Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Psychology. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 26 January 2021

Communication and Intergroup Relationsfree

  • Jake HarwoodJake HarwoodDepartment of Communication, University of Arizona


The subfield of communication and intergroup relations attempts to disentangle the ways in which human message exchange is influenced by, and itself affects, relations between social groups. Typically, the social groups considered are large scale groups (e.g., national, religious, ethnic groups), but similar processes can also be applied to smaller groups such as families or work groups. Specifically, the field of communication and intergroup relations considers how social interaction is changed when the interlocutors belong to (or perceive themselves as belonging to) specific social groups, and how everyday talk about groups changes perceptions and attitudes concerning those groups. The subfield also considers how broader societal messages relate to group memberships. For instance, how do media messages reflect the macrosocial position of particular groups, and do media messages influence how consumers think about group memberships and intergroup relations? Underpinning all study of intergroup communication is the belief that intergroup relations are forged, perpetuated, and modified in real-life everyday social communication.

Basic Principles and Assumptions

This article addresses communication as a fundamental process to understanding intergroup relations. The article first elaborates on this point, outlining some assumptions concerning communication that shapes work that falls under the “intergroup communication” umbrella. Subsequently, the article examines two very broad areas that capture the vast majority of research in intergroup communication. First, the article examines research on language and interpersonal communication—communication typically occurring on a local, “face-to-face” level. Second, research on mass-mediated communication is addressed, thus switching to a more macro-level analysis of communication’s influence. The article closes with a consideration of future research directions in the intergroup communication area.

A foundational assumption for scholars working in communication and intergroup relations is that messages are central to the nature of intergroup relations. Underlying this core principle, it is possible to identify a set of organizing assumptions with which most intergroup communication scholars would agree.


Messages express intergroup perceptions. When we talk, we reveal how we feel about groups, what we believe about groups, and how we understand intergroup relations. This may occur explicitly (“I hate group X”), or implicitly (e.g., via the types of story one chooses to re-tell or not about certain groups: van Dijk, 1987). In some cases, our verbal and nonverbal channels may tell different stories about our intergroup perceptions, with the nonverbal sometimes revealing “correct” attitudes, but also being subject to more dramatic misconstruals (Dovidio & Gluszek, 2012).


Messages construct intergroup perceptions. Not all messages are intentionally designed to be persuasive, but most messages have the potential to change attitudes or perceptions. As children, we are socialized by listening to what people say, and our socialization into intergroup relations in society comes through hearing what others say about groups. This process doesn’t stop at the end of childhood, and particularly as new groups become salient in our environments, we develop new understandings of those groups from the ways in which they are talked about. Indeed, it is through messages that we come to understand which groups matter: We pay attention to particular groups because people talk about them, while other groups are not discussed and hence ignored.


Messages are multilevel. The kinds of “talk” we encounter occur at many different levels of analysis. Certainly our interpersonal interactions (e.g., in a family) are important, but group relations are also expressed and constructed in broader political discourses, in mass communication, and the like. Throughout this article, the term “interpersonal” is used to indicate social interaction between a small number of individuals—typically a dyad or very small group (e.g., email exchanges between work colleagues, a chat between strangers on an airplane, spouses arguing).


Messages at one level influence other levels. Messages at a mass societal level establish the relevant vocabularies and themes for interpersonal talk—mass communication serves an “agenda setting” function for what matters on an intergroup level (which groups, what issues) (Scheufele, 2000). At the same time, interpersonal communication contributes to the media agenda, and can influence directly the content of media messages. For example, interpersonal influence from pressure groups such as the AARP or NAACP pressures media outlets to feature groups more positively.


Messages define intergroup belonging. At the simplest level, for many groups communication is essential to knowing about group membership. We often learn that someone is gay through them “coming out” to us: Many “invisible” group memberships operate similarly (Fasoli, Maass, & Sulpizio, 2016). For other types of groups, boundaries between membership categories are negotiable. For example, who counts as an “old person” relies on an arbitrary line being drawn on a continuum: Where and how that line gets drawn are figured out through social interaction about social categorization (Giles & Reid, 2005). Communicative cues can also be central to the categorization process when language signals group membership (e.g., via use of a particular language, an accent, or specific lexical or grammatical usage: Dragojevic, 2016; Gluszek & Dovidio, 2010). This process can be conscious and intentional, but also unconscious and even covert. People sometimes shift their language to attempt to “pass” as a member of a different group (Reid & Ng, 1999). Other times, people’s group memberships may be unwittingly disclosed by a single communicative “slip”—a shibboleth (McNamara, 2005).


