Constructing and Invoking Partisan Identities Through Narrative

Introduction

The rapidly crumbling veneer of civil discourse and functional democracy in the United States forces us to seriously consider its evolving political antagonisms, how we got here, and where this could be heading. It is in these grim conditions of increasing partisan hostilities that Manuel Almagro’s new book, The Rise of Polarization (2025), strikes a chord within our own homes, across the Atlantic, and around the globe. One of the central insights developed in Almagro’s book is that polarization is not just a matter of how we feel and identify ourselves; polarization is also about the narratives we proclaim, how tightly we grip to them, and how these narratives are used to vocalize our political attachments.

Almagro’s tour through the labyrinth of polarization exposes a tendency in previous theories to oversimplify polarization as a phenomenon that, at the first level, only concerns our identities and our attitudes. He shows that while important, focusing solely on these dimensions leads to an emotivist diagnosis of polarization, where group antagonisms are understood with respect to how closely (or warmly) one self-identifies with a group, and how hostile (cold) they are towards other groups. Almagro convincingly develops throughout the work that these are only a couple of the many interrelated dimensions of affective polarization.

The addition of narratives into theorizing polarization radically alters our views of polarization and of the very identities and emotions at play. With the work of political narratives in sight, we can see not only how individuals are polarized or describe a situation in terms of polarization. We also see how political discourse itself is polarized. Moreover, the narratives central to political discourse are vitally polarizing; they shape and transform the very social world that narratives attempt to frame. As he summarizes:

How do narratives play a role in constructing political identities? While Almagro does not explicitly take up the significance of this constructivist view of identity directly in the book, he does give many valuable insights into the relationship between identity and narrative in polarizing processes.

This paper has two primary focuses. First, I uncover this narrative constructivist view of identity implicit in Almagro’s book, against the individualist account often promoted in two-dimensional views. In contrast, his account of the nature of social identities affords great importance to the role of social environments and social relations, along with an individual’s world-views and practical engagements. This yields the insight that each of us has multiple socially, ideologically, and practically determined identities—ones that we might have without knowing we do, and ones we believe we have but actually fail to. With respect to the role of narratives, we can ask: Do partisans construct their own identity by using narratives, or are identities constructed within the narratives of partisan discourse? I highlight a key distinction between constitutive construction of narrator identities and discursive construction of narrated identities. While it is not entirely clear which sense Almagro intends at times, both constructivist senses are important as they co-create the networks of identities and affect relations in partisan narratives.

Second, and more critically, I raise the challenge that Almagro’s constructivist commitments betray part of his expressivist diagnosis of polarization. If identities are not simply forms of self-attribution, internal attitudes, or psychological states, then social identities are not expressible the way other mental phenomena are. In short, the expressivist model does not fit a constructivist view of identity. Identities and ideologies are constructed and participated with through social engagement and political discourse. They are embodied by the partisans who invoke them. I suggest that partisans “give voice” not just to their own attachments, but to certain identities and ideologies themselves by using divisive narratives. While partisans become affectively attached to certain ideological and identity constructions, this affective condition is created by partisans invoking identities through narrative construction. Thus, ideologies and identities are not something possessed by a partisan; perhaps instead they possess the partisan. I offer invocation as a conceptual model to better understand the participatory and socially constructing aspects of partisan identities, emotions, and narratives, which avoids the internalizing or essentializing tendencies in expressivist treatments of identity.

Our political lives are deeper than identities and narratives, and our emotions are more complicated than group antagonism. Almagro understands this personally, and it’s one of the reasons his personal vignettes are a breath of fresh air within sharp theoretical scholarship. Academics can often float in abstract theorization and miss the everydayness of our political worlds. Narratives of deportations are not deportations. In February of 2025, two weeks after Trump's second inauguration, my neighbor was abducted by ICE. Not only my neighbor, he is a husband, a father of a then-newborn child, a long-time Connecticut resident, and a worker in town. While in this paper, the discussion is largely at the level of political identities and narratives, I acknowledge that the reality of partisan politics is grim and embodied in the lives of the detained, incarcerated, starved, and mourning. It is with such people in spirit that I hope any subsequent insights into affective polarization and partisan politics may serve.

1. Polarized Identities

Almagro builds upon and contrasts his views with what he calls two-dimensional views of affective polarization. According to these kinds of theories, at its core, affective polarization is about the ways partisans identify in politics and the way partisans feel about politics. In this sense, the two dimensions involved in affective polarization are political identities and partisan attitudes.

While discussing two-dimensional views of polarization, Almagro briefly takes stock of the notion of identity utilized by such theories. Since the two major components of these views are identity and emotion, a significant deal resides in their conceptualization of these phenomena. We will focus here on both the kind of identities theorized, the metaphysics of what an identity is, and how one comes to “have” an identity. The upshot of this will be that Almagro not only adds additional dimensions to understanding polarization, but also has a different conceptualization of identity than the two-dimensional views he critiques. Almagro is right to utilize a more robust notion of identity, rather than the limiting account in two-dimensional views that conflate political identities with self-identities.

