The Public Sphere Is “Too Darn Hot”: Social Identity Complexity as a Basis for Authentic Communication

Abstract

A growing body of research suggests that the contemporary media environment enables motivated reasoning, which intensifies affective polarization. This is especially the case in the U.S., where elections are capital-intensive and media are largely commercially owned. From a normative perspective, these commercial forces may interfere with authentic communication by hijacking the “lifeworld” and thus undermining the sincerity of our speech. From a psychological and empirical perspective, this means we are an affective public steeping in “hot cognitions” that unconsciously motivate us toward processing (mis)information in biased and distorted ways. This kind of cognitive limitation intensifies as current affairs heat up, but starts well before, as a function of media market boundaries aligning with human psychology. Through a synthetic literature review of theory and empirical research, this essay argues that “social identity complexity” may help to overcome some of the worst outcomes of motivated reasoning, pointing toward a developmental basis for more authentic communication in the public sphere.

1. Introduction

Evaluations of deliberative democracy, political campaigns, and media institutions in the U.S. have varied widely, often in relationship to hopeful and less hopeful political moments. In the early to mid-2000s, the interactive capacities of the Internet promised an expansive potential for the reinvigoration of the public sphere and with this, the rise of a new kind of politics—witness the campaign of Barack Obama in 2008, the first “Internet President”, who was able mobilize young people and start a social media and networked inspired movement with a message of “hope” and “change”. For many observers, the Internet seemed limitless, expansive, democratizing, and generally pro-social.

But, over the course of the last three election cycles or so, strategic and commercial forces online have multiplied, and media have become ever more sophisticated, abundant, and ubiquitous, all of which seems to be distorting authentic human communication or communicative action and fortifying psychological tendencies toward motivated reasoning and affective polarization. Through selective and algorithmic exposure to predominantly likeminded (e.g., Bennett and Iyengar 2008; Stroud 2011; Sunstein 2009), uncivil (e.g., Sobieraj and Berry 2011), mis/disinformed, or sensational media content and group communication, “rationalization” rather than rational thinking (e.g., Lodge and Taber 2013) appears as the predominant mode of cognitive processing, and affective polarization as the current psychological state (e.g., Abramowitz and Webster 2018; Iyengar et al. 2012).

The normative and empirical question of whether the Internet might revive prospects for authentic communication and a “public sphere”, undistorted by the influence of markets, now seems quaint. The perceived impracticality of communicative action has always caused some critics to reject it as an ideal (e.g., Mouffe 2000; Young 2000), but the empirical reality of the current moment is even more persuasive. Americans may be talking about politics, perhaps more than ever, but their minds are already made up. The Internet, it seems, has created an “inability to communicate across lines of difference” (Waisbord 2016, p. 2), and rather than connecting us, it “enhances disconnection and segmentation” (Pfetsch 2018, p. 60).

And yet, to some extent, nostalgia for pre-Internet U.S. democracy is misplaced. The very recent hope for the reinvigoration of a public sphere only really became possible with the emergence of the Internet, fraught as the social experiment has been. The problem is not the Internet or even social media, per se, but commercially/corporately dominated, niche market systems which undermine authentic human communication through their reinforcement of social identities. Niche media markets, oriented toward consumer boundary maintenance, fuel hot cognition, motivated reasoning, and affective polarization by reinforcing rigid and simple social identity structures through their use of narrowly and densely drawn consumer boundaries. In liberal democratic societies, the extent to which these boundary dynamics dominate psycho-social conditions relative to other, possibly more authentically “communicative” boundary dynamics seems in part proportional to the dominance of markets as an organizing principle for social systems.

This review essay begins with an explanation of communicative action (Habermas 1984) as an authentic form of communication, the public sphere, and the barriers to their achievement associated with human psychology, especially affectively motivated cognition. It asks: Are humans psychologically capable of deliberation? Or more specifically, does the theory of communicative action depend fatally on the presumption of a rational public, capable of non-motivated reasoning? Through a synthesis of normative theory and empirical research, I argue it does not, especially if: (1) social systems gain greater clarity about what communicative action/authentic communication entails and which parts we may rescue, considering affect driven cognition; and (2) if they help to foster social identity complexity such that people habitually hold a wider variety of social identity groups in warmer regard. I examine how media system boundaries, both “niche” and “communicative”, might align with the development of social identity complexity and its capacity to be used as a basis for more authentic communication among citizens.