Communication is where intergroup relations happen. While it is possible to hear about and learn about how groups get along in an abstract sense, many of our most vivid experiences of intergroup relations occur when we directly observe or participate in intergroup contact. Moments of intergroup conflict, harmony, discussion, negotiation, and intimacy are all moments of communication. Even apparently depersonalized events (e.g., drone strikes) are typically the culmination of a long history of increasing hostility, much of which will have occurred through intergroup communication processes.


Almost all communication has intergroup elements. While it might be possible to imagine some message exchange that has no intergroup component, such events are rare. For the most part, our interactions—even the most intimate and apparently “interpersonal”—feature elements that reflect our group memberships. Heterosexual marriages are shaped by the gender dynamics of the participants, close friendships are very frequently between individuals sharing many group categories, and our ability to “ignore” group memberships is very limited (Giles, 2012).

Hence, the abstract concepts of social psychology—prejudice, stereotyping, entitativity, identity, and the like—have living, breathing reality in human message exchange. Groups are not static entities into which we are categorized and with which we identify (or not). They are dynamic shape-shifting constructions that ebb and flow in relevance and value with the ebb and flow of social discourse. With these thoughts in mind, it is perhaps worth noting that most work on intergroup communication has focused on racial and ethnic groups, with a good amount of research also focusing on gender groups, language/national groups, and age groups, and less work on religious or political groups, or groups defined by (dis)ability or sexuality (Giles, 2012). The trends as to which groups receive research attention and which do not is perhaps itself a product of communicative activity, as we build convincing arguments that certain groups deserve attention or as the media establish one group’s situation as worthy of consideration. One area that probably deserves more work across the social psychology and communication research communities is the extent to which research findings and theory from one intergroup arena apply to other intergroup contexts; in principle theories in this area should be applicable across different intergroup contexts, but in practice the explicit examination of this form of generalization is rare. Our theories of interracial communication, for instance, have clear applicability to communication between gender groups or age groups, but the specific histories, stereotypes, and inherent features of groups undoubtedly change important dimensions of intergroup interaction (Deaux, Reid, Mizrahi, & Ethier, 1995; Lickel, Hamilton, & Sherman, 2001).

The remainder of this article overviews contemporary research and theory in intergroup communication focusing on the two broad areas that have dominated research in the field: (a) language and interpersonal communication, and (b) mass communication.

Language and Interpersonal Communication

Intergroup Relations Influence Language

Considerable research in communication and social psychology has examined the ways in which intergroup relations (attitudes, stereotypes) affect message production, and particularly the linguistic details of message production. Probably the most productive paradigm in this area is work on linguistic intergroup bias (Maass, Salvi, Arcuri, & Semin, 1989) and the associated linguistic expectancy bias (Wigboldus, Spears, & Semin, 2005). Most broadly, the linguistic intergroup bias suggests that we frame linguistic details of messages in terms that reflect our group-based perceptions. This research builds on the linguistic category model (Semin & Fiedler, 1991). The linguistic category model shows that we can describe the same event using four linguistic categories. If we observe one person hugging another, we can say that “Judy hugged Bob” or we can say that “Judy is affectionate.” The former (a concrete verb: hugged) describes the event in a descriptive and transient manner, while the latter (an adjective: affectionate) implies some more stable and enduring characteristic of the person. Work on the linguistic intergroup bias shows that we are likely to use the more stable linguistic forms when describing positive ingroup and negative outgroup behaviors, but to use the more transient descriptors to talk about negative ingroup and positive outgroup behaviors (Maass et al., 1989). Thus, linguistically, members of the outgroup doing bad things are portrayed as (dispositionally) bad people, while a bad action by a member of the ingroup is linguistically framed as an isolated one-off event. Such linguistic framing does not merely represent bias—the use of such language results in inferences about both the identity of the person being described (Porter, Rheinschmidt-Same, & Richeson, 2016) and their likely characteristics (Werkman, Wigboldus, & Semin, 1999).

