Intergroup contact and collective action : A match made in hell , or in heaven ?

ince Wright and Lubensky (2009) suggested that intergroup contact and collective action seem strategically incompatible when it comes to social change, social psychologists have been inclined to see their potential match as one made in hell, rather than in heaven. Against this backdrop, I review and discuss the contributions to this special issue, most of which seem to suggest that intergroup contact and collective action are a match made in heaven, not hell. To account for these seemingly divergent perspectives, I suggest that both intergroup contact and collective action be conceptualized as relational, interaction‐based phenomena within which the forces of harmony and conflict—intergroup contact and collective action, respectively—reflect two sides of the same coin, namely individuals’ need to regulate their relationships within their social network. As such, it is individuals’ embeddedness in (networks of) social relationships that determines whether intergroup contact and collective action work together, or against each other, toward social change. I discuss the need for a broader and integrative theoretical perspective that does justice to the underlying psychology of these phenomena in terms of relationship regulation.


| INTRODUC TI ON
This Special Issue of the Journal of Theoretical Social Psychology taps into one of the more interesting debates in social psychology over the last decade: Whether intergroup contact or collective action are strategically incompatible in fostering social change (Wright & Lubensky, 2009). As these authors put it (p. 293), "… they may not be compatible at all. In fact, the underlying psychology required by these two approaches may place them in direct conflict with each other." As a brief backdrop, the social psychology of collective action (e.g., Klandermans, 1997; for a meta-analysis see Van Zomeren, Postmes & Spears, 2008) suggests that there is collective agency among individuals to achieve social change, and that such action typically takes forms that are conflictual and antagonistic (e.g., social protest).
Through collective action, it is assumed, groups make use of their power in numbers while trying to achieve group goals, such as social change. The Black Lives Matter movement is a good example of such a collective and conflictual push for social change, which seeks to enforce equal rights and treatment through collective action. The real problem, in this view, is structural inequality and the real solution is to engage in intergroup conflict to enforce structural change.
Social change can also be achieved, however, by more harmoniously reducing the prejudice of those who hold it. In the context of Black Lives Matter, for example, one could argue that if only the prejudice toward Blacks would be reduced, the group would be treated equally and thus the movement would be obsolete. Specifically, theory and research on intergroup contact has long promoted the idea that positive and frequent contact between different groups may lead its members to forego their prejudices and treat each other in more equal ways (Allport, 1954;Pettigrew, 1998 for a meta-analysis, see Pettigrew & Tropp, 2006). Thus, the real problem, in this view, is individuals' prejudice and the solution is establishing positive and frequent contact between members of different groups in order to harmoniously reduce it.
The presumed incompatibility between the two seems to rely on a belief that the psychological forces of conflict and harmony are mutually exclusive-you cannot fight friends, and you cannot like enemies. Indeed, Wright and Lubensky (2009) argued that creating harmony between groups lowers the disadvantaged group's motivation and ability to achieve actual social change (coined the irony of harmony; e.g., Saguy, Tausch, Dovidio, & Pratto, 2009; see also Dixon, Durrheim, & Tredoux, 2005). Similarly, engaging in collective action may exacerbate intergroup differences and thus fuel, rather than reduce, prejudice (e.g., Reynolds, Oakes, Haslam, Nolan, & Dolnik, 2000). As such, although both approaches may appear to aim for similar goals, they also seem strategically incompatible. To only very slightly overstate this claim from the perspective of those valuing social change, the match between intergroup contact and collective action is one made in hell.
By contrast, the key message of this article is that the match between intergroup contact and collective action, at least when considering the contributions to this Special Issue, actually seems to be one made in heaven. Indeed, a number of contributions show that intergroup contact facilitates collective action (e.g., Carter et al., 2019Hoskin, Thomas, & McGarty, 2019Römpke, Fritsche, & Reese, 2019, which can, tongue-in-cheek of course, be considered the irony of the irony of harmony. Moving beyond these findings, however, I will also explain them by identifying the underlying psychology of both intergroup contact and collective action as one based in individuals' need to regulate social relationships in their social networks (Fiske, 1992;Van Zomeren, 2016). This is most clearly visible in the operationalization of intergroup contact as intergroup friendships (e.g., Carter et al., 2019;MacInnis & Hodson, 2019). Yet the very same observation has been made in the collective action literature, where a key predictor of participation in such action is: whether one has been asked to participate by a friend (e.g., Schussman & Soule, 2005). Yet, what seems to be lacking in the literature is the integration of such an emphasis on social relationships, social interaction, and social networks. I will develop this argument further with an eye to what I think is a dire need for a broader and more integrative perspective on how intergroup contact and collective action promote social change.
