Emotions, social coordination, and the danger of affective polarisation

ABSTRACT

Smooth social interaction requires interindividual coordination. This Theory Section addresses the nature of the processes involved and the potential dangers of malfunctioning coordination. In her invited article, Butler provides a general overview of the processes involved, including interpersonal synchronisation, and advocates a dynamic systems framework for further research. In their commentary, Carré and Cornejo concur in principle but highlight the importance of the meaning attributed to the spontaneous expressive movements in naturally occurring interactions and the nature of the respective social situations. Suhay, in her commentary from a political science perspective, highlights the need for synergy based on social coordination for successful democratic governing. In commenting on the problem of political polarisation because of malfunctioning social coordination, introduced by Butler, Suhay adds the important role played by social identity and emotion. Of particular importance are affective reactions based on “evaluative biases” in favour of information that bolsters one's views and rejection of information that challenges them. I conclude this editorial by referring to recent data on affective polarisation, pleading for more multidisciplinary research on the phenomenon, and arguing for a more substantial contribution from the cognitive and emotion sciences. In particular, I outline the “predictive coding” framework and the potential contributions from emotion science.

KEYWORDS:

• Interpersonal coordination
• interactional synchronisation
• dynamic systems
• social identity
• political polarisation
• predictive coding

The topic of this Theory section was inspired by the choice of Prof. Emily Butler, University of Arizona, to devote her invited contribution to the complex interplay between cognition and emotion with social interaction and social coordination, a domain that has immensely profited from her contributions over the past years. Sadly, Emily Butler suddenly passed away in a very untimely fashion after having submitted the final version of her invited article in December 2022, as a consequence of a serious illness. We are very grateful to Prof.'s Iris Mauss, David Sbarra, and James J. Gross for having accepted to write an obituary in which they highlight Emily Butler's important contributions to the field (this issue). The Editors cherish the fact that Emily Butler has strongly highlighted the important role of affective polarisation in cognition and emotion research at the current critical period in modern society, as well as providing some leads for potential applications. In consequence, we will devote another theory section to this important and timely topic soon.

In recent years, there have been growing concerns about the unravelling of the social fabric in modern societies and increasing polarisation between social groups with different values and beliefs. This issue is frequently discussed in public forums and is increasingly studied in the social and political sciences. While it seems obvious that emotions play a major role in this process, so far emotion scientists have rarely addressed these worrying trends. This Theory Section attempts to broach these phenomena and their underlying causal mechanisms via contributions from different disciplines and research traditions.

In her invited target article, Emily Butler (this issue) focuses on the essential role of interpersonal coordination in social systems. She proposes to use dynamic systems theory to model the processes of changing interpersonal relationships and interactions, focusing on how intrapersonal processes (e.g., emotion) affect interpersonal processes (e.g., conflict or cohesion). In reviewing the literature, Butler describes different forms of coordination such as synchrony, interactive alignment, complexity matching, and synergy as well as the components involved in the coordination process (e.g., brain activity, autonomic physiology, and behavioural movements). She demonstrates that coordination can have both positive and negative effects, which requires consideration of moderating factors such as energy expenditure or cultural interaction rules. Butler then applies the dynamic systems framework to the emotionally charged issue of social polarisation, such as the growing conflict between adherents of the Democratic and Republican parties in the United States, generating hate and violence. She wraps up by enumerating the requirements for appropriate research efforts in this area such as the need for 1) conducting longitudinal studies that allow to examine nonlinear processes and control for group size and homogeneity, 2) focusing on real or virtual social interaction, and 3) measuring all relevant components of the system, all of which will require substantial amounts of cross-disciplinary collaboration.