Much like life and boxes of chocolates, identities come in all different kinds, and in scholarship, you’re often not sure what you’re going to get. For our purposes here, we will consider three distinct, yet related, kinds of identity: personal identity, social/political identity, and self-identity. Our personal identity is the sense of being me, as philosophers have long asked, who am I, and what makes myself ‘me’. Our social/political identities concern relationships, whether relations with people such as being a mother, or relations with ideologies such as being a conservative. Our self-identities are how we identify ourselves, a form of self-attribution concerning the identities we take ourselves to have or desire others to view oneself as having.

The primary source of the two-dimensional view that Almagro engages in the book is the joint work of Santo Iyengar, Gaurav Sood, and Yphtach Lelkes. In their 2012 research on affective polarization, they use survey data over the past fifty years to compare partisanship with in-group out-group emotional valencing. They find that respondents were reporting increasingly severe “cold” feelings towards those in the out-group over the past fifty years, while “warm” feelings towards the in-group have stayed relatively consistent (Iyengar et al. 2012). Crucially, their study relied upon what they call “social identity theory” which takes an individual’s salience of self-identification as a basis from which to predict inter-group prejudice. To measure partisanship, the surveys used in their research had respondents self-report their identification by asking which party they support, how strongly they identify themselves as Democrat or Republican, or whether they consider themselves “closer” to either party. In this study, we see implicitly the conflation of self-identity and social identity, which is made explicitly in Iyengar’s latter work, where it is claimed,

The role of self-identification is important to partisan political identities; however, this model of partisanship as self-identification oversimplifies the nature of partisan identities, and as Almagro convincingly argues, focuses on a simplistic “Identity First” causal explanation of affective polarization as caused by identification. Moreover, we see Iyengar et al. attribute the division of in-group and out-group as a result of identifying with a political party. These are significant social metaphysical claims about the role of self-identification regarding group membership, suggesting that one is not already immersed and engaged in a host of in-groups before the activity of self-identification. When we step outside of viewing partisanship as a political self-identity, we begin to view partisan identities among the myriad other features of our political environment. Might the intensity of our self-identification be a product of increasingly hostile attitudes driven by divisive narratives used by our families, coworkers, neighbors, news media, and ourselves?

This approach to understanding affective polarization and partisanship is explicitly taken up by Almagro in forging the multi-dimensional account of polarization. Ultimately, not only are narratives crucial for conforming to political identities (Brankovic 2022; Levi et al. 2023; McLaughlin et al. 2019; Warner 2019), Almagro also argues that “political identities and the emotions associated with them are significantly shaped by narratives and the reasons underlying them” (Almagro 2025, 54 emphasis added). Here we begin to see a multidimensionality emerge regarding the reciprocal influence between identities, emotions, and narratives. What most interests us for the purposes here is Almagro’s claim that affective polarization involves “the construction of certain identities around such narratives” (ibid., 57).

2. Narrating Identities

While Almagro does not provide explicit details of a constructivist theory of political identities, his further theoretical insights into the nature of political identities do accord well with constructivist accounts. I will connect his views with scholarship on social construction to better theorize the narrative construction of partisan identities. First, we can distinguish constructivist accounts from individualist theories of social ontology, where features of individuals (e.g., patterns of attitudes or actions) are key determinants of social phenomena, as in the case of partisan identity being determined by individuals’ self-attribution in the previous section.1

Moving beyond individualism, Almagro discusses at length how our identities are not essentially or primarily of our own choosing. Some identities are quite explicitly not up to an individual’s choices, such as racial, national, and kinship identities. At the same time, deciding to enroll in educational programs, work in certain occupations, raise children, or join political parties are all crucial factors of our social identities. These decisions and identities are embedded in our life activities, experiences, and shared social worlds. In short, actions that carry great significance for our identities are possible given certain opportunities and influences afforded by our particular social environments. As such, identities are not merely features of an individual; they are just as much a part of our social worlds. Following Almagro, it is clear that persons possess multiple identities, where we are part of many social groups, as categorized according to our relations and engagements.

A key insight here is that we do not come pre-equipped with many of our identities, nor are they of our own making; they are acquired by our forms of participation within social worlds. Already, there are strong resonances here with Judith Butler’s theory of constructed identities as performative accomplishments through constitutive actions (Butler 1988). Crucially, Butler argues that performatively accomplishing an identity is not just up to our own actions, but also depends upon an audience believing that a person’s actions have successfully constituted an identity. In this sense, cultural perceiving and identity-assigning practices are dispersed among “actors” and “audiences”.

Almagro takes up these themes when addressing that our political identities “are grounded in our worldviews, which are in turn rooted in practices— forms of life” (Almagro 2025, 79). Significantly for Almagro’s work, he distinguishes between the political identities one does have and the identities one believes they have (ibid., 79). This distinction is developed through the practical and perspectival requirements of political identities. Partisanship involves adopting certain worldviews and ideological perspectives in the abstract, while also appearing to require that these views impact the concrete everyday practices of partisans. Insofar as the abstract beliefs of partisans do not match their concrete activity, it stands to question whether they have a political identity or merely believe that they do. Further, Almagro presents the issue of misalignment concerning the political identities a person attributes to theirself and the identities others attribute to them. I may fail to have an identity that I believe I have, and others may not attribute that identity to me. Similarly, I may have an identity I do not believe I have, while others do attribute that identity to me. The attribution of others could likewise be wrong about a person’s identity. The point here is that there is some real sense to whether or not one has an identity, and this identity is not fully determined by the attributions of any individual, whether myself or any other person. The question is then: how do we come to have certain identities, and how does attribution impact our identities?