Before doing this, however, I will first review what we can learn from the interesting contributions to this Special Issue when it comes to social change. I will then outline a relational perspective on intergroup contact, collective action and social change and review some recent findings about intergroup contact and collective action that support such a perspective. Finally, I close with recommendations for the future by suggesting that we need to better understand when and why intergroup contact and collective action can be a match made in hell or heaven.

| THE CONTRIBUTI ON S TO THIS S PECIAL ISSUE
Each of the articles in this Special Issue have their own unique message to bring to the table, nicely fitting the call for papers on this topic. First, the article by MacInnis and Hodson, entitled Extending the benefits of intergroup contact beyond attitudes: When does intergroup contact predict greater collect action support?, reviews the paradox that "intergroup contact is generally associated with lower collective action participation and support for disadvantaged group members, but heightened collective action participation and support for advantaged group members" (p. 14). The authors discuss how being friends with outgroup members may have so-called "sedative" effects, decreasing perceived injustice and collective action motivation; while being friends with outgroup members may have what one may call "awareness-raising" effects, increasing perceived injustice and collective action motivation.
Importantly, on the basis of their review the authors propose that for intergroup contact to foster collective action among the disadvantaged and the advantaged, a certain contact threshold is required (with which they mean: potential intergroup friendships), as well as consensus about group differences and inequalities. The authors thus argue that intergroup friendships can be called upon to mobilize individuals for collective action, but only when all involved already agree about the need for social change. This suggests that we need harmony (e.g., contact) in order to mobilize for conflict (e.g., collective action). In fact, this may be why intergroup frienships are so important-individuals will be structurally invested in friendships more than more superficial forms of contact, and thus need to regulate this relationship more. The authors thus conclude that (p. 57): "students' involvement in collective action on campus is influenced, at least in part, by their perceptions of the injustice that marginalized students experience on campus, and the racial composition of students' friendship networks predict these perceptions of injustice." This is interesting in at least two ways. First, it supports the idea that individuals' need for relationship regulation-in terms of intergroup friendship-may be part of the underlying psychology of both sedative and awareness-raising processes. Furthermore, it explicitly links this to social networks, which provide structural opportunities to form and maintain such friendships. Second, it is also interesting to observe what is not explicitly mentioned-for example, the article does not report or discuss other aspects of psychology related to collective action, such as individuals' group identification and their group efficacy beliefs-yet these are core motivations for collective action (Van Zomeren et al., 2008), and arguably the more unifying and empowering aspects of engaging in collective action (Drury & Reicher, 2009). As such, the Carter et al. article does not speak to whether intergroup contact can unify and empower individuals toward fighting for social change.
The latter is important because, third, Thomas and colleagues offer an answer to this question in their contribution, entitled Transnational contact and challenging global poverty: Intergroup contact intensifies (the right kind of) social identities to promote solidarity-based collective action for those low in social dominance. In their study, they offer a longitudinal test of the idea that (transnational) contact promotes (solidarity-based) collective action because it intensifies supportive (opinion-based) social identities among advantaged group members.
These authors draw on the notion of (opinion-based) group identities, which basically entail MacInnis and Hodson's second criterion-that of consensus about a need for collective action. Indeed, Hoskin et al.'s notion of opinion-based group identity is based on opinions about how the world should be, and by identifying with such a psychological group, all involved effectively share the same opinion.
Contact, in their view, "intensifies" that shared reality and thus motivation for collective action. They find support for these ideas in their study, although only for individuals who score low on Social Dominance Orientation, which I interpret as meaning that they would be very likely to score high on valuing social change toward inter- This line of thought seems to imply different types of "we" in the underlying psychology of intergroup contact and collective action.
One implies intergroup friendships within which "we" think alike (for instance about the need for social change), whereas the other implies psychological group membership within which "we" think alike (for instance through identification with an opinion-based group).
Yet it remains unclear, from the contributions in this Special Issue at least, how these two types of "we" should be conceptualized and understood in conjunction. with an eye to social change. Indeed, to this end I believe we need to start taking each type of "we" seriously to the same degree.
The tendency toward category-based forms of "we" is continued in the fourth contribution by Römpke et al., entitled Get together, feel together, act together: International personal contact increases identification with humanity and global collective action. The authors propose that although intergroup contact may have sedative effects in intergroup conflicts, such contact may actually be ideal for bringing people together from different groups to act collectively against global crises (e.g., climate change). They predict and find across a number of studies that individuals engaging in international contact identify more strongly with "all of humanity" and are more willing to engage in relevant forms of action.