The invited commentary by David Carré and Carlos Carnejo (this issue) takes up Butler's suggestion to use dynamic systems theory to investigate the processes of interpersonal coordination. While agreeing to the general utility of a dynamic approach that focuses on interrelated changes in social processes over time, Carré and Carnejo point to potential problems in over-generalizing dynamic systems theory, a fundamentally abstract, mathematical structure applicable to a large variety of human and non-human systems, to specific research questions on interactional synchronisation of human beings in social contexts. Carré and Carnejo propose that human interaction processes cannot be understood without consideration of the meaning attributed by the interacting persons to the actions and movements involved. They cite a study (Cuadros et al., 2020) on the differences in behavioural synchronisation between spontaneously occurring adult-child interactions, such as story telling, to an experimentally produced condition of mechanical, choreographed movements of the adult interacting with a child. The authors argue that an indiscriminate application of the dynamic systems framework to different kinds of interactions may result in neglect of the specific aspects of spontaneous expressive movements in naturally occurring interactions. Carré and Carnejo suggest that specific “affective atmospheres” of the respective situations determine differences in magnitude, time, and form of the coordination on the motor level.

The complexity of interactional synchrony phenomena has been addressed in a previous theory section in this journal, focusing on emotional mimicry (Cognition & Emotion, 2022, vol. 36, issue 5). Research in this area has become increasingly frequent (and sophisticated) over time, often with an emphasis on brain-to-brain synchrony. A recent special issue in Social Cognitive and Affective Neuroscience (SCAN; Schirmer et al., 2021) provides an overview of the issues addressed and the methodological advances achieved in this domain. Thus, Hoehl et al. (2021) review three recurrent themes in this literature: the eliciting stimuli, the underlying mechanisms and the benefits of interactional synchrony for individuals and groups. The contributions in the special issue of SCAN demonstrate how researchers explore various forms of interpersonal coordination to important constructs and outcomes in areas such as teamwork, social relationships, conversational settings, and clinical psychology.

In her invited commentary, Elizabeth Suhay (this issue) suggests that Butler's network approach can inform the study of politics by highlighting 1) the need for synergy based on social coordination for successful democratic governing and 2) the ubiquity of social influence patterns in society via norms and rules in shared, homogeneous social environments. As to the central issue of polarisation raised by Butler, she emphasises two factors of major importance, social identity and emotion. In earlier work, she has shown that social influence within groups is dependent on emotional processes, especially the “self-conscious” emotions of pride, embarrassment, and shame (Suhay, 2015). She asserts that emotions contribute directly to polarisation between social groups. In a recent article, Suhay and Erisen (2018) reported two online studies, one of which included a quasi-representative sample of Americans, finding, as predicted, that individuals felt negative emotions toward arguments that undermined their attitudes and positive emotions toward arguments that confirmed them. The authors proposed that anger plays a major role in motivating individuals to engage in “biased assimilation” of political information (Lord et al., 1979; Lord & Taylor, 2009)—an evaluative bias in favour of information that bolsters one's views and rejection of information that undercuts them. Suhay and Erisen suggest that the valence of the political attitude and associated considerations increase the likelihood of anger in response to challenging information, which then increases the likelihood of bias.

Suhay introduces three further perspectives from political science on the polarisation issue, the nature of the political institutions (e.g., the two-party system in the US), the role of historical contingencies (e.g., the battle over racial equality and the rise of feminism and gay rights movements), and the effects of economic interests and political power. With respect to using scientific insight to strengthen democracy, as suggested by Butler, Suhay suggests to proceed cautiously, in collaboration with people outside academia, focusing on lessening partisan polarisation by moving toward an inclusive and synergistic system of human interaction.

In introducing the polarisation issue in her invited contribution, Butler cites McCoy et al. (2018), who define social polarisation as “ … a process whereby the normal multiplicity of differences in a society increasingly align along a single dimension and people increasingly perceive and describe politics and society in terms of ‘Us’ versus ‘Them’”. She proposes that this worrying phenomenon is a prototypical example of an emergent property in an interpersonal dynamic system, which could be better understood by applying this theoretical framework. Both Butler and Suhay mention potential factors responsible for social polarisation, such as sharing political opinions or different group norms. In contrast, Webster and Albertson (2022) in their review article on “Emotion and politics” place major emphasis on the role of emotion in “biased information processing” and “affective polarisation.” As to the former, they cite the work by Suhay and Erisen (2018) mentioned above. As to the latter, “affective polarisation” refers to the strong negative feelings expressed by adherents of opposed parties towards each other, for example democrats and republicans in the United States. Iyengar et al. (2019) review the literature on this rapidly spreading phenomenon, tracing its origins to the power of partisanship as a social identity, and exploring the factors that intensify partisan hostility. A frequently voiced opinion attributes the increase in partisan hostility to biased media coverage of political events.

