Masters, 1986; Williams, 1997). It is thus unsurprising that researchers initially focused on targets, developing models that were primarily target-centric (e.g., Williams’s model of ostracism; Williams, 1997) and devising paradigms to examine the effects of being ostracized in laboratory settings (e.g., “Cyberball”; Williams, Cheung, & Choi, 2000; “O-Cam”; Goodacre & Zadro, 2010).

Yet in this quest to understand the consequences of being ostracized, researchers (paradoxically) ignored the sources of the ostracism. They were barely acknowledged in early models of ostracism and were absent entirely from later versions (e.g., Williams’s model of ostracism; Williams, 2009). Moreover, in most laboratory paradigms, the sources of ostracism were either played by confederates or were computer generated, which eliminated any opportunity to examine the effects of ostracizing (Williams, 2007).

Initially, what was known about sources was found using such qualitative methods as structured interviews (e.g., Zadro, Arriaga, & Williams, 2008), diaries or event-contingent self-reporting records (e.g., Williams et al., 2001), and self-report narratives (e.g., Sommer, Williams, Ciarocco, & Baumeister, 2001; Williams, Shore, & Grahe, 1998). These qualitative studies provided rich information about ostracizing that could not have been uncovered in the laboratory (e.g., the long-term consequences of ostracizing; Sommer et al., 2001).

Unlike being ostracized, which uniformly leads to negative psychological and behavioral consequences for targets (see Williams & Nida, 2011), ostracizing can often have positive consequences for sources, according to some qualitative studies (see Zadro, Godwin, & Gonsalkorale, in press). For instance, real-world sources reported that ostracizing afforded them a sense of power over the target that was not possible during an argument (e.g., Zadro et al., 2008). Moreover, they stated that ostracizing often led to personal gain; for instance, targets often behaved ingratiatingly toward the source in an attempt to mend their relationship (e.g., buying them gifts, begging for forgiveness; see Zadro, Godwin, & Gonsalkorale, in press).

However, the consequences of real-world ostracizing are not uniformly positive; some sources expressed distress, remorse, and guilt over their ostracizing behavior (Zadro, 2004). Moreover, the aversiveness of ostracizing seemed to depend on multiple factors, including personality traits of the source, the identity of the target, and the duration of the ostracism episode (see Sommer et al., 2001; Zadro et al., 2008). Thus, the picture emerging from the qualitative studies reveals that the act of ostracizing can lead to a complex pattern of consequences that warrant further investigation.

Despite the clear need for systematic examination of the nature and consequences of ostracizing, however, only a handful of experimental studies have focused on sources (for a review, see Zadro et al., in press). Given that the effects of ostracism on targets have been investigated in the laboratory since the mid-1990s (see Williams, 1997), why has it taken so long to experimentally examine the effects on sources? One reason is that experimenters have found it challenging to devise paradigms that reliably induce participants to ostracize others.

The Challenges Associated With Investigating Sources in the Laboratory

One of the challenges faced by ostracism researchers is to devise ecologically valid paradigms that immerse participants in a powerful and believable ostracism situation. Although effective paradigms have been achieved for targets (see Williams, 2007), devising source paradigms has been more problematic, primarily because it requires inducing individuals to violate a strong norm (i.e., to behave in an inclusionary rather than an exclusionary fashion; see Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013). We will refer to the four broad categories of currently existing paradigms as follows: recall paradigms (e.g., Bastian et al., 2013), direct source inductions (e.g., Poulsen & Kashy, 2012), induced source inductions (e.g., Ciarocco, Sommer, & Baumeister, 2001), and contagion source inductions (Wesselmann et al., 2013).

In a recall paradigm, participants are instructed to recall an incident in which they ostracized others (e.g., Bastian et al., 2013). Although this paradigm is easy to administer and personally relevant to the participants, it does not provide a standardized ostracism experience; for example, recalled episodes may differ with respect to the motive and severity of the incident as well as the number of cosources. Moreover, autobiographical recall is subject to various memory distortions and selection biases (e.g., recency effects, telescoping bias). Thus, it is debatable whether recalling an ostracizing event and actually ostracizing during a laboratory paradigm result in comparable effects.

