Whitehouse presents a theory of extreme self-sacrifice by integrating previous theoretical insights with new empirical evidence. At the core of this theory lies the concept of identity fusion (Swann et al. Reference Swann, Gómez, Seyle, Morales and Huici2009), a psychological construct referring to the alignment between personal and social self, which has been found to predict endorsement of violence against out-groups when one's own group is threatened.

One outstanding question with respect to identity fusion is whether it describes a single phenomenon or, instead, is an umbrella term for a number of related phenomena, as this construct has been used to measure bonds that range from dyads (Vázquez et al. Reference Vázquez, Gómez, Ordoñana, Swann and Whitehouse2017) to countries (Swann et al. Reference Swann, Gómez, Huici, Morales and Hixon2010b). Indeed, the proponents of the theory argue that there are two distinct types of fusion (Swann et al. Reference Swann, Jetten, Gómez, Whitehouse and Bastian2012; Whitehouse Reference Whitehouse2013): “local fusion,” characteristic of small groups, and “extended fusion,” found among large “imagined communities.” The latter type involves the projection of properties associated with local fusion onto a larger group of anonymous individuals.

There are, however, various settings that do not clearly fall into one category or the other, such as when large groups of people experience fusion as a result of shared arousal, including massive religious and secular rituals, athletic events, demonstrations, and riots. One example of particular relevance for the discussion of extreme self-sacrifice is sporting events, which in many contexts can result in acts of violence, committed by individuals who are willing to put themselves at great risk to antagonize their rivals.

Sports hooliganism is a widespread, ancient (Cameron Reference Cameron1976; Frosdick & Marsh Reference Frosdick and Marsh2013), and puzzling phenomenon. Unlike other forms of violence that might be attributed to particular worldviews, hooliganism typically lacks such an overarching ideological framework, other than the existence of the group itself. Fans do have collective narratives, symbols, and scripts, but those are trivial to the transmission and maintenance of the “faith,” so to speak. Extreme levels of commitment to the team are forged not by explicit dogmas, but through the act of participation in collective events that have optimal structural arrangements for generating fusion.

During the game, dedicated fans move and chant in synchrony for hours. Laboratory and field studies show that shared psycho-physiological arousal and behavioral coordination can generate identity fusion (Páez et al. Reference Páez, Rimé, Basabe, Wlodarczyk and Zumeta2015), trigger the release of endogenous opioids associated with group bonding (Dunbar Reference Dunbar2010; Fischer et al. Reference Fischer, Xygalatas, Mitkidis, Reddish, Tok, Konvalinka and Bulbulia2014), and promote pro-social attitudes and behaviors (Xygalatas et al. Reference Xygalatas, Mitkidis, Fischer, Reddish, Skewes, Geertz, Roepstorff and Bulbulia2013), even among strangers (Lang et al. Reference Lang, Bahna, Shaver, Reddish and Xygalatas2017; Páez et al. Reference Páez, Rimé, Basabe, Wlodarczyk and Zumeta2015). In the context of a game, these effects are amplified by the size of the collective, the affordances for emotional contagion (Lakin et al. Reference Lakin, Jefferis, Cheng and Chartrand2003), and an abundance of symbolic group markers. What happens in the pitch, then, is, in this sense, secondary to the fusion process. The dynamics of the game provide emotionally salient external stimuli and goals to which the crowd is attuned, but die-hard fans are forged in the terraces, not on the couch. Fusion, in this context, is created through the “very act of congregating” (Durkheim Reference Durkheim1915).

So, which type of fusion are we dealing with in this case? According to the target article, local fusion is associated with imagistic practices, characterized by high arousal, dysphoria, and low frequency and performed among relatively fixed, small-scale, face-to-face groups (Whitehouse Reference Whitehouse2004). Extended fusion, on the other hand, involves alignment with large, anonymous communities and relies on less direct means, such as shared schemas, scripts, and doctrines not anchored in personal experience but rather acquired through social learning detached from any episodic event. Fusion among football fans does not seem to fall clearly into either mode, involving highly arousing, frequently repeated, ritualized events performed among massive, transient communities.

The question is important for the theory outlined here, as it is suggested (albeit tentatively) that local fusion can motivate extreme self-sacrifice, while extended fusion cannot. To resolve this matter, we need theoretical refinement as well as empirical data. Future research on identity fusion should manipulate its constituent parts independently to study their effects in isolation. For example, will attending a game as part of a small, face-to-face group rather than a large, anonymous crowd result in stronger fusion with the group? I would not be surprised if we found the opposite. Does attending a game in the stadium produce more fusion than watching it on television? I expect it would. And is dysphoric arousal especially potent in producing fusion among fans? A first attempt to answer this question (Whitehouse et al. Reference Whitehouse, Jong, Buhrmester, Gomez, Bastian, Kavanagh, Newson, Matthews, Lanman, McKay and Gavrilets2017) produced inconclusive evidence, as the study compared fans of different teams. Future designs should develop longitudinal measures focused on fans of the same team(s).

Importantly, is fusion a predictor of self-sacrificial behavior? So far, the extensive literature on fusion has relied almost entirely on self-reports on hypothetical scenarios, such as trolley dilemmas, that have little to do with the extreme behaviors the theory extrapolates. In fact, some of the findings may be more perplexing than enlightening. For example, college fraternity/sorority members who underwent initiation rituals later expressed a strong willingness to sacrifice themselves for the group (Whitehouse et al. Reference Whitehouse, Jong, Buhrmester, Gomez, Bastian, Kavanagh, Newson, Matthews, Lanman, McKay and Gavrilets2017). Given that we never see sorority members engage in suicide missions, this finding highlights some of the problems with such self-reported measures.

Finally, is the hypothesized distinction between local and extended fusion tangible? If so, what might be some of the behavioral and/or physiological correlates of each type? Could we find a behavioral signature for each type, for example, at the level of interpersonal space, postural mimicry, verbal interaction, aggression toward outsiders, and so on? Prior research has found that social proximity predicts higher synchronicity of autonomic responses and brain activity (Parkinson et al. Reference Parkinson, Kleinbaum and Wheatley2018; Xygalatas Reference Xygalatas2015). Might we observe similar differences between types of fusion?

Whitehouse is clearly aware of these outstanding issues, as well as the importance of framing the theory in terms of falsifiable predictions. It is now up to the research community to put those predictions to the test.