Chapter 1 opens with a narrative detailing the countless ways in which we interact with others about politics on a daily basis. We illustrate that previous research on the topic has left several assumptions untested, such as that individuals feel “upset” or “anxious” when faced with political disagreement, that social harmony can affect our political discussion behavior, and that the decision to engage in or avoid a discussion is an active choice. The goal of this chapter is to summarize the book’s key contributions, persuade readers of its importance, and preview the remaining chapters.

Chapter 4 commences the empirical tests of our theory, beginning with Stage 1 of the 4D Framework: detection. We directly tackle a question buried implicitly in previous findings, as well as our own, that people prefer like-minded discussants: How do people detect the political views of others? The stakes of discussion may be higher in a polarized environment, but the readily available cues stemming from a divided and politicized society make the process of sorting into amicable discussions easier. We show that individuals are able to use a variety of cues to infer political leanings, including more obvious cues like demographic characteristics and extremely subtle cues, such as first names, pet preferences, and movie preferences. We then explore the existence of stereotypes that individuals hold about partisans, under the assumption that these attitudes could affect our ability to recognize others’ views and our willingness to engage in a discussion. We find that, consistent with research on affective polarization, individuals ascribe more negative personality traits to outpartisans and consider them to be ill-informed, ignorant, and overly reliant on partisan media.

In Chapter 2, we outline the theoretical core of our inquiry. To fully understand the experience of political discussion, we must think more broadly about the full set of considerations that structure people’s decisions. We introduce the concept of the 4D Framework to the process of political discussion, articulating what happens at each of four stages preceding, during, and after the opportunity to discuss politics. The stages include Detection, Decision, Discussion, and Determination. Individuals make choices at each stage of the cycle as a result of their unique individual dispositions and the social context, both of which contribute to unique motivations. To rigorously examine the motivations behind political discussion preferences, we develop the AAA Typology, which characterizes motivations as accuracy, affiliation, and affirmation. We argue that in contrast to previous research emphasizing instrumental goals in political discussion (i.e. to learn more information, an accuracy goal), most individuals are driven to preserve their self-esteem (affirmation) and the social ties with their potential discussants (affiliation).

Chapter 8 considers Stage 4 of the feedback loop: Determination. We examine how individuals anticipate relationships changing after political conversations and how discussion behavior is correlated with social distancing and social polarization. We use nationally representative survey data to capture individuals’ reflections on their own social distancing behaviors as well as their projections of such behavior onto hypothetical characters in vignette experiments. We uncover that about a quarter of Americans have distanced themselves socially from a friend because of politics. Americans have done so in a variety of ways, including stopping all political discussion, forbidding their children from playing together, and severing all social ties completely. Vignette experiments revealed that individuals are more likely to avoid future political and social interactions with others who disagree with them. Using data from the 2018 Cooperative Congressional Elections Project, we find that strong partisans in the most like-minded discussion networks were more likely to be socially polarized, compared to strong partisans who were in disagreeable discussion networks.

Chapter 6 offers the first look into the Discussion stage (Stage 3) of the 4D Framework. We focus on what people feel during a political discussion, captured using psychophysiological measurement during two different lab experiments. In the Psychophysiological Anticipation Study, we measure changes in participants' heart rates and skin conductance as they anticipate a political discussion. We find that individuals had a larger psychophysiological response to even the thought of engaging in a political discussion, compared to observing contentious discussions on video. In the Psychophysiological Experience Study, we measure variation in heart rate and skin conductance during real conversations. We find that individuals exhibit physiological signs of discomfort while in these conversations, especially when the conversation is disagreeable.

Under what conditions are people most likely to discuss politics? Our focus in Chapter 5 is on the moment of decision itself (Stage 2). We use three novel approaches to answer this question. The True Counterfactual Study asked participants to reflect upon and describe either political discussions in which they had recently engaged or political discussions in which they could have engaged, but chose to avoid. Comparing these descriptions revealed that avoided discussions had larger groups with more disagreement. We then used vignette experiments to manipulate various features of a conversation, finding that individuals were more likely to avoid a discussion if they were in the political minority, less knowledgeable than the others, or conversing with weak social ties. The Name Your Price studies asked people to report how much they would need to be paid to discuss various topics with different groups. Individuals demand more compensation to discuss both political and nonpolitical topics with those who disagree, especially when that disagreement is defined in terms of partisan identity.

Chapter 7 continues our examination of the Discussion stage (Stage 3) of the 4D Framework, aiming to better understand what people actually verbally express in political conversations. We use a series of vignette experiments and a lab experiment to examine the extent to which people express their true opinions to the group, or engage in other expressive behaviors like self-censorship, silencing, or conformity. We find lab experimental evidence of conformity in real conversations and our vignette experiments revealed that individuals were less likely to reveal their real opinions when they were in the political minority and when they were less knowledgeable. We also find that individuals who were less knowledgeable were more likely to have affiliation concerns that explained their expression behavior. Finally, we analyzed the transcripts from the conversations in the Psychophysiological Experience Study to examine variation in what and how individuals discussed politics. We found that weak partisans were most likely to "hedge" their language when revealing their opinions and political identities.