Pietraszewski's framework allows for the generation of novel hypotheses relevant to understanding group dynamics on social media platforms, where coalitions are prominent, problematic, and societally impactful. There is much debate on the precise nature of interactions of groups and individuals on social media, with often conflicting accounts (reviewed in Barberá, Reference Barberá, Persily and Tucker2020). Some research suggests that social media facilitates “echo chambers” of amplified like-minded views; others have found that social media facilitates increased exposure to cross-cutting, diverse information (Cinelli, Morales, Galeazzi, Quattrociocchi, & Starmini, Reference Cinelli, Morales, Galeazzi, Quattrociocchi and Starmini2021; Nguyen & Vu, Reference Nguyen and Vu2019). There is also debate regarding the specific effects of cross-cutting interactions, with evidence suggesting that such interactions can both reduce and exacerbate conflict (Paluck, Green, & Green, Reference Paluck, Green and Green2019). Understanding social media interactions through the lens of evolved coalitionary psychology may help to clarify and resolve many of these tensions.

Social media may amplify factors that exacerbate affective polarization, leading to intergroup conflict. Pietraszewski's model suggests that identity is the primary factor that scales up to define a group. The primacy of identity neatly accounts for the observation that social media (and, more generally, partisan media) tends to fuse political and social identity, as well as amplify identity awareness (Barberá, Reference Barberá, Persily and Tucker2020; Hutchens, Hmielowski, & Beam, Reference Hutchens, Hmielowski and Beam2019; Iyengar, Lelkes, Levendusky, Malhotra, & Westwood, Reference Iyengar, Lelkes, Levendusky, Malhotra and Westwood2019; Iyengar, Sood, & Lelkes, Reference Iyengar, Sood and Lelkes2012). Because of algorithms purpose-built to maximize “engagement,” social media disproportionately elevates the voices of the most polarizing and hostile individuals, granting them great influence over others (Brady, Wills, Burkart, Jost, & Van Bavel, Reference Brady, Wills, Burkart, Jost and Van Bavel2019; Iyengar et al., Reference Iyengar, Sood and Lelkes2012; Shore, Baek, & Dellarocas, Reference Shore, Baek and Dellarocas2018). The “clickbait” driven business models of social media companies incentivizes threat-related, emotionally charged, and moralizing messages. Such messages appear to “hack” our evolved psychology, with such messages spreading more virally than neutral or positive messages (Berger & Milkman, Reference Berger and Milkman2012; Blaine & Boyer, Reference Blaine and Boyer2018; Boyer, Firat, & van Leeuwen, Reference Boyer, Firat and van Leeuwen2015; Brady et al., Reference Brady, Wills, Burkart, Jost and Van Bavel2019). People tend to dislike opposing elites more than the average outgroup member, while simultaneously conflating their own views with their group's views. Consequently, repeated exposure to inflammatory messaging may be a salient factor contributing to increased affective polarization (Druckman & Levendusky, Reference Druckman and Levendusky2019). More generally, competition and perceived disadvantage may contribute to the elevation of risky, antisocial, and/or uncooperative behaviors both on and off social media (Mishra, Barclay, & Sparks, Reference Mishra, Barclay and Sparks2017).

For individual coalition members, social media allows for amplified social signaling. Actors can gain outsized reputational benefits by broadcasting to an audience far larger than any experienced in ancestral environments, thereby receiving disproportionate social reinforcement via “likes,” “reactions,” or other socially engineered forms of operant conditioning (Dos Santos & Rankin, Reference Dos Santos and Rankin2010; Petersen, Reference Petersen and Buss2015; Tooby & Cosmides, Reference Tooby, Cosmides and Høgh-Olesen2010). Although signaling is not necessarily related to conflict, the factors described above, along with a significantly diminished possibility of incurring retaliatory costs, may create an environment where rewards for hostile intergroup signals are artificially inflated while costs are artificially suppressed.

Pietraszewski's model suggests a set of falsifiable hypotheses regarding the causes of social media amplification of coalitional conflict. News organizations and other actors with financial incentives tied to social media engagement disproportionately broadcast moral-emotive or threat-related messages to their followers. These signals function to mentally co-register (coordinate) followers, increase emotional engagement, amplify the salience of group identity, and facilitate the adoption of one of the four group-roles (generalization, alliance, defense, and displacement) that the model identifies (Barberá, Reference Barberá, Persily and Tucker2020; Petersen, Reference Petersen and Buss2015; Tooby & Cosmides, Reference Tooby, Cosmides and Høgh-Olesen2010). The average group member is more inclined to act (consistent with one's evoked role, as predicted by the model) propagating conflict. From the outgroup's perspective, attacks increase threat perceptions, placing outgroup members in a defensive posture. This process makes the formation of a defensive coalition more likely, with an associated greater likelihood of preemptive attacks (Karlsen, Steen-Johnsen, Wollebæk, & Enjolras, Reference Karlsen, Steen-Johnsen, Wollebæk and Enjolras2017). This cycle becomes self-perpetuating. This hypothesized cascade can readily explain why cross-cutting information, typically associated with a reduction in conflict/polarization, has been recently found to have the opposite effect (Barberá, Reference Barberá, Persily and Tucker2020).

It would be illuminating to test the above cascade by examining social media behavior of different coalitions following a polarizing event that sets groups in opposition (a recent example being the murder of George Floyd). Such high-profile events offer salient opportunities for influential group members to coordinate a punitive coalition with moral-emotive or threat-related messages (Tooby & Cosmides, Reference Tooby, Cosmides and Høgh-Olesen2010). Such circumstances could also be used to investigate dynamics between groups that do not typically associate. For example, during a recent violent flare up in conflict between Palestine and Israel, leaders and news organizations tied to the Black Lives Matter movement shared posts in support of Palestinian causes with higher frequency, whereas those associated with a group often acting in opposition to Black Lives Matter activists – Donald Trump supporters – signaled pro-Israeli sentiments with higher frequency. Consistent with Pietraszewski's model, we would expect to observe a clear increase in willingness to impose costs on the opposing groups in support of their allies mediated by an increased identification with the allied group.

Another potential area of exploration relevant to Pietraszewski's model is the effect of removing prominent individuals from platforms. The model suggests that removing a prominent node may serve as an effective form of coalitional warfare by decreasing the opposing group's co-registration and eroding their group identity. However, target group members may see this act as an attack in and of itself, which would have the effect of strengthening group identity, allowing for the more expedient formation of a defensive coalition (Petersen, Reference Petersen and Buss2015). Both effects may also occur with group identity being strengthened in the short term based on the perceived threat, but weakened over the long term as a unifying voice is lost.

Pietraszewski's model specifically, and the lens of evolved coalitionary psychology more generally, may help to clarify group dynamics and resultant conflict on social media. These frameworks may help to clarify much of the uncertainty in understanding of causality that exists in the social media landscape, helping to engineer online environments that may help de-escalate conflict, rather than inflame it.