It is often alleged that polarization threatens democracy. Yet, the nature of the problem polarization poses for democracy often remains undertheorized. Any philosophical theory of polarization’s relation to democracy must answer four key questions. It must explain the nature of polarization. It must diagnose its causes. It must explain how polarization impugns democracy ideals or threatens to undermine democratic institutions. Finally, it must explain how polarization might be reversed, prevented, and managed.

Robert Talisse has recently developed a novel theory of this kind. He contends that citizens are prone to belief polarization, a process by which like-minded groups’ interactions move them towards increasingly extreme positions and that this exacerbates political polarization. Talisse believes that the resultant dynamic undermines our ability to relate to one another as equals and thus undercuts the democratic ideal of self-government among equals. Talisse argues that since polarization is driven by pressures internal to democratic citizenship, polarization can at best be managed and offers a distinctive account of what citizens must do to sustain democracy.

This review provides an overview of Talisse’s position as articulated in Overdoing Democracy: Why We Must Put Politics in Its Place (2019), Political Argument in a Polarized Age: Reason and Democratic Life (2020), and Sustaining Democracy: What We Owe to the Other Side (2021).Footnote 1 I first sketch Talisse’s account of democracy and his understanding of polarization (1). I then summarize the challenge Talisse believes that polarization poses (2) and outline his proposal for how citizens should respond to it (3). I argue that Talisse overstates the degree to which polarization arises from the good-faith practice of democratic citizenship and downplays the extent to which polarization is caused by elites and exacerbated by social structures; this leads Talisse to overlook structural approaches to managing polarization and leaves his account of how citizens should respond to polarization incomplete (4). I conclude by suggesting that Talisse’s account can be integrated into a broader approach to polarization, which can avoid these shortcomings (5).

1 Preliminaries: Democracy and Polarization

Democracy, for Talisse, denotes more than a form of government. Democracy is the moral ideal of self-government among equals. Talisse believes democracy is not only arguably good in itself but that it is a “capital social good” (2019: 12), which makes possible a range of other valuable social goods (2019: 172). Part of any conception of democracy as a moral ideal is a conception of democratic citizenship. If democracy is self-government among equals, democratic theory needs an account of how citizens should conduct their lives together. Talisse emphasizes that the democratic ideal demands much of citizens. It demands that they promote justice, that they offer one another reasons in support of their views, and that they do all this while treating one another as equals. Talisse argues that it is herein—in the demands of the office of democratic citizenship—that polarization has its roots.

Talisse distinguishes several meanings of ‘polarization’ (2019: 97–106; 2020: 30–38; 2021: 73–91). Political polarization refers to the “ideological distance between two or more political groups” (2021: 73). This can be understood in different ways—as (e.g.) the distance between party platforms (platform polarization) or as the degree to which rank-and-file citizens embrace negative attitudes towards the opposition (popular or affective polarization).Footnote 2 However, Talisse argues that political polarization is driven by a more basic phenomenon.

Belief Polarization is a process that occurs within groups: interactions between like-minded people tend to result in each person adopting a more extreme version of their shared views (2021: 79). This process is well-studiedFootnote 3 and occurs in basically any homogenous group: socialists interacting with other socialists become more socialist; racists interacting with other racists become more racist; curling-fans become more passionate about curling. This process, Talisse innovatively suggests, is driven by perceived corroboration: “Corroboration by our peers makes us feel good about our shared beliefs and thus makes us feel affirmed in our group membership, our social identity” (2021:88).Footnote 4 This moves us towards extremity: we become more attached to and confident in increasingly radical beliefs.Footnote 5 Belief polarization and political polarization “work together in a degenerative spiral,” which Talisse calls the polarization dynamic (2021: 93).

Talisse believes that the polarization dynamic has its roots within democracy. It emerges from pressures internal to democratic citizenship. Democracy requires that citizens work to advance justice. This requires forming political alliances with like-minded compatriots—just the groupings wherein belief polarization can take hold (2019: 80–83). Democracy requires of citizens that they deliberate together and exchange reasons. This generates pressure to expand “democracy’s reach”: to treat all social spaces as presumptively venues wherein democratic politics to be enacted. Social space thus becomes saturated by politics. This heightens exposure to belief polarization: we express our political convictions more often and receive greater affirmation from ever more like-minded compatriots. And so, it occurs that well-intentioned democrats are led down the road towards polarization.

2 How Polarization Threatens Democracy

This road, Talisse argues, threatens the enterprise of democracy itself. Talisse’s argument is first detailed in Overdoing Democracy, the central thesis of which is that the political saturation of social space undermines democracy. Talisse argues that democracy depends on the capacity of citizens to sustain their investment in democracy in the face of what they regard as “momentous normative lapses and errors”: citizens must remain faithful to democracy even when it yields results that they regard as deeply unjust (2019: 144). To do this, citizens must be able to regard their compatriots as their “civic friends”: they must view their opponents as reasonable, good-faith actors who remain committed to democracy. Polarization erodes the capacities required for democratic citizens to do this—our capacities for “democratic sympathy” (2019: 147). As affective antipathy towards the opposition mounts, we become increasingly unable to take the perspective of our political rivals. Thus, we come to regard the political opposition as unreasonable and unfit for democratic citizenship (2019: 122–4; 2020: 39–41). Overdoing democracy thus undermines democracy.