In a direct source induction, participants are instructed to ostracize a target during a subsequent interaction. In the O-Train simulation, two participants—who play the role of sources—are instructed to ignore and exclude the participant sitting between them (the target) for the duration of a 5-minute imaginary train ride (Zadro, Williams, & Richardson, 2005). Dispensing with the artifice of a role-play scenario, other researchers have simply explicitly instructed participants to ostracize a naive participant (e.g., Legate, DeHaan, Weinstein, & Ryan, 2013; Poulsen & Kashy, 2012). Direct inductions are easily administered and allow researchers to examine the consequences of ostracism for both targets and sources. However, for participants, following instructions to ostracize has been found to lead to a diminished sense of autonomy (Legate et al., 2013). This means that findings may pertain only to real-world situations in which people are pressured to ostracize and may not generalize to situations in which people choose to ostracize of their own volition.

In response to this concern, an induced source induction encourages, rather than instructs, participants to ostracize. Ciarocco et al. (2001) gave participants the choice to either exclude or include another person in an upcoming interaction; however, to encourage participants to exclude, the experimenter stated that most participants so far had chosen to include their interaction partners and that it would be helpful if more participants chose to exclude their partners. This paradigm is successful in the sense that participants typically do choose to ostracize. Nevertheless, participants may feel that they have no real choice, given that they have been informed by an authority figure (the experimenter) that ostracizing the target will assist the study.

The ecological validity of direct and induced source inductions is somewhat limited because ostracizing occurs at the direct or indirect behest of the experimenter. These inductions do not reflect situations in which the source is personally motivated to ostracize without intervention from an authority figure. In the real world, sources are typically motivated to ostracize by a specific goal, such as punishing the target (i.e., punitive ostracism; see Williams, 1997), correcting the target’s behavior (e.g., a time-out in school; Gruter & Masters, 1986; removing deviant members; Schachter, 1951), or protecting oneself from harm (i.e., defensive ostracism; Zadro et al., in press).

Researchers have recently sought to provide a personal motivation for participants to ostracize in a laboratory setting. Wesselmann et al. (2013) made the target act in a burdensome manner (i.e., by being slow to throw the ball) during a four-person ball-tossing game consisting of the participant and three computer-generated players. Although researchers who use this paradigm aim to examine ostracizing behavior, the act of ostracizing is actually initiated by two other confederate players—the participant then chooses whether to follow their lead and ostracize the target or to include the target. Thus, this paradigm may be more accurately described as a contagion source induction, which assesses whether people who observe an ostracism event will actively model the other sources and participate in the exclusion when given the opportunity to do so, thereby adhering to the perceived group norm (i.e., that ostracizing others is acceptable behavior). Future research is needed to determine whether this contagion paradigm induces the same effects as the more direct source paradigms.

It is clear that each paradigm has its limitations, and findings may not be generalizable beyond the specific conditions created by the particular paradigm. However, as we review in the next section, each paradigm has yielded valuable insights into the consequences of ostracizing.

The Positive and Negative Consequences of Ostracizing

According to Williams (2007), ostracism is unique from other forms of interpersonal conflict because it affects four primary needs: belonging (the need for social acceptance and connection to others), control (the need for a sense of mastery over oneself and one’s environment), self-esteem (the need to have a positive feeling of self-worth), and meaningful existence (the need to have a sense of purpose and acknowledgment from others). The effects of ostracism on the needs of targets have been well documented (see Williams, 2007). But to what extent does ostracizing also affect primary needs? Moreover, does ostracizing threaten these needs or fortify them?