To answer this question, we must look more closely at what is involved in social construction.2 Sally Haslanger (2012) distinguishes between three kinds of social construction that are important for our purposes of understanding the narrative construction of identities: causal, constitutive, and discursive constructions.

First, something is causally constructed when social factors play a causal role in its existence, or social factors are causally responsible for something being the way it is. We see causal social construction at play when we understand narratives espoused by partisans not merely as a result of their own qualities but as caused by their social environment and previous experiences.

Second, when social factors are necessary components in defining something, it is constitutively constructed. In this sense, partisan identities and political narratives themselves are constitutively constructed, where social features are centrally important to the kind of partisan identities and narratives involved: party identities and narratives about gender politics are defined, in part, by social factors.

Third, when something is the way it is largely due to the attributions of it, that thing is discursively constructed. By naming, classifying, characterizing, describing, etc., these attributions become elements of the thing so discursively embellished. Identities and their associated qualities are key examples of discursive constructs precisely because of the role of narratives in constructing classificatory schemes that in turn guide our judgements and evaluations.

Before returning to see how narratives play these constructive roles, we must first get clear on what narratives are in the sense used here. I will primarily follow Almagro’s conception regarding the general phenomenon of narratives, while I point to the smaller-scale discursive phenomena where narratives find their way into everyday speech. While there are many types of narratives, such as personal, literary, religious, and so forth, the focus here and in Almagro’s work is on political narratives. Almagro provides a concise characterization:

He likewise points out that narratives are perspective dependent—narratives “encode perspectives” (ibid.) in the sense that when one uses a narrative, they take up the perspective from which that narrative is articulated. An example Almagro uses is the skeptic of climate change, who claims that while there may be some climate change, there is too much alarmism because the global climate changes throughout history. On the first level, this is a sequence of events about climate change: how current climate conditions are related to past conditions. On the second level, there is the political narration of “alarmism”: there are certain actors who are overexaggerating the political significance of climate change.

Importantly, Almagro points out that narratives are often allusive or non-explicit, regarding what the main point is or who precisely the actors are. As in this example, neither specific people, statements, nor actions are called alarmist, making it highly ambiguous, while other narratives may cite precise people or actions, making it less allusive. Crucially, there is no distinction between narratives at the allusive end and fact-reporting at the explicit end; narratives exist along a spectrum of explicitness. It is easy to spot narratives when they are generalizing a series of events and actors. It is more difficult to spot them when narratives take on the guise of “just the facts”. For example, consider the skeptic who claims “climate change is overexaggerated”; is this a narrative, a political judgement, a verifiable assertion, or some other kind of discursive phenomenon? I contend here that this claim is a narrative, though in a highly abbreviated form that simply states a major point of the larger narrative. Indeed, this appears to be a primary way that narratives abound in political life, as slogans, chants, soundbites, and memes encapsulate the narrative into a convenient phrase. We can call such phenomena narrative judgements in the sense that these judgements are both components of broader narrative trends and not apprehensible without a comprehension of the narrative in which that judgement is a part.3 Both speaker and audience must already have some familiarity with the climate skeptic narrative to grasp the meaning of “overexaggerated”.

One indicator of a narrative judgement is that if an audience asks the speaker what they mean, what they are talking about, or why they believe that, the speaker will have to supply more narrative context, e.g., they will likely make statements similar to those in Almagro’s original climate skeptic example.4 In this sense, we could say that to apprehend the judgment that “climate change is overexaggerated”, one must comprehend the political narrative about climate change in which this judgment is a part. Alternatively, one may learn more about the general climate skeptic narrative by drawing upon the meaning of climate change being “overexaggerated”. In reciprocal fashion, one learns about whole narratives by becoming familiar with narrative judgements, and one better understands narrative judgements through familiarity with larger narrative structures. As I take narrative judgments to be parts of narratives, I will often simply refer to these judgements as narratives, in the sense that one does not need to explicitly state a whole narrative to use a narrative in discourse or in action. Many of the examples in what follows are narrative judgments for the sake of brevity and because they easily show the identities involved in narratives. Likewise, I will give a partial narrative reconstruction to contextualize the examples.

With this view of narratives, we can return our attention to the constitutive and discursive kinds of social construction which are especially relevant for understanding how narratives are involved in the construction of political identities. By looking more precisely at the role of narratives in constructing identities, we can see the pragmatic requirements of narrative use to obtain an identity and the networks of attributions and attitudes involved within narratives categorizing social worlds. To begin, we must recognize a key distinction in the two sides of narrative identity construction: the construction on the side of the narrator and the side of the narrated.