As such, this literature review works toward some starting points for a developmental model of authentic communication—one that attempts to confront apparent psychological impediments to productive deliberation and to suggest practical mechanisms for overcoming them. Rather than remaining at the purely theoretical level, the analysis relies heavily on a review of empirical research to be as precise about the mechanisms by which social identity complexity may be enhanced or diminished by social systems, and in particular, media systems.

2. Communicative Action as Authentic Public Sphere Communication

In the Structural Transformation of the Public Sphere, Habermas (1989) proposes a time and space in when something like a set of public spheres emerges as enlightened counterbalances to monarchical rule, whose power comes not from brute force, but the force of reason. The “classic public sphere” comprised a set of publicly accessible media and forums for political discussion, including salons, coffee houses, newspapers, books, and pamphlets. Collectively, these institutions embodied enlightenment ideals of rationality, love, civility, diversity, and freedom, facilitating the unfettered deliberation of ideas and, thus, the development of the will of the people in the form of “public opinion”. Habermas theorizes regarding the structural underpinnings of the classic public sphere and the changes that transformed it and why one has yet to emerge in replacement. His central thesis is that when political–economic “systems” and media institutions become too deeply intertwined, the strategic logic of the former comes to distort communicative action.

In contrast with more instrumental forms or rationality, communicative action uses a form of rationality rooted in human communication rather than the strategic action of systems. Importantly, communicative action is directed toward the goal of true understanding and consensus building, rather than instrumental persuasion. As a normative ideal, people assume a critical stance toward their own speech so that their claims may be more easily identified and criticized. Mutual understanding, rather than instrumental persuasion, is achieved through authentic communication that justifies the truth, the right to speak, and the sincerity of the utterances. Taking a phenomenological stance, Habermas sees this form of communication as an authentically human activity, even as it is challenging, and one which leads to more stable forms of consensus and rational public opinion (Habermas 1984).

Yet, even scholars in general agreement with these normative ideals see little reason for optimism that anything like it could exist as an empirical reality. While research suggests that deliberative processes promote several pro-democratic benefits, including improved “argument repertoire” (Cappella et al. 2002) and especially increased tolerance (Mutz 2006), appraisals of group discussions in the actual world, especially the Internet, have been less than sanguine (e.g., Gregson 1998; Hagemann 2002; Santana 2015; Sydnor 2019). Normative ideals need not be reflected in reality—this is why they are ideals. However, if we conclude that human beings are psychologically incapable of any approximation of this ideal, or even democracy itself, then all hope for the ideal seems lost.

3. Psychological Barriers

And in recent years, perhaps the worst news for advocates of communicative action comes from a large body of empirical research suggesting that human beings are not rational (i.e., not persuaded by good reason), may be psychologically incapable of objective reasoning about politics, and instead process political information in entirely affect-driven, biased ways (e.g., Erisen et al. 2014; Lodge and Taber 2013). From a theoretical perspective, this research suggests an “affect” model for human cognition, which means that unconscious emotions have primacy over and precede all conscious thinking processes.

Further, political information processing is especially fraught according to the “hot cognition hypothesis”, which posits that socio-political concepts are highly affectively charged and that exposure to political concepts activates unconscious affect within “milliseconds”, much faster than any conscious appraisal process (Lodge and Taber 2013). Neuroscientific research supports the automaticity and speed of this response—even arbitrary group distinctions can lead people to process ingroup faces differently than outgroup faces in as little as 170 milliseconds (Ratner and Amodio 2013). Information processing is thus biased by an initial affect valence, which automatically directs/motivates reasoning processes toward conclusions that result in “affective congruence” (Erisen et al. 2014). In what seems like a fatal blow to deliberative democracy, Lodge and Taber (2013) conclude that “political behavior is the result of innumerable unnoticed forces, with conscious deliberation nothing more than a rationalization of the outputs of automatic feelings and inclinations” (p. i).

Importantly, political sophistication does not mitigate affectively motivated reasoning, but only strengthens the reasoner’s capacity to employ reason in the service of affect. Taber and Lodge (2006), for example, found that politically sophisticated individuals were more, not less likely than to evaluate arguments based on the extent to which they conformed to their own political beliefs, likely because they have more information at their disposal, making political concepts even more affectively charged and efficient for them to refute arguments that do not conform to their affective state. Similarly, Kahan (2013) found that effortful thinking actually strengthened motivated reasoning. Political sophisticates, it seems, find attitude reinforcement especially satisfying, exhibiting more activation of the amygdala (associated with emotional salience) and ventral striatum (associated with reward processing) (Gozzi et al. 2010).