Attitudes also manifest more explicitly in language in the form of ethnophaulisms. These are disparaging words used to describe an ethnic group or an individual-as-group-member. Ethnophaulisms are one component of a larger category of antilocutions—speech against groups—which would encompass the equivalents of ethnophaulisms for non-ethnic groups (e.g., ageist, sexist, homophobic slurs). Ethnophaulisms and antilocutions are hate speech—an even broader concept that would also encompass jokes that target particular groups, for instance (Haas, 2012). In some countries the use of hate speech has legal ramifications—such speech is expressly forbidden. In others (including the United States), such speech is broadly protected under free speech constitutional provisions, unless the speech explicitly advocates violence toward another group, for instance. Ethnophaulisms vary in complexity and valence, with research suggesting that smaller (demographically) and more foreign (i.e., less similar to the host culture) ethnic groups are represented by more negative and less complex terms (Mullen, Rozell, & Johnson, 2000). The use of ethnophaulisms appears to predict the exclusion of certain groups from society (Mullen & Rice, 2003), and is associated with negative effects for the targeted groups (Mullen & Smyth, 2004).

In a sense, the “opposite” of hate speech is the idea of “politically correct” speech. Political correctness broadly represents a set of norms for how to communicate about particular groups (or more broadly political issues) in ways that are respectful of all. Behind this general idea resides an ideological battle over which ways to communicate actually are offensive, and which are appropriate (Robinson & Reid, 2016), as well as broader concerns over the extent to which speech can or should be regulated (i.e., freedom of expression versus censorship).

Two final areas at the interface of communication and intergroup language use are worth mentioning here. A continuing debate on the use of the generic masculine (using “men” to refer to all people, or “he” to refer to someone of unknown sex, for instance) has received some empirical attention. For the most part, such investigations suggest that generics serve to exclude women and reinforce traditional gender hierarchies (Stahlberg, Braun, Irmen, & Sczesny, 2007). Such issues become considerably more complex when considering the broader ways in which gender can be marked across languages (e.g., the absence of gender-based pronouns in Mandarin, the gendering of articles in German) and the consequences of those linguistic features for the salience of gender (Gabriel & Gygax, 2016). Finally, for groups that are invisible, communication serves a fundamental role in disclosing group membership and identity to others. “Coming out,” for instance, refers to the personal disclosure of sexuality, and “outing” to the involuntary disclosure of someone else’s sexuality (Fasoli et al., 2016). Both are communicative acts fraught with rhetorical challenges. While less discussed, the same issues apply to disclosure of religion, mental illness, transgender status, and numerous other group/identity-relevant dimensions of personal status. The social consequences of such disclosures can be mixed: Disclosure might trigger anxiety or discrimination in the recipient, or greater sensitivity to, and more positive attitudes toward other groups (Day & Schoenrade, 1997; Gross, Green, Storck, & Vanyur, 1980). Relatively little work has examined the communicative processes of coming out, or how such processes are influenced by attitudinal or relational characteristics (but see Denes & Afifi, 2014).

Many more specific pieces of work examine particular communicative “moments” and how they illustrate or reveal group perceptions and attitudes (“Some of my best friends are . . .,” “I’m not prejudiced, but . . . “Sure, but where are you really from?”: Bonilla-Silva, 2002; Sue, 2010). More conceptual work is needed to provide broader theoretical frameworks for the immense variety of ways in which everyday talk reveals prejudice (see Gallois, Cretchley, & Watson, 2012) for more detail on the methodological challenges in examining such language use).

Intergroup Relations Influence Interpersonal Relations

Considerable work has examined the influence of intergroup relations on interpersonal communication—particularly interpersonal communication between members of different social groups. The most enduring theoretical perspective in this area is communication accommodation theory (Giles & Gasiorek, 2013). This theory began as a way to account for the adjustments we all make in our speech based on our interlocutor. We all tend to converge our language toward the style of a conversational partner, particularly when we are seeking approval or are attracted to that person. Similarly, we tend to diverge when we are seeking distinctiveness or are in conflict with the partner (Giles, Taylor, & Bourhis, 1973). Such convergence or divergence occurs across a wide array of linguistic dimensions (e.g., speech rate, pitch, accent, length of utterance, etc.). In bilingual/multilingual contexts, accommodation can manifest as shifting completely from one language to another (e.g., shifting from English to French in bilingual Quebec). When applied to intergroup communication, accommodation strategies reflect intergroup dynamics in complex ways. For instance, we diverge from members of groups that we don’t like, particularly so when the interaction raises the salience of group memberships or when an outgroup partner is seen to be threatening our ingroup identity (Bourhis & Giles, 1977).

