The “affective” in affective polarization

Public opinion in many contemporary democracies shows increasing signs of polarization. In particular, animosity and dislike across political lines have become increasingly characteristic of Western mass publics (Boxell et al., 2024; Finkel et al., 2020; Wagner, 2024). Several empirical methodologies designed to measure citizens’ opinions, which include questions about their feelings, started to report a shift in partisans’ emotional self-attributions. Specifically, there has been an increase in the distance between positive feelings toward the ingroup and negative feelings toward the outgroup. This is, in a nutshell, affective polarization (Iyengar et al., 2012).

Affective polarization doesn’t necessarily imply an absence of beliefs: citizens’ political opinions may still play a role in these polarization processes. It simply emphasizes that the most prominent attitudes involved in these polarization processes are affective rather than cognitive in nature.

But what is the difference between cognitive and affective attitudes? To intuitively grasp this difference, imagine two guys chatting about, say, Taylor Swift, each with different opinions about her music and influence. Let’s consider two possible scenarios. In the first scenario, these guys neither like nor dislike Taylor Swift; they’re simply intrigued by the Taylor Swift phenomenon, and just want to discuss what they think about it. In the second scenario, one of them loves Taylor Swift and passionately defends her, while the other hates her music and what she represents.

The second of these two scenarios typically involves affective attitudes, i.e., emotion- and desire-like mental states such as loving or hating, whereas in the first scenario, the more prominent attitudes involved will surely be belief-like mental states that are still compatible with positive or negative valence.

Affective polarization has more to do with the second scenario than with the first one: it is characterized by involving salient affective attitudes.[1] However, there is a lack of conceptual clarity regarding “affect” in the literature on affective polarization. Researchers in this field have not reached a consensus on its definition, and then operationalize it in different ways. For example, surveys sometimes ask about “cold and warm feelings,” while other times they use expressions such as “dislike,” which is often used interchangeably with “antipathy,” “animosity,” or even “distrust,” to name just a few expressions found in the literature (see Rollicke, 2023).

Attempts to address this lack of clarity typically focus too heavily on empirical research, which, despite the value of their contributions, leaves us almost where we started (see, for instance, Bakker & Lelkes, 2024, and Webster & Albertson, 2022). So we are in an urgent need of a philosophical analysis of the “affective” in affective polarization.

We might begin to address this lack of conceptual clarity by distinguishing emotions from evaluations. When responding to opinion surveys, citizens might be reporting their emotions, but they could also be making evaluations. Let’s first consider the possibility that they are reporting their emotions. If that’s the case, what exactly are they reporting?

In the fifth chapter of his book, The Concept of Mind, Gilbert Ryle argues that the term “emotion” can be used to refer to at least three distinct categories, which he labels as “inclinations,” “moods,” and “feelings.” “Feeling” refers to what somebody experiences internally when they have, for example, episodes of joy, comfort, anger, rage, etc. Thus, feelings are phenomenal episodes or sensations, characterized as occurrences, i.e. episodic experiences that one undergoes.

Inclinations, on the other hand, include emotions such as vanity, avarice, patriotism, or kindness. These are characterized as dispositions or propensities to behave in certain ways, including having certain feelings. For example, to be an irascible or kind person is to have the disposition to act as irascible or kind people typically do. Finally, “moods” are also forms of dispositions or inclinations but are temporary. A person might be in a happy, angry, sad, or melancholic mood fora while, which means they have a temporary disposition to behave and feel in ways characteristic of these moods.

Thus, when we attribute to someone an emotion –e.g. anger–, we might be claiming at least three different things: they are an irascible person, they are in an angry mood, or they are experiencing sensations of anger. Therefore, when affectively polarized people report having certain negative feelings, dislike or animosity toward particular topics or individuals, they might be pointing to their feelings, their moods, or their inclinations.

But there is another possibility: maybe they are not reporting their emotions in any sense, but are rather making evaluations, i.e. expressing their non-cognitive attitudes toward certain political views.

Non-cognitive attitudes are also affective attitudes, but they encompass more than just emotions; they also include values, convictions, perspectives, and identities (Beaver & Stanley, 2023; Lynch, 2022). Back to our previous example: when someone says they love Taylor Swift, they might be conveying their emotions, whether dispositional or episodic. But this person is also expressing a set of values, a perspective, and an identity. We can infer and predict a lot beyond their emotions about someone who expresses love for Taylor Swift and fervently defends her.

Similarly, we can infer a lot about people who say they have cold feelings toward Democrats, liberals, or feminism. This is so because our evaluations –the non-cognitive attitudes we express– are inherently connected to specific values, perspectives, and practices.

Thus, when respondents to these surveys claim they have cold feelings toward a certain issue, that they “dislike” certain partisans, or that they would be displeased if their son or daughter married a conservative/liberal, they may not be (merely) reporting their emotions or feelings. Instead, their responses could represent negative evaluations, through which they express their non-cognitive attitudes. That is, their answers could be a way of negatively evaluating a certain political perspective, which provides information about the political parties they are likely to vote for, the political perspectives they have, and even to what degree they are affectively attached to such perspectives (Almagro, forthcoming).

These two ways of interpreting surveys’ responses suggest very different pictures of what affective polarization is. According to the first, the “affective” is confined to our emotions, whether dispositional or phenomenal. According to the second, the affective aspect encompasses positive and negative attitudes toward certain political identities and narratives. Each picture might call for very different detection and intervention strategies, so it could be crucial that they are not conflated.

Considering the diverse terminologies used in questionnaires to track affective polarization, it seems we’re faced with two possibilities: either affectively polarized individuals are reporting a surge in various emotions, or this terminology represents different ways of measuring essentially the same thing –namely, the non-cognitive attitudes they hold toward specific political identities and narratives. My money’s on the second option.

References

Almagro, Manuel (Forthcoming). The Rise of Polarization. Affects, Politics, and Philosophy. Routledge.

Bakker, Bert N. and Yphtach Lelkes (2024). Putting the affect into affective polarisation. Cognition and Emotion, 38(4), 418–436. https://doi.org/10.1080/02699931.2024.2362366

Beaver, David and Jason Stanley (2023). The politics of language. New Jersey: Princeton University  Press.

Boxell, Levi, Matthew Gentzkow, and Jesse M. Shapiro (2024). Cross-Country Trends in Affective Polarization. The Review of Economics and Statistics, 106(2), 557–565. doi: https://doi.org/10.1162/rest_a_01160

Finkel, Eli, Christopher A. Bail, Mina Cikara, Peter H. Ditto, Shanto Iyengar, Samara Klar, Lilliana Mason, Mary C. McGrath, Brendan Nyham, David G. Rand, Linda J. Skitka, Joshua A. Tucker, Jay Van Bavel, Cynthia S. Wang and James N. Druckman (2020). Political sectarism in America. Science, 370(6516), 533-536.

Lynch, Michael P. (2022). Memes, misinformation, and political meaning. The Southern Journal of Philosophy, 60(1), 38-56.

Ryle, Gilbert (1949/2009). The Concept of Mind. London: Routledge.

Wagner, Markus (2024). Affective polarization in Europe. European Political Science Review, 1–15. https://doi.org/10.1017/S1755773923000383

[1] Note that the affective scenario could also involve cognitive attitudes: the parties still have different opinions. The main difference between both scenarios is just how salient affective or cognitive mental states are.

Picture is produced by ChatGPT