A cause and result of hotly charged cognitions and motivated reasoning are feelings of affective polarization, which are on the rise in the U.S. (e.g., Abramowitz and Webster 2018; Iyengar et al. 2012). Affective polarization pulls partisans to have generally favorable impressions about their ingroup and universally negative views about the outgroup. In the U.S., research by Iyengar and colleagues suggests that an increase in negativity in campaigns is at least partially responsible for increases in partisan affective polarization (Iyengar et al. 2012). Using survey data, they found that inter-party animosity has increased steadily and dramatically since the 1960s, with Republicans disliking Democrats somewhat more than the other way around. Affective polarization appears to intensify over the course of U.S. election campaigns and is felt most acutely in battle ground states with high exposure to negative advertising. Importantly, this trend was unmatched in the U.K., where affective polarization is not as high and has not increased by nearly as much.

A citizenry thus divided poses substantial problems for communicative action. Affectively polarized people engage in motivated reasoning as a form of “identity defense”. They will disbelieve and resist empirical assertions (e.g., Liu and Ditto 2013; Taber and Lodge 2006) and even facts (Iyengar and Westwood 2015) that threaten the standing of the social identity and will be more inclined to believe social-identity-affirming misinformation (Garrett et al. 2019). Further, affective polarization leads people to be less likely to seek diverse perspectives (Valentino et al. 2008) in order to make compromises with one another (MacKuen et al. 2010) and to hold themselves to a standard of civility and tolerance (e.g., Layman et al. 2006).

4. Memory/Developmental Models vs. Affect-Based Models

It would seem, then, that the central role of affect in political reasoning processes poses a serious challenge to normative theories of democracy, which hold the “informed” citizen as the “good” citizen and which propose education as an antidote to poor reasoning processes on the part of the public (e.g., Dewey 1916). Developmental models suggest that over time, citizens can improve through the development of complex mental structures and through increased knowledge. While “memory-based” models have long taken into account bias in information processing (e.g., Zaller 1992), they assume that political sophistication aids in minimizing bias (which would seem to leave room for hope of improvement). They do not acknowledge the primacy of affect and the extent to which politically sophisticated citizens tend to have especially strong affect associated with political concepts due to the extra consideration they have given them.

“Affect-based” models of political psychology would seem, then, to upend developmental models, and with that any hope of communicative action as an authentic mode of engagement. And yet, the developmental perspective is not without support—research finds that reasoning capacity improves with age (Piaget [1932] 1965), a liberal education (Pascarella and Terenzini 1991), and deliberation (Bago et al. 2020). And while politically knowledgeable people are less likely to be persuaded or change their minds overall, they are more likely to change their minds than the less informed when presented with good and novel reasons (Zaller 1992).

Perhaps there is a way in which developmental and affect-based theories can be reconciled. These approaches, with the former seemingly more optimistic about communication than the latter, may not be hopelessly incompatible. Further, affect-based reasoning may not deal quite as fatal of a blow to communicative action as we might first imagine.

First, “affect”, in and of itself, may not be anathema to “reason”. As suggested by Dahlgren (2018), “At bottom, political passions always have reasons, even if they are not always immediately accessible to us, there is some goal or object that is valued. Thus, political passion, even if it is partly anchored in the unconscious, is not blind; it involves some sense of the good…. Reasons, in turn, incorporate emotions…” (p. 2058). The characteristics of affect, positive or negative, should furthermore not be thought of in purely individual or social psychological terms. They should vary considerably based on shared social experience or “civic cultures”, as well as available resources (Dahlgren 2018).

Second, affect-based theorists seem most concerned about a particular type of reasoning capacity, which is to say the ability to be persuaded by good reason. Hot cognition makes this nearly impossible. But in these cases, deliberative processes are largely defined as individual cognitive processes by affect-based theorists (internal deliberations), which very much leaves open the question of how authentic human communication processes, which is to say the kind of deliberation imagined by deliberative democracy itself, might impact unconscious affect over time.