As such, intergroup contact and collective action once more seem joined at the hip, in this case through the notion of transnational contact and identification with all humanity. At the same time, it is important to note that global crises are fundamentally different from the contexts within which Wright and Lubensky (2009) identified the strategic imcompatibility between intergroup contact and collective action. In fact, it is difficult to see the same tension between harmony and conflict when focusing on a collective problem relevant to everyone involved. Moreover, one can wonder about the meaning and sustainability of group identities that comprise all "humanity." In the absence of a clear outgroup, one might expect that individuals need to feel distinct within such an abstract, encompassing psychological group. Once more, then, the underlying psychology here requires a view on different types of "we." Finally, Zagefka's paper is entitled Triadic intergroup relations: Studying situations with an observer, an actor, and a recipient of behaviour. Her key theoretical observation is that intergroup contact and collective action researchers typically conceive of social inequality in dichotomous ways, for instance, reducing such a context to members of disadvantaged and advantaged groups. Zagefka argues convincingly that unlocking a triadic perspective, although modestly increasing the number of collective actors from just two to just three, already offers more scope for understanding contact, collective action, and social change.
I agree whole-heartedly with this suggestion, not in the least because this raises rather fundamental questions about what type(s) of "we" can be expected in contexts that move beyond dichotomous intergroup relations. Although some may argue that such contexts evoke an even stronger psychological need for individuals to categorize the world in ingroup-outgroup dichotomoties (e.g,. Subašić, Reynolds, & Turner, 2008), the emphasis in the other contributions on intergroup friendships may suggest that more complex contexts evoke a need to regulate relationships within the social networks we are part of. To this end, in the next section I will describe a recent line of research (Klavina & Van Zomeren, 2018) that shows how intergroup contact facilitates collective action in such a triadic context.

| A REL ATI ONAL PER S PEC TIVE
In this section, I apply a relational perspective on what moves and motivates individuals in their lives to the underlying psychology of intergroup contact and collective action. In doing so, I conceptualize both intergroup contact and collective action as different forms of relationship regulation, and thus of regulating potentially different types of "we" (be it friendships or psychological group memberships).
Whereas the rationale for conceptualizing collective action as relational interaction can be found in detail elsewhere (Van Zomeren, 2014, 2015, 2016, the case for intergroup contact seems even more clear-cut, especially given the operationalization of such contact as intergroup friendship (Carter et al., 2019MacInnis & Hodson, 2019. The simple observation in either case is that both phenomena occur within (networks) of social relationships, and would be hard to imagine outside of them.
Social psychology is rife with pointers toward the pivotal importance of social relationships, social interaction, and embeddedness in social networks for individuals' health and happiness.
Indeed, humans are an ultra-social species with a strong need to belong (Baumeister & Leary, 1995). Individuals can empathize with others in need (Batson, 1990), suffer psychologically and physically from loneliness (Cacioppo & Patrick, 2008), and display attachment behavior regulate relationships long before our self-concept arises (Ainsworth, 1979(Ainsworth, , 1989Bowlby, 1969Bowlby, , 1973Bowlby, , 1980Mikulincer & Shaver, 2007a, 2007b. Social interaction and the development of relational ties provide us with "safe havens" (to seek shelter in; e.g., Van Zomeren et al., 2016) or "secure bases" (on which to explore the world), which need to be maintained and regulated (Fiske, 1992).
A relational perspective offers a number of additional insights.
First, embeddedness in social relationships imply restrictions to individuals' degrees of freedom. In this view, individuals are assumed to prioritize maintaining relationships over terminating them, which means that social relationships require continuous and effortful regulation. For instance, expecting others to be available and responsive to your needs also implies that others expect you to be available and responsive to their needs-an obligation that is called upon when, for example, a friend asks you to come to a demonstration next week.
A second addition is that relationship regulation implies forces of harmony as well as conflict-in this view, these are different sides of the same relational coin. For instance, if maintaining harmony within the group (e.g., one's family) is more important to a Black individual than risking conflict with Whites in one's neighborhood, then there is no necessary tension between engaging in collective action together against racial discrimination. Fiske and Rai (2015) even went as far as to suggest that our embeddedness in social relationships is so fundamental that it also explains why people turn violent. For instance, honor killings are, in this view, not acts of mad men, but of individuals trying to regulate their relationships within the group (e.g., defending the family reputation) through such violence against others (that threaten the family reputation).