On the first side, we have the sense that the narrator, the subject using a narrative, constitutively constructs their own identity via narrative practices. The narrator engages in political discourse in ways that construct their own identity. As Almagro notes, part of having a partisan identity is sharing world-views with other partisans. This involves the pragmatic requirement that one espouses this world-view and uses the correct narratives aligned with the partisan ideology. To be a white nationalist, one must incorporate certain tropes of blood and soil, foreign invasion, “the great replacement”, racial supremacy, etc., into the ways they see, speak, and act within the world. Including these social world-views and narratives into one’s engagement with the world constitutes white nationalism. These are some of the social factors that are criteria for coming to possess this political identity, and criteria by which others will attribute the identity.

On the second side, we have the sense that discourses construct a subject of narration or a narrated identity. Discursive construction in this case takes the form of classifying and categorizing social identities within the narrative itself. In this way, certain identities are constructed by being narrated within the explicit content of narration; these are who the narrative is about. Consider Elon Musk’s 2024 tweet that “The ‘homeless’ aka ‘drug zombie’ industrial complex is very real”. This is an example of a narrative judgement, claiming the reality of a “homeless industrial complex”, which is part of a broader narrative concerning the increased presence of unhoused people and allies, especially in urban centers. Here we have a clear example of the construction of a social identity, “drug zombie”, being used in an overtly dehumanizing way to refer to unhoused people suffering mental illness and addiction. This example explicitly shows how identities are actively narrated to be constructed and contested in political discourse.

I highlight the distinction between narrated and narrator identities not because they are different kinds of identities, but because they show different ways that narration contributes to identity construction. The narrated and narrator identities often overlap. This can be seen in the case of white nationalists in the Northeastern U.S. using the slogan “keep New England white”, where white New Englanders are the central identity both in the narration and presumably of the narrator. However, there are also non-explicitly narrated identities in this slogan: those who are racialized as being non-white (by no means a clear nor consistent category), whose existence is thereby a threat to New England as white. This narrative likewise erases the Native Nations and Indigenous peoples of Dawnland, who continue to live in what is now known as New England. In this way, narratives involve many identities, which may include the identity(s) of the narrator, the identity(s) explicitly narrated, and the identity(s) implicitly or associatively connected to the narrative.

In other cases, discursive antagonism of a narrated identity is, in part, used to constitute the narrator identity. This can be seen in antagonistic narratives platformed by Trumpian political commentators, such as “the radical Left is ruining this country”, where the Leftist narrated identity is attached to hostile anti-American associations. This narrative functions to distinguish the narrator as against or opposed to the narrated group.

Other narratives may not explicitly involve a narrated identity, such as “climate change is a hoax”, although implicitly liberal and leftist partisans are associated with climate change concerned policies, while academics and scientists are in on the hoaxing. In this sense, we could say these political narratives have implicative or associative narrated identities—those identities implicated in the broader story-frameworks of a narrative.

I am highlighting here the ways narratives are not only tools of identifying oneself, but they are part of the structure through which identities are cast upon others. An obvious and dangerous current example of this is the identification of undocumented immigrants as “illegals”. Political narratives such as “illegals are committing violent crime across the nation’s cities” are engaged in shaping the very landscape of identification through perpetuating a derogatory narrated identity. The identity-label move from “illegal immigrants” to simply “illegals” functions to recast the identity of migrants in two ways: first, it guarantees their criminalization through what the Right conceives as an indubitable tautology: all illegals are criminals. Second, it strips them of residency, of having immigrated, which rhetorically delegitimizes them from any belonging on U.S. soil.

Here, the political alliance that notoriously beats their chest as the side with “facts and logic” masks a narrative of dehumanization in simple carceral logic to justify blanket incarceration and deportation for those deemed to be “undesirables”. Florida Governor Ron DeSantis casts people as “illegals” with great frequency in his narratives to normalize this identification. Throughout his tour of the incarceration-deportation camp dubbed “Alligator Alcatraz”, he proudly points to all the cages that will confine the thousands of “illegals” he looks forward to concentrating in the rushed tent camp built on a barren tarmac in South Florida on stolen Miccosukee ancestral lands.

“Illegals” in this sense is a discursive construction that is built up from criminality discourses and carceral apparatuses. While likely no one would self-attribute theirself as being an “illegal”, this identity has been brought into being by regimes bent on dehumanizing vulnerable peoples to further white national agendas. In many cases, we must face the reality of political identities whether oneself would attribute them or not. The reality of identity is not just in whether someone “really is” this or that way, it’s also about whether being attributed that identity could have you detained, deported, or worse.

While it is not always clear which of the two sides of narrative identity construction Almagro has in mind when analyzing this relationship between narratives and identities, I believe his book primarily addresses the first side, where the narrator constructs their own identity by engaging in political discourse. This is evidenced by the importance of the credence dimension of polarization, where the level of one’s credence given to a narrative indicates the degree of their partisan membership. Similarly, the partisan narrator “voices their attachments” or expresses attitudes regarding ideologies and identities by using narratives. In both analyses, the identities within the content of a narrative are not the essential element, but rather, it concerns how certain narratives are used with respect to the narrator’s own political identity and attitudes.

Almagro makes a significant contribution to understanding the role of narrations for partisan identities, though this is primarily with respect to the involvement of narrator identities. By this further clarification of the kinds of constructivism at play, we can illuminate both how partisans take up certain narratives and how narratives contribute to the construction of certain identities.