Moreover, if we are to hold communicative action as a normative ideal, persuasion is not essential. In fact, from a Habermasian perspective, the type of reasoning processes that arise out of persuasive attempts may be inauthentic and more associated with a kind of side effect of the strategic logic associated with the system. Moving beyond the normative, research suggests an empirical foundation for Habermas’s contention that there are, in fact, two types of rationality, one communicative (authentic) and the other strategic—indeed, Schaefer et al. (2013) found they are associated with different parts of the brain. When research participants engaged in communicative rationality, MRI imagining showed that the parts of brain that lit up were those associated with moral sensitivity, emotional processing, and language control. However, when participants were engaged in strategic rationality, these areas of the brain did not light up, leading Schaefer et al. (2013) to conclude, “strategic reasoning is associated with reduced social… cognitions” (p. i). Thus, it is possible that if the goal is mutual understanding, as Habermas proposes, rather than instrumental persuasion, human beings may be more psychologically equipped to achieve it should they manage to light up the more communicative areas of the brain.

Nevertheless, whether we are talking about internal deliberation or deliberation with one another, an important question is whether there is a kind of education and moral development that humans may undergo to tame some of the more disturbing consequences of affect. If political knowledge is sometimes helpful (helps us be persuaded by good reason), but in some cases harmful (enables successful motivated reasoning), we can conclude that political knowledge is useful, but not sufficient. If it is true that affect precedes reason, when there is reason to be had at all, then the characteristics of affect matter greatly.

Narvaez (2010), attempting to reconcile “intuitive” or “affective” models of moral reasoning with developmental approaches, specifies the development needed to for increased moral complexity, which would prompt automated yet mature moral responses through habitual practices. “Habituated empathy”, the author suggests, can be developed through practice. Even in intergroup contexts, levels of habitual or automated empathy (measured as neurological responses) can vary in critical ways—while people generally exhibit greater empathy toward ingroups, the size of this gap in empathy is predictive of people’s propensity to engage in “costly helping” of the outgroup (Hein et al. 2010). Interestingly, Taber and Lodge (2016), who are among the most prominent scholars working in the “affect-based” tradition, wonder, albeit skeptically, if “Egalitarian values can be triggered outside of awareness to control stereotyped driven inclinations and discriminatory behavior. Perhaps cooperative responses can be learned (conditioned?) to automatically override zero-sum or retaliatory behavior even in the presence of competitive cues” (p. 63).

Perhaps an underutilized portion of Habermas’s (1989) writing is his discussion of the psychology and education of the bourgeoisie, which in theory prepared them for the rigors of public discussion. This education was not just geared toward becoming more informed. Rather, he discusses the importance of a “cultivated personality” developed in the private/familial realm: “The three elements of voluntariness, community of love, and cultivation were conjoined in a concept of humanity…” and they became persons capable of entering “purely human relationships with one another”. The sort of quiet subjectivity building, diary writing, and novel reading that came alongside this distinctly bourgeois form of development perhaps seems too rarified for contemporary society, and yet it does suggest that some a kind of psychological and moral education could be needed for authentic communication. The idea of a liberal education is not too far away from this general idea, although it may place not enough emphasis on education that enhances “subjectivity” or, as is more pertinent to the concerns of the current analysis, “affect”.

If we attempt to reconcile affect-based models with normative ideals, a realistic developmental perspective recognizes emotional constraints on human reasoning and acknowledges consistent findings that group deliberation, under the right circumstances, can produce better reasoning, but perhaps even more importantly, tolerance and mutual understanding. Even if we are incapable of “cool cognition” as we process information on politics, the outcome of mutual understanding may be sufficient, and may be one that people are more psychologically equipped to achieve. Human interaction involves not just the processing of information, but intergroup dynamics, which can very likely influence the characteristics of affect and how deliberation is thus constituted.

A deliberative democracy, capable of authentic communication, might then be composed of citizens with habitual positive affect toward one another. Even if minds remain unchanged when faced with good reason, mutual understanding may still occur, which may make democratic compromise, or even rational consensus, more plausible. In the arena of developmental psychology, this would seem to require social identity complexity. The development of the “socially complex” citizen may be at least as important to the possibility of authentic communication as an “informed” or politically sophisticated one. While not referring to the psychological concept directly, Dahlgren (2005) nicely articulates its potential normative significance:

Citizenship is a formal status, with rights and obligations. However, it also has a subjective side: People must be able to see themselves as members and potential participants with efficacy in social and political entities; this must be a part of people’s multidimensional identities (p. 159).

5. Social Identity Complexity

As a psychological concept, social identity complexity is defined by (1) people’s ability to see out-groups as being composed of heterogeneous or differentiated individuals (Ben-Ari et al. 1992) and (2) the extent to which their cognitions reflect a sense of belonging to numerous and differentiated social groups (Roccas and Brewer 2002)—”people with low social identity complexity see their ingroups as highly overlapping and convergent, whereas people with high complexity see their different ingroups as distinct and cross-cutting membership groups” (Brewer and Pierce 2005, p. 428). In this way, simple social identities might have as few as one ingroup to which all other groups are subservient, such as “Republican” or “Democrat”.