A study by Lu and Lee (2019), using a large national survey sample in the US, provides support for this notion, showing that consumption of pro-party television programs is more likely to induce negative emotions such as anger and fear toward the out-party candidate, which in turn, leads to a higher level of affective polarisation. Using the framework of appraisal theory of emotion, the authors hypothesise that anger is induced when one perceives (a) others being responsible for negative events and (b) a violation of certain standards. This toxic sequence is likely to be catalysed by TV programs with a partisan slant blaming the other side for national problems, and inform the audience about their opponent's misconduct and/or the peculiar claims made by the other side. Fear, by contrast, could be seen as arising from (a) threatening stimuli in one's environment associated with (b) a lack of certainty and personal control, suggested by biased reports of negative consequences of political programs represented as potential threats to personal well-being. A study by Gill (2022), using a large national sample, supports this hypothesis, focusing on the central role of anger. The data show that cable TV news, a medium frequently described as uncivil, tends to increase negative affect toward the out-party candidate, an effect mediated by feelings of anger and certainty in one's political attitudes.

The most disturbing consequence of increasing polarisation is the decrease in confidence and trust in political and social institutions. In a recent article, Brady and Kent (2022) report survey data documenting fifty years of declining confidence and increasing polarisation in trust in political and nonpolitical American institutions. The increase in polarisation and the decline in trust has been particularly dramatic during the period from 2010-2021. The authors conclude that this development will make it increasingly difficult to fight epidemics, maintain faith in policing, and deal with problems such as climate change.

Concern with this major threat to social relationships and democratic functioning is rapidly increasing in the research community. In a Special Feature of the Proceedings of the National Academy of Sciences on the dynamics of political polarisation (see Levin et al., 2021, for an overview) political scientists and systems theorists approach this development from different angles. While most of these are concerned with collective phenomena and system dynamics, Kawakatsu et al. (2021) propose a multi-level dynamic approach that analyses the danger of partisanship for interindividual cooperation, opening a perspective that includes individual behaviour and social interaction.

In concluding this editorial, I would like to briefly point to the role of emotion in these processes and sketch out a theoretical framework that may be of interest to both cognition and emotion scientists as well as political and social scientists. At this stage of the developing research efforts in this area, it seems important to identify potential paths towards interdisciplinary collaboration on the major threat for social cohesiveness and democratic governance represented by affective polarisation.

As described above, it is likely that biased information processing (such as “biased assimilation”) is a central causal factor for the development of affective polarisation, which in turn is likely to further strengthen biases in information processing, a powerful example of cognition-emotion interaction cycles. As to the underlying mechanism, in researching the issue, I came across the “predictive coding” approach, a “new look” computational framework on how the brain processes information (for a comprehensive account see Clark, 2013a; 27 commentaries on this article and a rejoinder of the author can be found here http://www.frontiersin.org/Theoretical_and_Philosophical_Psychology/researchtopics/Forethought_as_an_evolutionary/1031). Briefly summarised, the authors propose to change from the classic bottom up approach (the brain collects and organises information from sensory data), to a top down approach according to which a structure of superordinate expectations constrains and interprets incoming sensory information by comparing them to the predictions based on expectations. The assumption is that the structure of prior expectations is relatively stable as it represents the best fitness estimate for the individual. Error minimisation consists of neglecting small deviations, assuming that an error will have minor consequences. Only if the prediction error becomes very large, expectations will be adjusted. In recent years, this framework has been applied to many different phenomena, including mental disorder such as delusions or schizophrenia (Sterzer et al., 2018).