Overlooking Democracy describes how the problem of polarization looks from the outside. Sustaining Democracy considers how matters appear to citizens engaged in democratic politics. Talisse suggests they face what he calls “the democrat’s dilemma.” As a citizen, one is rightly moved by one’s sense of justice (2021: 47): one’s actions and identity reflect one’s convictions about what is just. Yet, democracy invariably involves deep disagreement. Citizens thus find themselves at odds about what justice requires. Under such conditions, it can come to seem not only naïve but positively unjust to uphold democracy. Democracy requires accepting that justice may lose out and this can seem like capitulation to injustice. The question—why not suspend, at least temporarily, our democratic commitments to advance justice? —presses. This stands in tension with the commitment to treating those with whom we disagree as our equals.

The dilemma arises from within the Janus-faced ethos of democratic citizenship, which asks that citizens pursue justice and treat those at odds with justice as their equals (2019: 165–6). Polarization exacerbates this predicament because it threatens the capacities required to view our opponents as equals. We thus come to regard our opponents not as mistaken democrats acting on their sincere commitments, but as “a monolithic, unnuanced, irrational, untrustworthy, and depraved mob” (Talisse 2021b: 142). Reasonable political disagreement with them seems impossible. Polarization can thus shake our own commitment to democracy, for we become increasingly willing to disregard the demand that we treat those with whom we disagree as equals.

3 Sustaining Democracy in the Face of Polarization

Talisse’s argument raises two key questions. The first is normative: why should we sustain democracy under conditions of polarization? Why relate to our rivals as equals, given that they appear so unreasonable? The second is practical: how can we sustain democracy? How might citizens undo the polarization dynamic?

Talisse’s answer to the normative question comes in two parts. One part consists in reminding us of democracy’s value and that justice itself requires treating one’s political adversaries as equals (2021: 41). Yet, this alone does not fully address citizens in the grip of the democrat’s dilemma. So, Talisse offers a second answer: that managing belief polarization is necessary to preserve our political alliances. The more polarized our beliefs become, the more our capacities to sustain democratic relations with our allies erode. Belief polarization rears its head within our political coalitions, which become less accepting of reasonable disagreement (2021: 99). We thus lose our capacity to maintain equality in the face of disagreement. Our alliances thus fracture, which undermines our ability to pursue justice. Hence, “we need to uphold democratic relations with our reasonable opponents if we are to maintain the capacity to treat our allies as our equal partners in the pursuit of justice” (2021: 102–3).

But how can we reverse the polarization dynamic? Talisse argues that although “more democracy” solutions—that political rivals should deliberate more together, (e.g.)—may perhaps help prevent polarization, they cannot reverse polarization. Instead, when members of opposed parties come together in politically saturated spaces, this can exacerbate their animosity towards one another (2021: 117–121; 2019: 21 − 5). This claim will likely prove controversial, as the evidence for Talisse’s position seems mixed.Footnote 6 But Talisse’s alternative is worth considering. Talisse first argues that “in order to sustain democracy, we need to manage belief polarization within ourselves” (2019: 124). The requires recognizing our own vulnerability to belief polarization (2019: 165), learning to “regard our political views as works in progress” (2021: 125), focusing on “reasonable criticism of one’s own political views” (2021: 128), changing how one engages in political argument (2020: 104–114), and cultivating distance from the “politics of the moment” (2021: 134). These measures may help us recognize that we’re caught in the same polarization dynamic as our rivals and thereby affect a change in attitude towards them.

The second part of Talisse’s proposal involves putting politics in its place. Our social interactions are increasingly expressive of our political views and, due to social sorting, are increasingly politically homogeneous. This exacerbates the polarization dynamic as we more often express our political identities and receive greater corroboration for doing so. To break this dynamic, Talisse thinks we must immerse ourselves in nonpolitical activities which allow us to engage with our political opponents as something other than political rivals (2019: 163–5). This doesn’t mean retreating from politics; rather, Talisse’s point is that sustaining democracy requires occasionally doing something other than politics with one another. Just as leading a happy life requires adopting ends other than one’s own happiness, fostering civic friendship requires doing something together besides politics.

Despite the persuasiveness of Talisse’s argument, there remains some unclarity about its scope. Part of the force of Talisse’s account seems epistemic: in the grip of belief polarization, we are prone to misrepresent our reasonable rivals as unreasonable anti-democrats. But sometimes our rivals really are unreasonable: their views are “incompatible with a commitment to a society of self-governing political equals” (2021: 58). Members of resurgent authoritarian political movements plausibly fit this description. Talisse believes that “proper democratic citizens need not uphold democratic relations with unreasonable citizens” (2021: 57) but does not fully address how citizens should respond to those who are beyond the pale. The injunction that democratic citizens—particularly, racial and ethnic minorities—engage in non-political activities with (e.g.) white-nationalists seems at best counterproductive. Talisse perhaps agrees, but it is striking how little he says about such cases. Nevertheless, our political rivals don’t always espouse such noxious views and so Talisse’s account is an important intervention.