We also shift our language use in more subtle ways, depending on the specific content of intergroup perceptions. So, for instance, younger people will often “overaccommodate” older adults. Overaccommodation is a linguistic strategy that goes beyond the necessary accommodations and often accommodates to a stereotype of the other person based on dependency or incompetence (Ryan, Giles, Bartolucci, & Henwood, 1986). Thus, younger people are often seen to talk in a patronizing or “baby talk” manner to older adults, reflecting a stereotype of older adults as needing care, and an emotional response of pity (see Cuddy, Fiske, & Glick, 2007, for more detail on links between stereotypes, emotion, and interpersonal behavior). Such communication is often viewed as demeaning, and may result in the recipient being viewed in a manner concordant with the speech (“blaming the victim”: Giles, Fox, & Smith, 1993).

A second approach to intergroup interaction has focused on the specific communicative dynamics as reflecting various aspects of stress, anxiety, or avoidance. Trawalter, Richeson, and Shelton (2009) synthesize this into an innovative framework involving threat, stress, and coping. In particular, they explain how individuals’ orientations to intergroup interactions will vary depending on the extent and type of threat experienced, and their perceived and actual resources for coping with that threat, resulting in responses along traditional categories of fight, flight, freeze, or tend and befriend.

Interestingly, members of different groups may be anxious about different things in interaction. For instance, in the United States, interracial encounters may be influenced by social stereotypes that Whites are prejudiced and that Blacks are incompetent or unintelligent. As such, Blacks may be concerned about self-presenting as competent, while Whites wish to self-present as likeable or moral (Bergsieker, Shelton, & Richeson, 2010). Such divergent goals can result in interactions during which individuals from different groups operate at “cross-purposes” and emerge dissatisfied. Similar patterns of effects are apparent in terms of preferences for discussing racial differences and inequalities: Whites prefer to avoid such topics, while Blacks prefer to engage them (Saguy, Dovidio, & Pratto, 2008). This work extends beyond simple examinations of groups not liking one another, and illustrates how group differences can impede effective communication even when all the parties involved may have positive orientations and intentions.

Interpersonal Relations Influence Group Attitudes

Since Allport’s (1954) classic statement of the importance of contact between groups in propagating positive intergroup relations, a massive literature has emerged demonstrating the truth of that basic proposition (Pettigrew & Tropp, 2006). That literature now includes considerable work showing the importance of specific features in determining positive intergroup contact (see the next section, “Factors Implicated in Positive Intergroup Relations,” for a broader discussion of which factors encourage positive intergroup interaction). The positive effects of intergroup contact on reducing prejudice largely occur through affective channels, and particularly through the ability of positive and frequent intergroup communication experiences to reduce intergroup anxiety (Stephan, 2014).

Research has consistently demonstrated that contact needs to retain some minimal level of group salience in order for attitudes to generalize from an individual to an outgroup as a whole (Brown & Hewstone, 2005). If, on one hand, I engage an outgroup member on a purely “interpersonal” level, I may like her but not infer anything about her group from my liking of her as an individual. On the other hand, if I am aware of her group membership while in contact with her, I am more likely to infer something about her entire group from my experience with her (if I like her, I will like her group). Unfortunately, research also demonstrates that awareness of group memberships can have negative consequences for intergroup interactions—group salience predicts less positive intergroup communication experiences (Paolini, Harwood, & Rubin, 2010), and negative contact does not improve intergroup attitudes (Hodson, Hewstone, & Swart, 2013). Beyond the narrow confines of members of different groups engaged in active face-to-face conversation, the contact literature is broadening to allow for less direct forms of contact. In interpersonal interaction, a promising line of work is that considering computer-mediated interaction. New technologies provide the ability for “thinking before talking” that is often precluded in regular talk, and also allow conversation at a distance (critical in segregated or hostile contexts) (Amichai-Hamburger, 2012; Walther, 2009).

Factors Implicated in Positive Intergroup Relations

The number of variables that are implicated in positive outcomes of intergroup interaction are too numerous to mention here. However, a limited number of particularly important variables are worth attention. Four subcategories are considered.