Several researchers using direct source paradigms have found that ostracizing leads to need fortification. For instance, in the O-Train paradigm, participants instructed to ostracize reported a more fortified sense of belonging and self-esteem compared with those who were instructed to include (Zadro et al., 2005). Likewise, Poulsen and Kashy (2012) found that sources who had been instructed to ostracize a target viewed themselves and their three cosources as more likable than the target—a finding consistent with the idea that ostracizing enhances self-esteem. Groups may benefit from the fortified belonging and self-esteem that individuals feel when they ostracize with cosources, because such feelings enhance group cohesion and commitment (e.g., Gruter & Masters, 1986).

In contrast to laboratory studies involving ostracizing with two or more cosources, interviews with real-world sources suggest that ostracizing the target one-on-one can threaten belongingness and self-esteem. Real-world sources reported that ostracizing often diminished the bond they shared with the target, who was typically a loved one (e.g., child or partner), thereby threatening belongingness (Zadro, 2004). Because a sole ostracizer must take full responsibility for ostracism outcomes, ostracizing can also threaten self-esteem; for instance, some sources felt disappointed in themselves for engaging in ostracism tactics with loved ones.

The potential for ostracizing to both fortify and deplete needs is also illustrated in relation to control. Sources wield enormous power over targets and the interaction, and the effectiveness of ostracism in enhancing control is often acknowledged by sources in recall paradigm or narrative studies (e.g., Sommer et al., 2001; Williams et al., 1998). Moreover, as researchers found during an O-Train study, ostracizing fortifies control more than does arguing (Zadro et al., 2005). However, to be effective ostracizers, sources must constantly monitor their behavior in the presence of a target (Zadro et al., 2005), which may, paradoxically, lead to a loss of control. Indeed, short-term ostracizing in a laboratory setting impairs performance on subsequent tasks that require self-control (Ciarocco et al., 2001), whereas in the case of long-term ostracism, sources have reported that they find it hard to stop themselves from ostracizing because it has become a difficult habit to break (Zadro et al., 2008). Thus, although ostracizing may initially impart a sense of control, ultimately sources may find themselves just as helpless in the situation as the target.

Finally, it could be argued that ostracizing may fortify meaningful existence (for a discussion on the links between ostracism and terror management, see Williams, 1997). Targets often resort to drastic behavior (e.g., ingratiation) in an attempt to manage the ostracism episode (Zadro et al., 2008). This may enhance sources’ perception of being acknowledged by others, thereby strengthening their sense of purpose. Although such fortification of meaningful existence can be seen as a result of real-world ostracizing, experimental studies have not yielded strong support. Consequently, the links between ostracizing and meaningful existence need to be further explored.

Future Directions

For almost two decades, ostracism research has centered on targets. Researchers need now to redress this balance by turning their attention to sources, which means addressing the limitations of current source paradigms. First, researchers need to improve ecological validity by developing paradigms that better reflect situations in which ostracizing occurs in the real world. Existing source paradigms focus on situations in which sources model the ostracizing behavior of others or are induced to ostracize. Thus, what is missing is a paradigm that motivates participants to engage in ostracism of their own accord. Gooley, Zadro, Gonsalkorale, Williams and Sethi (2013) induced such a motivation by presenting participants with an objectionable interaction partner during a getting-to-know-you game. During a subsequent game of Cyberball, the majority of motivated sources chose to ostracize the obnoxious partner of their own accord, thereby successfully inducing a motivated ostracism situation in the laboratory.

Second, whenever possible, researchers need to compare the ostracism experiences of targets and sources. This can be accomplished by adapting current target paradigms, such as Cyberball, to include sources (e.g., by adding a source induction—whether induced or motivated—before the game; see Gooley et al., 2013).

Ultimately, creating paradigms that allow researchers to look at different facets of the ostracizing experience (e.g., manipulating different motives for ostracizing, ostracizing alone vs. with cosources) will assist researchers not only to develop models that capture the dynamic interplay between targets and sources but also to devise ameliorating strategies that can be used in real-world ostracism settings, whether in the schoolyard, workplace, or home.