Taken together, a relational perspective offers an underlying psychology of intergroup contact and collective action that may help to explain why their match can be one made in hell, or in heaven.
This perspective acknowledges that there is one type of "we," for example derived from intergroup friendship, that is fundamentally different from another type of "we" that is derived from psychological group membership. Individuals' regulation of the first type of "we" needs to fit with their regulation of the second type of "we," if we want intergroup contact and collective action to be a match made in heaven (such as when participation in collective action is

| EMPIRI C AL ILLUS TR ATI ON S OF A REL ATIONAL PER S PEC TIVE
In this section, I illustrate a relational perspective with relevant empirical findings. For instance, individuals' participation in collective action depends crucially on the social networks in which they are embedded (Schussman & Soule, 2005; see also Klandermans, Van der Toorn, & Van Stekelenburg, 2008). Their embeddedness in specific activist networks, for example, offers them structural opportunities to participate in collective action, as one's friends may frequently ask one to join an action (and one may ask others in the network in turn). This is a clear example of relationship regulation in (collective) action, suggesting that individuals' hearts and minds are certainly important, but that our hearts and minds are intimately connected to those of others around us. Furthermore, the relationships in our networks provide us with a safe haven to seek shelter in, or as a secure base to explore (and potentially change) the world around us.
Activist networks, however, are rather rare, as most people are not embedded in such networks. So what happens in non-activist networks? Van Zomeren et al. (2016) surveyed Indonesian ethnic minority members in two studies, linking individuals' seeking shelter in their social relationships (i.e., turning to friends and significant others in times of trouble) and their ethnic group identification to wellbeing and collective action against ethnic discrimination. Results across the two studies showed that despite perceiving their group to be discriminated in society, individuals' shelter-seeking positively predicted their well-being above and beyond their ethnic group identification; furthermore, neither of those variables predicted their collective action tendencies. Thus, whereas activist networks may be great mobilizers because of individuals' need for relationship regulation, non-activist networks may be great harmonizers for the very same reason. Importantly, in either case ethnic group identification did not explain unique variance in these psychological processes.
Furthermore, Górska, van Zomeren, and Bilewicz (2017)  The underlying psychology here, I would argue, is one in which (changes in) social structure affords individuals' relationship regulation, with prejudice reduction and collective action as potential downstream consequences.
Finally, let me describe a set of three studies we conducted (Klavina & Van Zomeren, 2018) on the underlying psychology of intergroup contact and collective action. In this line of work, Zagefka (2019) would be overjoyed to see that we focused on triadic contexts in which intergroup contact and collective action were both psychologically and contextually relevant. We examined whether the same core motivations for collective action that we typically find to apply to disadvantaged and advantaged group members (i.e., group identification, anger, and efficacy beliefs; Van Zomeren et al., 2008) also apply to "third groups" -groups outside of an unequal intergroup relationship (e.g., Blacks and Whites) that nevertheless could be psychologically or actually affected (e.g., Latinos in terms of police brutality in the US). The research question we asked across three studies with different triadic contexts was whether third group members would be motivated for collective action to protect their own ingroup from future disadvantage, or to protect the disadvantaged outgroup from actual disadvantage. Furthermore, we explored how intergroup contact (with the disadvantaged outgroup) would affect individuals' motivation for collective action on behalf of that group.
The third and final study of this set was conducted in the US, with Latino participants eying the intergroup relationship beween Whites and Blacks. But the other two studies had perhaps lesserknown contexts that require somewhat more explanation. Study 1 was conducted in Latvia, after the Russian annexation of the Crimea (formerly in the Ukrain). We thus asked Latvians about their motivations for collective action in order to protect themselves, or to protect the Ukrainians. Similarly, in Study 2 we used a Dutch context within which a government-associated gas extraction company came under scrutiny after the emergence of earthquakes in the populated gas extraction area in the North of the Netherlands. We surveyed those close to, yet still outside of, the affected areas about their willingness to engage in collective action to protect their own region from becoming part of those areas in the future, or to protect those in the affected areas.
Across the three studies, and consistent with previous work, we found that individuals' willingness to engage in collective action to protect their ingroup was predicted by their ingroup (rather than outgroup) identification, anger, and efficacy beliefs. Thus, Latinos were more likely to act to protect Latinos against police brutality in the US, when they identified more strongly with their ingroup (measured with often-used items such as "I identify with Latinos"). Furthermore, we found that individuals' willingess for collective action to protect the disadvantaged outgroup was predicted by their outgroup (rather than ingroup identification, anger and efficacy beliefs. Thus, Latinos were more likely to act to protect Blacks against police brutality in the US, when they identified more strongly with that outgroup (measured with the very same often-used items yet applied to this outgroup, such as "I identify with Blacks"). The same core motivations for collective action thus applied to third group members' collective action intentions.