The causal, constitutive, and discursive kinds of social construction are especially significant for appreciating the importance of ideological or narrative critique. The very concepts, categories, identities, and attitudes employed in discursive practices can produce profound impacts on the very political processes in which they are embedded. In this sense, political narratives don’t simply speak about the world; they are causally constructed by political processes, and they are engaged in the very transformation of these political processes. This is one of the invaluable insights by Michel-Rolph Trouillot (1995), who shows how socially produced historical narratives hold great power in shaping political processes. Part of the infamous strength of the MAGA slogan is its swift construction of a historical narrative that calls for a future project: make America great again. When was it great? What was great about it? Who made it great? How will it be made great again? The simple slogan is silent in all of these respects. The silences that permeate such a narrative are not bugs; they’re features. It allows narrators and audiences to imaginatively fill in the silences with what made America and Americans great in the past, and how they’re going to make it happen again. This narrative is a device for voicing political attachments to American exceptionalism. The details are only so important; what really matters are the alignments of attitudes, identities, and ideologies.

It is crucial that we see in the MAGA slogan how the implicitly narrated national identity, “American”, is taken up by Right-wing partisans. It simultaneously speaks about the Unamerican, those responsible for unmaking America’s former greatness. In these narrative conditions, in this framework of discursive identity construction, the logicality of the phantasmic fixation on “illegals committing violent crime and making cities unsafe” has a clear direction: these Unamericans are responsible for the denigration of a once great nation, and they must go.

Ultimately, for many Right-wing partisans, a thin blue line is all that divides the American and the Unamerican: the law upholding, and the law breaking, which is the same as the Nation building, and the Nation ruining, the Good and Evil, the White and Black. This compounding of national, criminal, moral, and racial identities as narrated is crucial to understanding the shifting ways in which people themselves form their self-attributed identities and their other political attitudes.

At the same time, the identity “illegal” is an enclosure of a person into the moment of an original sin, a permanent transgression: arrival. The way someone has arrived at this place carries a cross-temporal significance for their very identity as defined both by the legal apparatuses of the state and by partisans who translate legal statuses into identifiers, whether “undocumented migrants” or “illegals”. Embroiled in the partisan narratives, policies, legal codes, and judicial proceedings, these political identities carry immense weight for the people whose arrival and extended presence have marked them as bearers of such identities. Barred from being an immigrant, cast as unamerican, and guilty of destroying a nation, they are being excluded and violently removed from the American project.

I have been emphasizing the role narratives play in not only constructing narrator identities, but also in constructing narrated identities. One purpose of this discussion specifically on narrated identities such as “illegals” is to showcase the socially constructed and non-internalized aspects of social identities. As I began, social identities are not just a matter of self-identification; they are intricately related to the social narratives used by others as much as they are related to the narratives we use. Having shown that, much like meanings, identities aren’t just in our heads, we can now address how this narrative constructivist view of identities works with the expressivist diagnosis.

3. Expressing Attitudes, Invoking Identities

There is a rising tendency in political philosophy to theorize the ways in which certain social practices express or signal a person’s political attitudes and identities. The expressive function of actions like poll responses (Ross and Levy 2023; Schaffner and Luks 2018), voting (Achen and Bartels 2016), social media posting (Marwick 2018), stating beliefs and commitments (Lynch and Tosta Forthcoming), and even reasoning (Kahan 2017) are taken to explain why people act in certain ways, i.e. they act because of the political expression of their actions.

Almagro makes a valuable contribution to this scholarship with his expressivist diagnosis of affective polarization. Particularly, Almagro argues that polarizing narratives are used for their expressive function: a partisan uses a narrative in order to “voice their attachments” to certain ideologies and identities. In this sense, participants of political discourse recognize ideological and identity associations with narratives, whereby using a narrative is an opportunity to express their attachments to ideologies and identities. For Almagro, these attachments are credences—how strongly one holds to the core beliefs of the ideology associated with a political identity. Moreover, he takes political judgements and narratives to “function as vehicles for expressing a political identity, encapsulating the speaker’s preferences and values” (Almagro 2025, 126).

This explanation of partisans’ expressive use of narratives runs together two forms of expression. The first is political attitude expression: the expression of people’s attitudes regarding political or social phenomena. The second is political identity expression: the expression of the person’s social identity through some mode of political engagement. With this distinction, we can appreciate the difference at the level of analysis and conceptualization between the expression of attitudes and the expression of identities. The principal issue here is whether social identities are the kinds of things that can be expressed, or whether an expressivist model is correct and insightful for understanding how people engage in activities that “show who they are”.

One might wonder whether the expressivist model of signaling, showing, or displaying is completely sufficient for theorizing social identities. As I will briefly argue in this section, the expressivist model is insufficient primarily because social identities are not internalized features of individuals. As we have been arguing, social identities are social constructs and thus are not internally possessed, therefore, they cannot be expressed or “pressed out” from within oneself.