Individual Differences

At the level of the individual, prior intergroup contact is a critically important variable. Indeed, even if prior intergroup contact has been anxiety provoking, contemporary models indicate how such contacts, over time, can accumulate to reduce anxiety in future encounters and increase perceptions that intergroup encounters are manageable and potentially positive (MacInnis & Page-Gould, 2015; Paolini, Harris, & Griffin, 2016). Related, individuals with more positive attitudes toward the outgroup are typically better equipped to manage intergroup encounters well (McConnell & Leibold, 2001; cf. Shelton, Richeson, Salvatore, & Trawalter, 2005).

Relational Features

Considerable evidence (and intuition) demonstrate that encounters in established and intimate relationships are more successful than those between strangers (Van Laar, Levin, Sinclair, & Sidanius, 2005). Beyond this rather obvious point, intergroup interactions tend to go more smoothly and have more positive outcomes when the participants are of equal status, and when they have some probability of future interaction (Pettigrew & Tropp, 2006).

Interaction Features

The nature of the interaction context is critical to predicting positive versus negative outcomes. Structured forms of interaction are more effective than unstructured settings. For example, unstructured “get to know each other” conversations elicit high levels of anxiety (Stephan, 2014; Towles-Schwen & Fazio, 2003). While beyond the scope of the current article, the medium of interaction is also critical. As noted earlier, mediated communication channels (e.g., the Internet) can reduce intergroup anxiety and result in different (and perhaps more positive) intergroup contact dynamics (Amichai-Hamburger & McKenna, 2006; Harwood, 2010).

(Macro) Social Context

Finally, interpersonal scholars must retain awareness of broader sociopolitical contextual considerations. Situations characterized by high levels of intergroup conflict and a high degree of group segregation will result in higher levels of anxiety and a reduced likelihood of positive intergroup interaction (Stephan, 2014). In such contexts, structured, long-term contact via mediated channels is particularly beneficial to achieve positive intergroup interactions (Walther, 2009). Broader patterns of intergroup suspicion or lack of trust will also impair opportunities for positive social interaction; in particular, it is important to be aware that a person’s anxiety about intergroup interaction might stem not only from concerns about outgroup behavior, but also from concerns about how fellow ingroup members might respond to seeing one engaged in intergroup contact (e.g., potentially being judged as a traitor or a black sheep: Belavadi & Hogg, 2015; Marques, Yzerbyt, & Leyens, 1988).

Mass Communication

Group Vitality and Media Portrayals

Group vitality was established by Giles, Bourhis, and Taylor (1977) as indicating the likelihood of an ethnolinguistic group surviving. Since Giles et al.’s original conceptualization, the term has broadened to encompass a more global measure of a group’s strength and status in society. Vitality is indicated by the group’s: (a) demographic strength (relative population size, growth/decline in population, intermarriage, etc.); (b) status (socioeconomic position, perceived position in the social hierarchy); and (c) institutional support (presence of the group in important religious, political, educational, etc., societal institutions). The media are an important social institution, and hence the presence and portrayal of groups in a culture’s media is a powerful indicator (and determinant) of the group’s vitality (Harwood, Giles, & Bourhis, 1994).

Content analytic examinations of group presence and portrayal are common in the field of communication, particularly with regard to U.S. media. These studies tend to show fairly consistent patterns that minority ethnic groups (Dixon & Linz, 2000), older adults (Zhang, Harwood, Williams, Ylänne-McEwen, & Thimm, 2006), women (Gerding & Signorielli, 2014), and sexual minorities (Raley & Lucas, 2006) are underrepresented (present in numbers smaller than their true population numbers) and negatively portrayed. For instance, Tukachinsky, Mastro, and Yarchi (2015) show serious underrepresentation of Asian Americans, Native Americans, and Latinos in popular U.S. television shows over a 20-year period, as well as showing that Latinos tend to be portrayed as hypersexual. In this study (and others: Harwood & Anderson, 2002), African Americans are portrayed in roughly the same numbers as their population presence, a function perhaps of the NAACP’s careful lobbying of media groups for better and more frequent representations (e.g., NAACP, 2016). These examinations extend across many genres and platforms of electronic media, with a particular focus on prime-time, children’s, and informational television; music video; newspaper and magazine coverage (and magazine advertising); and more recently, Internet platforms (e.g., Williams, Martins, Consalvo, & Ivory, 2009).