Social identity complexity allows for resistance to self-categorization processes and the capacity to view social behavior in a multidimensional way (Bieri 1966), which allows for more flexibility when evaluating group members (Markus and Zajonc 1985). Research further demonstrates that social identity complexity is associated with openness, low power orientation, universalism, positive attitudes toward outgroups, less ingroup favoritism (Brewer and Pierce 2005; Roccas and Brewer 2002), tolerance (Brewer and Pierce 2005; Miller et al. 2009), and the capacity to experience shame when ingroup members act violently (Costabile and Austin 2018). Conversely, as demonstrated by Mason (2018), social identity simplicity, measured as the degree to which partisans “sort” various social identities (racial, religious, ideological) into one simple mega-identity, results in cooler evaluations of partisan outgroups/affective polarization.

In their conceptualization, Roccas and Brewer (2002) suggest that social identity complexity can become automated or “chronic”, requiring very few cognitive resources, which is promising if unconscious affect indeed precedes all effortful thinking, as Lodge and Taber (2013) suggest. Chronic social identity complexity would seem to greatly assist deliberative, intrapersonal, and especially interpersonal processes, such that it would be a complexity “bias” and its associated “warm” (Mason 2018) affective states that would be the starting point for deliberation.

6. Media Systems and Social Identity Complexity: A Developmental Perspective

In summary, as media stimuli have grown hotter and more pervasive, affective polarization and motivated cognition have intensified. Indeed, in some ways, affective polarization and motivated reasoning may be a reasonable psychological adaptation to structural-level challenges and social-identity-related crises. As suggested by Kahan (2013), “When societal risks become infused with antagonistic social meanings, it is…individually rational for ordinary members of the public to attend to information in a manner that reliably connects them to positions that predominate in their identity-defining groups” (p. 420). More hopefully, however, the differences in affective polarization in the U.S. over time and the unparalleled effects found in the U.K. (Iyengar et al. 2012) do suggest the possibility that human psychology is, to some extent, malleable and responsive to pro-social cultural/societal-level influences.

A developmental approach to social identity asks us then to look to structural features of society that also vary and might suggest educational/developmental openings and healthier ways of coping with social identity crises. Understanding the antecedents of social identity complexity is therefore of critical importance. For Roccas and Brewer (2002), the level of chronic social identity complexity in a society is contingent on: (1) individual differences in cognitive style, (2) socialized values (especially universalism versus authoritarianism), and (3) experiences with a complex social environment. While individual differences in cognitive style are likely to be partly rooted in trait or personality differences (which may be stable across populations), socialized values and experiences vary across cultures, suggesting that chronic social identity complexity can also vary and may be increased through societal-level developmental processes.

Social environments and institutions do not cause people’s needs for social identity, certainty, or existential security, but they likely influence the shape and depth of their pathologies. As suggested by Roccas and Brewer (2002), “Social environments in which different bases for ingroup-outgroup distinctions are crosscutting rather than convergent confront the individual with knowledge about the differences in meaning and composition of different social categorizations” (p. 96). Similarly, Tajfel (1981), the most prominent social identity theorist, deeply questioned the extent to which “explanations of social conflicts…can be mainly or primarily psychological” (p. 7). Rather, he believed that social identity theory should provide a starting point for thinking about societal-level structures and social processes that exacerbate and fuel dysfunctional attachments to social identities. Speaking to the effects of media systems on identity development processes, Bennett (2012) further suggests “Mediating the direct impact of social change on personal identity are important intermediate institutions…and symbolic resources…that offer people creative options to think about their places in society and define their relations to others” (pp. 313–14).

Although the media are clearly not the only institutions that might respond to the current crisis of identity, it is the route by which most people come to understand the political and even social world, and could thus be central to the development of social identity complexity. In support of this possibility, comparative research suggests that different media systems are associated with different outcomes at the individual level, such as political knowledge (e.g., Curran et al. 2009). As we have seen with Iyengar et al. (2012), media systems with more robust public broadcasting options (e.g., the BBC) and different campaign finance structures have less affective polarization. This may in part be a function of public broadcasting’s ability to provide a greater supply of prime-time news (Esser et al. 2012), thus producing a greater “opportunity structure” for people to encounter cross-cutting political perspectives (Castro-Herrero et al. 2018).