Wheeler et al. (2020) suggest that the predictive coding paradigm can also be used to understand ideological polarisation by outlining how a “process of shared error minimisation may lead to shared ideologies and beliefs that allow group members to predict and cooperate with each other, and how, as a consequence, political polarisation and extremism may result” (p. 192). They propose the following mechanism: “since group members learn from, motivate, and regulate each other's beliefs, mutually constrained error minimisation may align individual experience and behaviour with shared beliefs and schemas that allow us to predict and cooperate with others. However, as a byproduct, particularly under conditions of uncertainty, this mutually constrained error minimisation may produce increased polarisation and extremism.” (p. 196)

Emotions are rarely considered in the literature on predictive coding. This is surprising as the superordinate expectations that determine the evaluation of sensory information consist of beliefs or convictions. Frijda and Mesquita (2000) have provided an extensive discussion on the relationships between emotions and beliefs, arguing that acute or dispositional emotions create and strengthen beliefs, as well as making them resistant to change. Furthermore, emotions influence the credibility of information and inferences, and the estimated likelihood that information is true, via affective biases in thinking. Socially shared beliefs are particularly important in the case of moral convictions, characterised by a very high level of emotionality (Ciuk & Rottman, 2021; Skitka et al., 2021). In consequence, it would be useful to devote further research efforts to improve our understanding of the origins and developmental history of beliefs and convictions.

Affect and emotions may also be involved in the process of determining the degree of fit between expectations and incoming sensory information. The literature on predictive coding mainly deals with prediction error in terms of the amount of discrepancy between a predicted and an observed outcome. It is likely that the evaluation of the discrepancy consists of a more complex appraisal process, similar to what appraisal theorists consider as the elicitor of an emotion episode. The appraisal outcome determines the nature of the resulting emotion, depending on evaluation criteria such as expectancy, valence, causation, coping potential, and norm coherence or moral acceptability (Ellsworth & Scherer, 2003). Thus, the prediction error is likely to be quite large in cases where the outcome of an event is evaluated as unexpected, negative, caused by another person, difficult to deal with, and violating social norms or moral prescriptions. It might be argued that top down processing in the brain occurs very rapidly and largely unconsciously, unlikely to generate an emotion experience (see Clark, 2013b, p. 241). However, appraisal theorists assume that appraisals occur on different levels of neural processing – sensorimotor, schematic, and conceptual (Leventhal & Scherer, 1987; Scherer, 2009; van Reekum & Scherer, 1997). Thus, even very rapid, unconscious evaluation processes can create complex judgments with accompanying affect or emotion. In fact, it might be the strength and the nature of the affect that determines whether the prediction error will lead to a change or rather a strengthening of existing convictions.

Given the steady increase in social and political polarisation and resulting upheaval in many societies, often generating public discord and even violence, a better understanding of the underlying causes and the nature of the evolving processes is of highest importance to allow elaboration of potential intervention procedures. This Theory Section shows that there is already a large body of research directly relevant to these concerns and suggests avenues for further research. However, it also reveals a certain lack of interdisciplinary orientation that can connect theoretical strands dealing with cognition, emotion, and social behaviour, factors that are closely intertwined and need to be studied in an integrative fashion. I therefore call for coherent interdisciplinary efforts between neurosciences, cognitive sciences, emotion sciences, and social and political sciences to remedy this situation. A strong theoretical framework, with multiple factors and an emphasis on dynamic systems, will provide the essential basis for large-scale, longitudinal, empirical studies across several cultures. I hope this Theory Section can move us closer to ongoing integration between the scientific study of political polarisation and emotion science – and, eventually, to recommendations for potential interventions to prevent further deterioration of the social fabric in modern societies. It is an issue of high priority and an excellent topic for interdisciplinary convergence on theory and research strategies. To be revisited.