The Influence of Mediated Communication

As might be expected, considerable experimental and survey research has examined the effects of viewing the kinds of portrayals outlined in the previous section. Globally, people who consume negative portrayals of outgroups tend to subsequently display more negative attitudes about those groups. This occurs both in short-term experimental exposures, and with measures of more long-term exposure. Mastro, Lapinski, Kopacz, and Behm-Morawitz (2009), for example, demonstrate that experimental exposure to media portrayals of Black crime are associated with subsequent judgments about African American criminality more broadly. Tukachinsky et al. (2015) show that historical trends in portrayals on television are associated with broader trends in social attitudes toward groups in society. These effects are not restricted to increasing prejudice or stereotyping in the dominant group. For example, Johnson, Trawalter, and Dovidio (2000) show that exposure to violent rap music has negative consequences for judgments of Black (versus White) defendants. Notably, these effects were consistent for Black and White respondents—Black respondents were as strongly negatively influenced by the violent rap music as the White respondents.

In contrast to the previous paragraph, considerable work has examined the potential for media portrayals to improve attitudes concerning particular groups. Much of this work builds on contact theory, framing contact between viewers and television characters as a form of parasocial or vicarious contact. While the terminology is not always used consistently, it is helpful to distinguish parasocial contact (where viewers have contact with an outgroup media portrayal) from vicarious contact (where viewers are exposed to media portrayals of communication between ingroup and outgroup characters). Parasocial contact can be understood as a mediated form of traditional “direct” contact, where liking for an outgroup media character might generalize to perceptions of the entire outgroup (Schiappa, Gregg, & Hewes, 2005). Vicarious contact can be understood as a media extension of “friend of a friend” forms of extended contact discussed in social psychology (Wright et al., 1997). In vicarious contact, a viewer is not only exposed to the outgroup member, but also to portrayals of intergroup interaction. Hence, such portrayals potentially offer models for how the ingroup might effectively communicate with the outgroup, and indeed for how the outgroup deals with the ingroup. Social cognitive theory (Bandura, 2001) would suggest that such models might be effective in improving viewers’ efficacy for future “real world” intergroup interactions (Mazziotta, Mummendey, & Wright, 2011; Ortiz & Harwood, 1997). Considerable work demonstrates that exposure to positive outgroup exemplars and portrayals of positive intergroup relations are effective in improving intergroup attitudes (Joyce & Harwood, 2014). Obviously, these portrayals need to be selected or constructed carefully—they do not represent the typical or average media portrayal. However, this work presents avenues for interventions that might be effective in improving intergroup relations, and that are easily disseminated to large numbers of people in the population.

While most of the media contact work addresses improving attitudes among dominant group members, media portrayals also have implications for members of non-dominant groups. This is particularly the case in terms of messages featuring fellow ingroup members being presented in positive or powerful roles. For instance, Dixon et al. (2009) show that exposure to rap music videos is associated with increased collective self-esteem for African American viewers. Such portrayals may be more likely when members of non-dominant groups have ownership rights or producing control over media (Harwood & Roy, 2005).

One emerging strand of research examines an alternate route through which media can enhance intergroup relations. Certain media portrayals are “elevating”—they make us feel connected to higher moral purposes, expose us to exceptional achievements, or “move” us emotionally to a higher plane. Such elevating experiences reduce prejudice against other groups (Oliver et al., 2015). As a result, such media messages have the potential to improve intergroup relations, even perhaps absent any direct contact with the outgroup.

Selective Seeking of Mediated Content

One prerequisite for media to influence us is, of course, that we are exposed to those media. Considerable work examines the reasons why we seek certain messages and avoid others, and some of this (dating back at least to Blumler, 1985) examines group-based motivations for media selection. People gravitate to media featuring members of salient ingroups (children enjoy shows featuring child characters, African Americans consume more media featuring African Americans than other ethnic groups, etc.). This effect persists even when the content of messages is controlled (Harwood, 1997). A social identity gratifications approach suggests that such consumption delivers social identity rewards (Harwood, 1999). This idea is supported by work demonstrating, for instance, that certain groups selectively seek media messages featuring positive portrayals of the ingroup, as well as enjoying messages portraying the outgroup in a negative light (Knobloch‐Westerwick & Hastall, 2010).