7. Media System Boundaries

In an effort to clarify the conditions under which social identity complexity may be developed or diminished, here, I focus on the role of media system boundaries in shaping the logics of developmental processes—those stemming from current “niche-market” logics and from a hypothetical “authentically communicative logic”. Niche media markets have created narrow and dense consumer boundaries. They are narrow in the sense that they are personalized and highly targeted—we are each swimming in our own lane, so to speak, or at least in a lane with other people we identify as “like us”. They are dense in the sense that it becomes harder to move across the boundaries or to “swim in other lanes”—they can seem hostile, hard to find, and not for “us”.

To characterize contemporary boundaries as dense and narrow may seem odd because, in some ways, the Internet makes social boundaries more porous. In fact, the notion of porous boundaries is evoked repeatedly in early Internet-related scholarship (e.g., Bimber et al. 2005; Brundidge 2010; Cammaerts and van Audenhove 2005). Cammaerts and van Audenhove (2005), for example, suggest that an “unbounded” form of citizenship as opposed to a “bounded” one distinguishes contemporary public life. By increasing the porousness of social boundaries, the Internet increased the “traversability” between “news” and “discussion” among citizens (Brundidge 2010), allowed new voices to be heard, made the formation of heterogeneous networks and “weak ties” possible, diversified information sources, and eased the costs associated with collective action (Bimber et al. 2005). While we can still see the effects of certain kinds of porousness at work, not least of which is the rising visibility and success of social movements (e.g., Papacharissi 2015), economic forces in conjunction with technology have, in effect, provided a powerful counterbalance to the “porous social boundary inducing” aspects of the Internet.

The problem of boundaries moves beyond the familiar issue of “echo-chambers”. Since the problem of interest here is affective polarization, disagreement over policy is not the main (or at least not the only) source of inter-partisan animosity. Mere exposure to alternative ideological arguments is not enough to mitigate outgroup hostilities, and may even reinforce them under certain conditions due to motivated cognition. Current conditions, such as those created on social media, appear to be highly conducive to such reinforcement—they suggest to people that they find and connect with one another within the boundaries of social identity, as opposed to, say, the more traditional and somewhat less politicized boundaries of the workplace or geography. In this space, partisan social identities become bound up with other social identities (Mason 2018), such as racial (Abramowitz and Webster 2018), gender, national, or religious identity, making it increasingly difficult to see the ways in which they are distinct. Törnberg et al. (2021) characterize current boundary conditions as creating “solid” (as opposed to “fluid”) information regimes. They find that “Politics in the solid regime represents a condition where partisan identity strengthens and begins to align with other social identities… new information technology triggered such a transition by facilitating the social psychological processes that form and strengthen social identity” (p. 12).

Functionally, narrow and dense consumer boundaries created by this “solid regime” may limit social identity complexity and promote affective polarization in at least two major ways: First, they promote media frames and the spreading of (mis)information that reinforces existing partisan social identities. Second, as niche market boundaries colonize the media landscape, increasingly drawn by a few powerful corporations, they may reduce the ease or “serendipity” (Pariser 2011) with which people have contact with a meaningful (e.g., warm, thoughtful, empathetic) diversity of perspectives and social identities as they traverse the public sphere.

In niche markets, framing largely serves to reinforce existing interpretations and affect. Jamieson and Cappella (2008), for example, found that conservative news outlets use frames to reinforce stereotypes of partisan outgroups. They suggest that this pattern of framing gives rise to incivility as an intrinsic (and profitable) characteristic of the news media market. Jones similarly points out (Jones 2012) “Fox News, in conjunction with other conservative media outlets, provides a steady and consistent diet of such overtly ideological symbolic material to sufficiently sustain viewer interest and commitments as a community” (p. 181). Nadler (2022) further argues that they appeal to their audience by “Invoking a specter of smug liberals and disdainful leftists” and by positioning “themselves as the defenders of their audience’s besieged identities” (p. 2637). By arousing a sense of “collective injury” (Polletta and Callahan 2017), conservative media is then able to offer its outlets as sources of therapeutic social identity repair (Nadler 2022).

Research further demonstrates how this approach to framing does indeed activate partisan identity, which exacerbates ingroup/outgroup thinking (Levendusky 2013), and, in terms of affect, may prompt high negative arousal and increased identity-based motivated reasoning (Boyer 2023). Some partisan outlets may be so clearly branded that frames and arguments are not even needed to achieve an effect; the response essentially becomes Pavlovian, activating partisan identities upon use (e.g., Knobloch-Westerwick and Kleinman 2012).