We then explored the role of intergroup contact, especially with an eye to the notion of "outgroup identification" (which, in the theoretical universe of social identity and self-categorization theory, cannot and hence does not exist, as identification with a group implies this group psychologically being or becoming an ingroup). Across the studies, we consistently found that intergroup contact (measured as frequent and positive contact with outgroup members) with the disadvantaged outgroup positively predicted identification with this outgroup, which in turn positively predicted collective action intentions. Furthermore, across the studies we also found that individuals' so-called relational models (i.e., their endorsement of specific ways of relationship regulation such as communal sharing and equality matching; Fiske, 1992)

| LOOKING FORWARD
So far I have looked back on the debate sparked by Wright and Lubensky (2009)  Such a multi-level approach to intergroup contact and collective action will also be informative for empirically examining Wright and Lubensky's (2009)  What this account does assume, however, is that psychological groups are formed through regulating relationships within the networks in which individuals are embedded. As a consequence, individuals are able to identify with outgroups (without self-categorizing as part of them or as part of a superordinate category), as discussed in the studies with the triadic context (Klavina & Van Zomeren, 2018). What group identity means, in this view, is thus dependent on the networks within which people put them to use through relationship regulation.
One important implication of this line of thought is that collective action can be viewed as a (safe) form of intra-group contact, which develops individuals' ingroup identity as a secure base on which to explore and change the world (Bowlby, 1969;Van Zomeren et al., 2016).
Similarly, intergroup contact can be viewed as an expansion of one's network with outgroup members-however, incidental forms of intergroup contact (outside of one's comfort zone) may differ fom more structural forms of intergroup contact that bring a stronger psychological investment to the relationship (such as intergroup friendships, within one's comfort zone), and this may precisely be why the latter seems more effective in reducing prejudice. Thus, we need a better understanding of the different types of "we" toward reaping the proverbial best of both worlds with an eye to social change.
Third and finally, a relational approach adds considerably to the intergroup contact and collective action literature by implying that psychological theories about social relationships and relationship regulation are missing from these literatures without good reason (Mackie & Smith, 1998). In fact, theoretical integration between these perspectives may help substantiate the "relations" in "intergroup relations," and force researchers to include, rather than to preemptively dismiss, measures of different types of "we" in studies of intergroup contact and collective action. This may help explain why scholars are arguing more and more for structurally positive and frequent forms of intergroup contact (such as close friendships); and why scholars are coming to the conclusion that participation in collective action is best predicted by the structural availability of others (in one's network) who ask you to be responsive and put your money where your mouth is.
One concrete example of this is including the role of attachment style, or more precisely, the working models people develop over the course of their lifetime that suggest what can be expected from others and the self in terms of availability (of/for others) and responsiveness (Bowlby, 1969). Indeed, Reis and Gable (2015) concluded that responsiveness is the key predictor of what makes close relationships harmonious. Hence, a perceived lack of availability and responsiveness of the other (or the perceived obligations about availability and responsiveness that others may force upon you) is what makes them conflictuous. A concrete suggestion for future research, then, would be that individuals' attachment styles and working models should be strongly predictive of both intergroup contact and collective action.

| CON CLUS ION
I am certainly not the first to suggest that theory and research on group processes and intergroup relations can learn something from theory and reseach on interpersonal processes, and vice versa (e.g., Mackie & Smith, 1998). I am also certainly not the first to suggest that social networks and social structure matter in intergroup contact and collective action (e.g., Gorska et al., 2017;Klandermans et al., 2008), and that we need a more diverse theoretical and methodological repertoire in order to better understand whether, for example, intergroup contact and collective action work against, or together with, each other (Pettigrew, 1998).
I am probably the first, however, to be in a position to discuss the interesting contributions to this Special Issue, and from that position I have expressed pretty much all the above things. The key message here is that Wright and Lubensky's (2009) argument about intergroup contact and collective action as a match made in hell does not seem to do justice to their underlying psychology in terms of relationship regulation, which can also lead to their match as one being made in heaven. Indeed, those who engage in collective action and intergroup contact are ultimately engaged in regulating relationships with other people around them. Perhaps, then, the current Special Issue signals that the time has arrived to start treating them as such.