The expressivist approaches generally presume that what we express are our internal psychological characteristics: our emotions, our attitudes, our inference patterns, our experiences. While expressivism is vital to understanding our communicative practices and social identities themselves, I suggest that social identities are not expressible because they are not reducible to internal features of mind. Instead, social identities might be more fruitfully understood as invoked through the expression of certain attitudes. To sketch this argument and shift towards an invocational model of participating with social identities, I will consider some expressivist views and outline some general features of expressivist models.

Expressivist theories generally regard a person’s psycho-physical characteristics as expressible through communicative practices. Mitchell Green (2011) offers a theory of expression wherein agents manifest their cognitive or emotional states through certain signaling activity. Expressivism is directly applied to understand the meaning of linguistic expressions by Luca Incurvati and Julian Schlöder (2023), where the meaning of ethical statements such as “lying is wrong” is the expression of a disapproving attitude about lying. A key insight here is that the meanings of our linguistic and signaling activities are the attitudes expressed. Taken together, our communicative practices pervasively facilitate the expression of our attitudes, and this is clearly the case when it comes to political discourse. Michael Lynch (2025) considers political discourse as performative when discourse is used mostly to express people’s commitments and emotions, and those expressions are intended to signal loyalty to a political identity, leader, or ideology. Here we see that Lynch approaches expressivism in politics as regarding our attitudes of commitment and loyalty to political features. This aids our analysis of the first kind of expressivism in politics noted earlier, i.e., concerning political attitude expression.

At this point, we must address whether Almagro shares this theoretical commitment to expressivism wherein expressibles are personally internalized features, like thoughts, disapproval, and loyalty. Almagro explains, “[a]ccording to expressivism, almost any sentence, uttered in the right context, can fulfill an evaluative or affective function, expressing the particular affective attitudes related to the speaker’s mindset” (Almagro 2025, 61-62). He later clarifies in greater detail how sentences have an expressive role with respect to the evaluative or affective functions of sentences. Crucially, Almagro makes the distinction between occurrent and dispositional affect. Occurrent emotions are “feelings”, mental sensations, or what is sometimes described as the phenomenal experience of emotion. Dispositional emotions are moods and inclinations that primarily reveal tendencies to act in certain ways, e.g., if I said I find something “disgusting”, I may not actually feel disgust, but I communicate my disposition to feel disgusted in the presence of said thing and to avoid or otherwise be averse to it.

These dispositional and occurrent forms of affect are both features of persons in the sense meant by personally internalized. This is clearly the case with occurrent feelings, where it is one’s own person who is having a particular feeling and expressing that feeling through speech or action. Dispositions are likewise internalized features in the sense that they are a person’s moods, tendencies, and psychophysical conditions. This is not to claim that dispositions are “merely” internal since they are highly involved in outward activity, for instance, someone having an angry disposition may have a propensity to yell. But this is a case where one’s disposition is expressively manifested by some typical behaviors; yelling is expressive of their anger. So too, a coach has a propensity to yell in their role as a coach, but this may not (though could) express the coach’s anger, and as I will argue, it is mistaken to claim that the coach expresses their coachness or expresses being a coach by yelling. Part of my goal is to motivate the significant difference between yelling out of anger and yelling in one’s role as a coach, which may also occur together, as for an angry coach. The former is behavior conditioned by an emotional disposition or mood, the latter is a behavior conditioned by a social role or identity. While our behaviors are often conditioned by both emotions and identities, the crucial difference to be addressed later concerns how they condition our activities.

Almagro explains that in political discussion, we do not primarily express occurrent emotions, but rather dispositional ones. This is not to say political discourse is engaged from a place of no feelings—clearly there are often highly intense feelings involved—rather, the affect expressed by political discourse is not always the same emotion felt during discussion. In practice, we express many more affective states, e.g., love, hate, admiration, fear, desire, in the span of a few sentences than we are usually capable of feeling in a strong way.5

Almagro’s point is that it is often not the case that we feel the affect we express. Instead, the expressed affects are dispositional attitudes, which can present as occurrent feeling or not. In particular, when it comes to expression in political discourse and narrative use, people often express their affective attachments. In the sense used here, attachments are the levels of confidence, or subjective probabilities, in the core beliefs of their political identities in a certain context (Almagro 2025, 121). A person’s greater credence in such core beliefs is more strongly attached to that ideology and political identity.

As discussed by Almagro, the attachments at the center of the affective expressive model are nonetheless personal attachments: an individual’s relations to certain views, practices, parties, and identities. At the same time, we have been arguing here that social identities are not constructed merely at the personal level; they are thoroughly social phenomena produced at the level of neighborhood and nation, not merely at the level of Patricia and Peter. Certainly, the personal and social are interdependent, as are attitudes and identities. It is my contention that the processes and dynamics involved in personal and social phenomena require different vocabularies that better accord with their particularities and that highlight their connections.

The argument I am motivating here is that the second kind of expressivism, political identity expression, is mistaken about the expressibility of social identities, because (i) expressivism deals with attitudes, internal personal characteristics, or psycho-physical features, and (ii) social identities are not merely attitudinal, personally internalized, nor psycho-physical (they are socially constituted, as argued in the previous section and in Almagro’s work). These conditions together show that social identities are incompatible with only an expressivist theory, and this requires additional theorization to address how political expressions relate to social identities. In short, the expressivist model is unable to accord with the socially constructed nature of social identities as not determined by any person’s constitution, but rather as emergent from the ways collectives participate within societies—that is, they are socially constituted.