One of the most important shifts in online boundaries has come along with the increased use of algorithms. Search engine algorithms, social media sites (and their algorithms), and marketing consultants have further created more segmented and highly targetable social spaces. Campaigns use the Internet to target voters with unprecedented scale and specificity (Bennett and Manheim 2006)—this was made especially visible in 2016 and 2020, with the proliferation of targeted mis/disinformation. By following the moment-to-moment behavior of citizens and by monitoring for slight shifts in their group loyalties and affective states, algorithms have become increasingly sensitive and responsive to social identity (see Soffer 2021).

8. Communicative Logic: Building Complexity through Porous Social Boundaries

In summary, niche markets impose substantial limitations on the extent to which media institutions may use enhance social identity complexity, reduce affective polarization, and assist in authentic communication. However, communicative logic suggests a move toward more permeable or porous citizen boundaries—enhancing citizens’ capacity and motivation to move from entertainment to information, the private to the public spheres, likeminded to non-likeminded perspectives, and one social identity to another. In this way, the Internet could become more akin to the kind of diverse geographic neighborhoods that appear to nurture social identity complexity (Schmid et al. 2013). Under these conditions, media institutions and regulatory bodies might use frames and digital media boundaries in ways that are designed to maximize social identity complexity and, thus, mutual understanding through authentic communication.

Unfortunately, experimental interventions designed to quickly manipulate increased levels of social identity complexity or reduced affective polarization are often unsuccessful and sometimes even backfire (e.g., Levendusky 2018), leading Lyons (2018), for example, to conclude: “the influence of trait-based identity does not appear to be as sensitive to manipulation as theorized. While social identity interventions may be more viable in certain contexts, strategic depolarization will require a more consistent approach” (p. 800). Indeed, while there appears to be abundance of literature demonstrating the ways in which humans can be made simpler, less research has examined the mechanisms by which we become more complex. Certainly, more research needs to be conducted, and yet, it still seems useful to think about concrete mechanisms where we can find them. In media systems where such mechanisms are pervasive as a norm, social identity complexity may be developed over time, perhaps even becoming chronic in a critical mass of citizens. These mechanisms, in any case, offer practical starting points for thinking about how micro-level social processes could be considered in conjunction with structural-level media designs.

8.1. Media Framing

As social-identity-inducing mechanisms, media frames seem to hold some potential. As suggested by Shah et al. (2004), certain frame combinations can stimulate “individuals to consciously feel more motivated to think about and respond to open-ended question[s]” and trigger “increased associations with existing mental structures, thus automatically activating a greater number of linked constructs” (p. 115). Frames could also further activate people’s identification with multiple groups, motivating a search for a more inclusive superordinate social identity and, thus, greater social identity complexity (Brewer 2009). Indeed, the robust use of competing political frames and arguments can be helpful for making multiple identifications meaningful and functional for citizens, which could reduce simplistic outgroup stereotypes (Brewer 2009), allow for less salient self-categorization processes (Hogg and Reid 2006), make reasoning less affectively motivated (Redlawsk et al. 2010), and reduce affective polarization (Levendusky 2017). As suggested by Grant and Hogg (2012), when a salient social identity is in competition with others, its prominence will be lessened and have less weight on subsequent cognition.

8.2. Online Boundaries

While psychological tendencies toward selective exposure to consonant information and likeminded social groups undeniably hamper the potential of technologies to connect people across lines of difference, more permeable online boundaries could serve to counter selectivity by exposing people to diversity, if only inadvertently (Brundidge 2010). Helpfully, people’s unwillingness to engage with diverse perspectives may not be as much of a barrier as once thought—research demonstrates that while people do indeed seek out consonant perspectives, they are comparatively less interested in avoiding dissonant messages (e.g., Garrett 2009; Zaller 1992).

Unfortunately, mere exposure to outgroup members is not usually enough and may sometimes lead to further affective polarization due to partisan-motivated reasoning. We must then consider the boundary conditions by which positive contact with out-groups could be facilitated. As posited by contact theory, group prejudice may be reduced when people connect in the spirit of friendship, without anxiety, and with an empathetic mindset (Dovidio et al. 2011).