4 Polarization’s Causes and Solutions

Still, it is hard to pin down how exactly Talisse’s story about belief polarization fits into his larger story about political polarization. On a natural reading, Talisse believes that belief polarization is the primary driver of political polarization in actual democracies. However, if this is Talisse’s view, he seems to overstate his case.

One reason is that belief polarization—holding increasingly extreme views with an increasing level of confidence—doesn’t necessarily undermine the capacities required for democratic sympathy. The key mechanism linking belief polarization to democratic sympathy seems to be affective polarization. Yet, it is possible, both in theory and in practice, to undergo belief polarization without undergoing affective polarization. As Thi Ngyuen (2021: 80) has argued, one can imagine a sect of passionate Talissians congregating and becoming increasingly convinced we’re all overdoing democracy. Surely this wouldn’t erode their civic capacities. The extent to which belief polarization poses a problem for democracy thus seems to depend partially on the content of citizens’ beliefs and on the degree of affective polarization this dynamic triggers.

Yet, levels of affective polarization (which Talisse calls popular political polarization) markedly vary across countries.Footnote 7 This fact can’t be explained by social sorting and belief polarization alone as these forces exist to some degree in all societies.Footnote 8 It must either be the case that other social forces explain affective polarization or that social sorting itself has deeper social causes. There’s ample evidence for both hypotheses. It’s well documented that elites in the United States (Talisse’s main example) have become increasingly polarized in ways that outstrip their counterparts in most other countries and that this drives mass polarization (Zingher and Flynn 2018). It’s clear too that this process has been asymmetrical: in the United States (e.g.), elite-level Republicans are more polarized than Democrats.Footnote 9

There’s also evidence that other social and structural factors drive social sorting and political (and, in particular, affective) polarization. It’s no coincidence that political polarization is particularly dramatic in the United States. Although elections themselves are key drivers of affective polarization,Footnote 10 certain features of the American electoral system, such as its two-party system, and certain features of American society, such as its history of racial and class-based segregation, plausibly exacerbate polarization and social sorting. Talisse says little about these deeper structural factors, treating social sorting as largely the product of individual choice. Given that there is no shortage of opportunities for individual associative choice in both political and nonpolitical contexts in other democratic societies beyond the United States and yet many of these societies are less polarized, choice expansion and social sorting seems at best only part of the story. So, even if belief polarization and social sorting partially drive the polarization dynamic, a theory of these processes should take into account their deeper social and institutional causes.

There are more interesting ways to read Talisse’s thesis, however. One is that elite-level polarization and structural factors work together with belief polarization to create a looping effect: given the structural constraints of electoral systems, belief-polarized voters select politically polarized candidates, triggering the polarization dynamic. Alternatively, perhaps Talisse is not making an empirical claim about what drives polarization in particular societies; rather, Talisse is merely suggesting that belief polarization is a particularly important phenomenon for democratic citizens because it is an occupational hazard of the office of citizenship. Talisse makes both suggestions in recent work and both are highly plausible.Footnote 11

These alternative readings point to the need to integrate Talisse’s insights into a broader account of political polarization. However, once one broadens one’s perspective on polarization, Talisse’s suggested remedies for polarization look inadequate. Talisse recommends highly individualistic solutions to polarization: things democratic citizens can do to counteract polarization. Yet, if social sorting and belief polarization are the product of structural forces, these individualistic strategies will likely have a limited impact. Talisse’s diagnosis of polarization’s causes dictates his preferred cures. But that diagnosis is only part of the story. Once one sees that elite-level and structural factors shape polarization, it because apparent that managing polarization requires not only individual action but structural solutions.

What might such solutions look like? If the structure of an electoral system plays a role in driving affective polarization, then sustaining democracy may require election reform. If elections are themselves drivers of polarization, then this provides reason to take more seriously alternative democratic structures, like mini-publicsFootnote 12 or lottocracies,Footnote 13 or constraining the role of political parties. If social sorting is partially driven by racial and economic segregation, then sustaining democracy may require reforming housing policy, such as promoting mixed-income housing or politically diverse neighborhoods, or reigning in economic inequality, which drives facilitates spatial exclusion. Most modestly, states might seek to create more public spaces in which cross-party interaction of the kind Talisse recommends is viable by creating more mixed-use public spaces and non-political public events. The claim that such reforms would be efficacious is, of course, somewhat speculative, though it is no less speculative than Talisse’s own proposals. But I conjecture that such measures would do at least as much to reverse polarization as Talisse’s proposed remedies.

5 Conclusion

Talisse’s work constitutes an invaluable contribution to the political philosophy of polarization. Talisse’s claim that belief polarization is partly driven by pressures internal to democratic citizenship and his suggestion that reversing polarization requires that citizens engage in nonpolitical activities with one another are both plausible. Yet, just as Talisse claims that politics must be put in its proper place, Talisse’s insights serve democracy best when they are integrated into a broader approach to polarization, which takes elite-level and structural forces seriously.