Despite this incompatibility, Almagro makes the claim that using certain statements or reasons, and sustaining certain disagreements can be vehicles for expressing a political identity. When he states this, the treatment of identity slips back into a rather individualized view of identity, where what is “encapsulated” by the expression of a political identity is “the speaker’s preferences and values” (Almagro 2025, 126). This seems to make political identity more about a person’s political views than it is about a socially constructed phenomenon, constituted by practices, narratives, and groups.

Taking social identities to be expressible is a broader tendency in scholarship, as referenced at the outset of this section. The difficulty here, which often gets glossed over by expressive treatments of identity, is the balancing of the criteria of membership and the political identity of membership. While one must have certain political views to be a member of a political group, to have an identity, this does not reduce political identity to a personalized set of political views and affect. So we want a theory that respects politics as a highly expressive place of persons’ views and affect, while also appreciating that political identities are social phenomena in all three causally, constitutively, and discursively constructed senses.

We can overcome the insufficiencies of the expressivist model, while incorporating some of its key insights into political life, by advancing an invocational theory of social identity participation. One way that political identities are invoked is by participating in politically expressive activity. Rather than internalized features of individuals, social identities are participatory phenomena that require the performative activities of people in certain circumstances for their invocation. One does not express who they already are, but becomes their identity in part by “giving voice” to it. We can see here a partial explanation of why expressive activity appears so embedded in political discourse: people must invoke their political identities through expressing political attitudes in order to “have” that identity.

As Almagro has shown, partisans possess certain political world-views that they articulate through the narratives they use. They must make certain evaluations, cite (mis)information, attribute identities, show up to meetings, vote, and even commit acts of violence, all according to both the ideological and practical expectations of the partisan group at a particular time. This requires not only continual upkeep of an identity, but also a continual collective construction of the political identity. These practical endeavors are not merely expressions of an identity one already has. Each stage along the way is the processive invocation of a socially constructed identity. We might call this the participational realizability criterion of social identities: one must engage practically in social relationships to embody their identities. Identity expressivism misses the order of embodiment: they take identities to prefigure or cause an expressive act, rather than the act being the cause, figuration, or invocation of an identity. In other words, social identities are “achieved” by practices; they are not internal agential causes of practices. There are no identity-ghosts in the machine.

At the same time, expressing certain attitudes is crucial to participating in social relations and meeting this criterion. Thus, this approach acknowledges the importance of expressivism in social phenomena. The coordination of affect around practical and ideological expectations is a well-observed part of partisan politics. I believe this is one of the motivating insights behind the expressivist diagnosis. I agree with Almagro when he claims that vital parts of political identities are the world-views that embed both evaluative and descriptive features. The affective, cognitive, and practical dimensions are all key constitutive elements of our social identities. Thus, I agree with the portion of the expressivist diagnosis that claims partisan narratives express their affect, but I disagree with the claim that partisan narratives express their identity. Instead, we can say that one way partisans invoke their identity is by expressing their affect through narratives. In this sense, partisan invocation through narratives comprises an even more complex picture of political discourse. Partisans are not merely showing who they are, they are constructing who they are with divisive narratives about social worlds in which they are a part.

The invocational model articulates how social identities are constructed by partisans who use narratives in the two senses discussed previously: The narrator invokes one’s own partisan identity, while the narrated social identities are invoked upon others. This way, we move beyond the expressivist model of inside-out identity showing, and we can see the ways social identities are actively constructed through engagement with political narratives—even against the perceptions of those who have identities invoked upon them. The invocational model also helps us differently conceive of identity possession by appreciating different social metaphysics. Political identities are not simply shown from narrative use by partisans who have an identity; partisans actively become their identity by using these narratives. This highlights the constitutive dimension in the narrative construction of identities—part of being a partisan is engaging with narratives in certain ways. It is a partial requirement of having a partisan identity that the person’s discursive practices follow social norms according to the partisan group’s expectations and shared interpretations.

4. Conclusion

We have here a classic three-body problem. While modeling the reaction between any two bodies can be relatively straightforward, the introduction of a third body significantly increases the difficulty of modeling their relationships and trajectories. Once identities and emotions are put into relationship with narratives, the level of complexity is exacerbated, forcing us to reconsider all the elements.

By focusing on the role of narratives in partisan politics and polarization, Almagro has engaged in a much more difficult problematic that strikes at the multiple spheres through which our political worlds are lived. The additional dimensions he introduces create a welcomed complexity that accords with the on-the-ground political crises that evade straightforward explanations.

In this way, I approach Almagro’s book not as a closed work of polarization theory, but as an open invitation to further collaborate in theorizing and attending to how the dimensions of narrative, credence, and attitude vocalization are interwoven in partisan politics and processes of polarization.

To this end, I first explored in more detail how we can theorize some ways political identities are constructed through narratives. As social constructions, social identities are not merely a result of our self-attribution, but require adopting worldviews, engaging in narrative practices, and other forms of social participation. This involved attention to how narratives are used to construct the narrator’s own identity and construct identities within the narrated world-views. In other words, narratives both constitutively and discursively construct political identities.