Fortunately, recent research has found that some forms of intergroup contact can go so far as to diminish existing affective polarization. For example, when people are enabled to see themselves as like outgroup members, they may come to wonder whether they are outgroup members at all (Wojcieszak and Warner 2020). Similarly, exposure to information that corrects misperceptions about outgroup members (even in the absence of contact) can furthermore reduce affective polarization by helping people to see that they are more like themselves than they originally believed (Druckman et al. 2022). In summary, positive contact with outgroup members (or positive information) can increase people’s social identity complexity by stimulating a greater awareness of shared group memberships and more differentiated understandings of their multiple ingroups (Schmid et al. 2013), thereby reducing affective polarization. While it may be difficult for media systems to facilitate positive contact outside of experimental contexts, the idea that information and, thus, the media can play a role in correcting misperceptions of outgroups, if not necessarily facilitating positive contact, is realistic and promising from a developmental perspective.

While the most central concern in this essay is affective polarization and its reduction through communicative action/authentic communication, not necessarily persuasion, we may not have to give up fully on the idea that people can be persuaded by good reason, affectively motivated as they may be. Redlawsk et al. (2010), for example, found that, in the face of repeated exposure to dissonant information, motivated reasoning does not go on indefinitely. More specifically, when participants were exposed to a minimal amount of negative information about their preferred candidate, they engaged in motivated reasoning which made them entrench in their original position—initial positive evaluations were even more positive. However, in the face of a critical mass of negative information about their preferred candidate, participants felt conscious anxiety overriding unconscious affect, which persuaded them to reevaluate their initial assessments. “Fact-checking” of misinformation can also prove effective, providing that the number of correcting statements outweighs the number of false statements (Aird et al. 2018). Contrary to the hypothesis that deliberation intensifies partisan-motivated reasoning, slow, deliberative thinking following exposure to false news headlines appears to reduce people’s belief in them (Bago et al. 2020). When affectively depolarized and better informed through critical masses of dissonant and corrective information, and through news that prompts deliberative thought, citizens may become more prepared to authentically communicate with one another.

9. Conclusions

While there is a growing body of research that suggests human beings are capable of depolarization, working their way out of motivated reasoning, and finding their way to more authentic communication, they clearly need to more support from media systems than is currently present. For now, it seems relatively clear that U.S. media markets are producing media frames and online boundary conditions that encourage social identities to be less complex. Short of a restructuring of economic imperatives, a more communicative media logic seems unlikely to emerge, and spaces for authentic communication may be increasingly difficult to find.

As suggested by Habermas (2006), “[m]ediated political communication in the public sphere can facilitate deliberative legitimation processes in complex societies only if a self-regulating media system gains independence from its social environments, and if anonymous audiences grant feedback between an informed elite discourse and a responsive civil society” (pp. 411–12). If complete media independence seems implausible, current levels of affective polarization in the U.S., so visibly and violently displayed in the 2021 Capitol insurrection, should be enough cause for concern to affirm a commitment toward developing more publicly owned media channels and networks that are committed to the public good and might at least compete with, if not entirely override, commercial mediated space.

Much may be learned from the initial openings that the Internet seemed to provide. At its advent, it seemed unbounded, limitless, expansive, and democratizing, if perhaps chaotic. Online technologies should move away from current algorithmic trends toward the reinforcement of consumer identities and work to make the boundaries between diverse networks of people and the ecosystems of accurate news and information more permeable, as well as to widen the boundaries around public spheres and make them more inclusive. News institutions should consider communicating in ways that bridge differences and enhance complexity, perhaps using framing, stereotype-correcting information, or critical masses of dissonant and accurate information as empirically plausible mechanisms (e.g., Aird et al. 2018; Druckman et al. 2022; Redlawsk et al. 2010; Wojcieszak and Warner 2020). In this way, journalism can perhaps assist in revving up the more pro-social and communicative areas of the brain (see Schaefer et al. 2013) and calm down the amygdala (e.g., Gozzi et al. 2010).

This review has attempted to synthesize theory and empirical research, with an eye toward reconciling disheartening research findings and realities with normative ideals. Communicative action is not fatally dependent on the absence of affectively motivated reasoning. However, the characteristics of affect, whether empathetic and warm, or polarized and “hot”, matter a great deal to our capacity for deliberation and mutual understanding. A goal for media systems, however they are structured, might be to assist in the development of “civic cultures” (Dahlgren 2005) that support citizens’ capacity to develop complex social identities that habitually foster warmer affect, allowing them to see others outside of narrowly defined social “ingroups” and “outgroups”—a developmental basis for authentic communication in the public sphere.