Future Research

Work on the ability of elevating media experiences to improve intergroup tolerance should be extended. Internet platforms such as Facebook increasingly offer troves of “feel good” content—videos of people doing good things for others, for instance. Such content should be elevating, and apparently is in demand from viewers (Krämer et al., 2016). Researchers should examine more carefully the potential for such messages to influence attitudes when they also feature portrayals of outgroups or intergroup relationships (Bartsch, Oliver, Nitsch, & Scherr, 2016), as compared to when they provide elevating content that does not feature group portrayals.

Related, media researchers should also develop new paradigms for examining new media. The Internet presents challenges for content analytic and effects research that were not present in times when television was dominant. Only a couple of decades ago, television presented a finite set of messages, and most consumers largely watched a similar small fraction of those messages (prime-time network TV). The Internet presents an almost infinite variety of messages, and many of them are sufficiently personalized (e.g., a Facebook feed) that methods based on manipulating exposure to a particular widely disseminated message are no longer ecologically valid. Similar points apply to new media’s impact on interpersonal relations. We know relatively little about how an email exchange might compare with a face-to-face conversation in terms of influencing intergroup attitudes.

Research should also explore a greater diversity of media in more detail. For example, relatively little research in communication or psychology has considered exposure to music as a form of intergroup communication. However, there is considerable potential to understand both the lyrics in music (Binder & Ward, 2015) and the music itself (Harwood, 2017) as having the potential to influence intergroup relations. Musical performance requires more careful synchronization than most other human activities, and hence group musical activity serves important coalition-signaling functions for groups (Hagen & Hammerstein, 2009). Likewise, musical activity across groups provides one of the more powerful displays of intergroup cooperation, and hence might perhaps serve a critical modeling function for intergroup synchronization in other spheres (Harwood, Qadar, & Chen, 2016). When combined with the intense emotional power of music, these dynamics deserve more attention.

Similar points could be made concerning video games. Gaming provides opportunities for the interaction of “real world” intergroup contact (e.g., with other gamers on one’s team or an opposing team, some of which may be mediated via some sort of online chat) along with virtual contact (e.g., with ingroup and outgroup avatars) (Eastin, Appiah, & Cicchirllo, 2009). Gaming also presents opportunities for direct conflict with non-human enemies (aliens, for instance), which has the potential to enhance “human” identities and thus increase appreciation for diverse groups of humans (Ellithorpe, Ewoldsen, & Porreca, 2015). This level of complexity presents exciting opportunities for theory building.

More research is needed to understand when and why people seek out contact with outgroups, both direct interpersonal contact and mediated messages featuring the outgroup (Dunne, 2013). We understand a fair amount about the factors inhibiting such voluntary exposure (e.g., anxiety, institutionalized and voluntary segregation: Dixon, Durrheim, & Tredoux, 2005), but very little about why people intentionally put themselves in positions where intergroup contact is likely.

Communication researchers are also well positioned to examine more carefully the intersections and interactions between mediated and interpersonal experiences of intergroup communication. To what extent does media modeling of effective intergroup exchange increase people’s willingness to engage in intergroup communication, and does subsequent interpersonal communication have different effects when it is potentially “comparable” against the mediated message? Might “real” intergroup communication sometimes be disappointing or frustrating because it doesn’t live up to an idealized media image of intergroup harmony? We should also continue examining the relative power of mediated and interpersonal exposures: Scattered survey work exists documenting the relative predictive power of interpersonal and mediated contact, but more data are needed for some coherent conclusions on the relative power of each (Ramasubramanian, 2013).

A final note on theory is merited. Communication scholars, on the whole, have been a little slow in developing theoretical perspectives related to intergroup communication. Most communication work in this area builds from psychological perspectives such as social identity theory (Tajfel & Turner, 1986), social cognitive theory (Bandura, 2001), self-categorization theory (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), and the common ingroup identity model (Gaertner & Dovidio, 2000). Communication accommodation theory (Giles & Gasiorek, 2013) may be the theoretical perspective that has been most heavily developed within the field of communication, albeit even that was initially formulated in psychology. In the mass communication area, theories of agenda setting (McCombs & Shaw, 1972) and priming (Scheufele, 2000) have been dominant, with some attention to cultivation theory (Gerbner & Gross, 1976) approaches. The latter approach developed within communication, but the others are borrowed from elsewhere. There is no doubt that communication scholars have developed a unique and communication-focused body of work on intergroup processes, and that they have contributed to the development of these theories, but influencing conceptual thinking on these topics across disciplines is an area for future growth for communication scholars.