A challenge regarding the connection between expressivism and identity constructivism was then raised. Where politics is acknowledged as deeply a place of expressive performance, it is easy to assume that our political identities, like our attitudes, are expressed by narratives. However, since the expressivist model treats internal states, attitudes, and psycho-physical features as expressible, and social identities are not internalized in this way, this strongly suggests that identities are not expressible. I have sketched an invocational model to conceptualize the ways political participation and attitude expression are key components of partisans constructing their identities, whereby partisans invoke their political identities through narratives.

REFERENCES

Achen, Christopher H., and Larry M. Bartels. 2016. Democracy for Realists: Why Elections Do Not Produce Responsive Government. Princeton: Princeton University Press.

Almagro, Manuel. 2025. The Rise of Polarization: Affects, Politics, and Philosophy. London: Routledge.

Ásta. 2018. Categories We Live By: The Construction of Sex, Gender, Race, and Other Social Categories. Oxford: Oxford University Press. https://doi.org/10.1093/oso/9780190256791.001.0001.

Brankovic, Jasmina. 2022. Narrative, Power and Polarisation: The Role of Influential Actors. June. https://doi.org/10.5281/ZENODO.10473718.

Butler, Judith. 1988. “Performative Acts and Gender Constitution: An Essay in Phenomenology and Feminist Theory.” Theatre Journal 40 (4): 519–531. https://doi.org/10.2307/3207893.

Epstein, Brian. 2015. The Ant Trap: Rebuilding the Foundations of the Social Sciences. Oxford: Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199381104.001.0001.

Green, Mitchell S. 2011. Self-Expression. Oxford: Clarendon Press.

Haslanger, Sally. 2012. Resisting Reality: Social Construction and Social Critique. Oxford: Oxford University Press.

Haslanger, Sally. 2022. “Failures of Methodological Individualism: The Materiality of Social Systems.” Journal of Social Philosophy 53 (4): 512–34. https://doi.org/10.1111/josp.12373.

Incurvati, Luca, and Julian J. Schlöder. 2023. Reasoning with Attitude: Foundations and Applications of Inferential Expressivism. Oxford: Oxford University Press.

Iyengar, Shanto, Yphtach Lelkes, Matthew Levendusky, Neil Malhotra, and Sean J. Westwood. 2019. “The Origins and Consequences of Affective Polarization in the United States.” Annual Review of Political Science 22 (1): 129–46. https://doi.org/10.1146/annurev-polisci-051117-073034.

Iyengar, Shanto, Gaurav Sood, and Yphtach Lelkes. 2012. “Affect, Not Ideology.” Public Opinion Quarterly 76 (3): 405–31. https://doi.org/10.1093/poq/nfs038.

Kahan, Dan M. 2017. “Misconceptions, Misinformation, and the Logic of Identity-Protective Cognition.” SSRN Electronic Journal, ahead of print. https://doi.org/10.2139/ssrn.2973067.

Levi, E., M. Bayerlein, G. Grimalda, and T. Reggiani. 2023. “Narratives on Migration and Political Polarization: How the Emphasis in Narratives Can Drive Us Apart.” MUNI ECON Working Papers, no. 7. https://doi.org/10.5817/WP_MUNI_ECON_2023-07.

Lynch, Michael P. 2025. On Truth in Politics: Why Democracy Demands It. Princeton: Princeton University Press.

Lynch, Michael P., and Jason Tosta. Forthcoming. “From Identity to Belief and Back Again: Perspectives from Philosophy.” In The Oxford Handbook of the Cognitive Science of Belief, edited by Neil Van Leeuwen and Tania Lombrozo. Oxford: Oxford University Press.

Marwick, Alice. 2018. “Why Do People Share Fake News? A Sociotechnical Model of Media Effects.” Georgetown Law Technology Review 2 (2): 474–512.

McLaughlin, Bryan, John A. Velez, and Joshua A. Dunn. 2019. “The Political World Within: How Citizens Process and Experience Political Narratives.” Annals of the International Communication Association 43 (2): 156–72. https://doi.org/10.1080/23808985.2019.1597635.

Ross, Robert M., and Neil Levy. 2023. “Expressive Responding in Support of Donald Trump: An Extended Replication of Schaffner and Luks (2018).” Collabra: Psychology 9 (1): 68054. https://doi.org/10.1525/collabra.68054.

Schaffner, Brian F., and Samantha Luks. 2018. “Misinformation or Expressive Responding?” Public Opinion Quarterly 82 (1): 135–47. https://doi.org/10.1093/poq/nfx042.

Sveinsdóttir, Ásta. 2015. “Social Construction.” Philosophy Compass 10 (12): 884–92. https://doi.org/10.1111/phc3.12265.

Trouillot, Michel-Rolph. 1995. Silencing the Past: Power and the Production of History. Boston: Beacon Press.

Warner, Benjamin P. 2019. “Explaining Political Polarization in Environmental Governance Using Narrative Analysis.” Ecology and Society 24 (3): art4. https://doi.org/10.5751/ES-10999-240304.