1 Introduction

In 2022, for the second time in a row, Marine Le Pen and Emmanuel Macron were the two options in the second round of French Presidential Elections. Many left-wing voters faced a hard choice between voting for Macron — a candidate they disliked for his economic agenda, voting for Le Pen — a candidate they disliked for her nationalist policies and anti-immigrant rhetoric, and abstaining as a way to demonstrate dissatisfaction with the entire political system. Thierry was one of such voters. In 2016, he had voted for Macron to block the extreme right. In 2022, he abstained. For him, this time was like voting between ‘plague and cholera’ (Sauvage 2022). In this paper, I shall give Thierry some reasons to think of his vote as a way to respond to the deterioration of political communication in contemporary democracies.

An ever-expanding body of social scientific research demonstrates that in several liberal democracies across the world, would-be representatives use derogatory expressions, which may not reach the threshold of legal prohibition, to unify and mobilise voters (e.g., Valentino et al. 2018). In such contexts, philosophers have so far focused on the key issue of defining specific linguistic mechanisms that can explain the success of explicit racist appeals (Saul 2019).Footnote 1 Yet, the success of explicit derogatory expressions in the political mainstream also suggests that standard forms of counterspeech, such as persuasion and direct linguistic interventions, may be blunt instruments when it comes to challenging public appeals made to the media and the general public. Since election results can define the public political rhetoric for years, this article investigates the individual act of voting against candidates who have used explicit derogatory expressions in democratic campaigning as one of the possible forms of counterspeech. Providing a compelling case for adding voting in large-scale democratic elections to the range of counterspeech actions requires that we unpack the experience of voting at least from the perspective of ordinary citizens who are not members of targeted minorities, despise the spread of offensive speech in the democratic public sphere, and have engaged in counterspeech or wish to engage in counterspeech (hereby, counterspeakers).Footnote 2

Counterspeakers may face particularly hard choices during elections in which some candidates use expressions targeting vulnerable groups and other candidates are perceived as incompatible with their ideological affiliation. Against this backdrop, I claim that in elections framed as a vote for or against candidates who have used expressions explicitly targeting vulnerable groups, individual electoral choices can contribute to reducing the relevance of disrespectful forms of expression in the democratic public sphere. Since the act of voting is one of the first counterspeech actions that must be taken so that other, and perhaps more controversial, actions can stand criticism on moral grounds, I argue that counterspeakers can expect other counterspeakers, who have engaged in visible forms of counterspeech, to vote against would-be representatives who have used derogatory forms of expression targeting vulnerable groups.

To support my claims, I begin by making clear the theoretical framework of this paper. In Section 3, I place the act of voting against representatives who use offensive speech in the context of ongoing debates about counterspeech. I recall the distinction between counterspeech as speaking with and counterspeech as distancing. In Section 4, admitting that it is extremely unlikely that any one vote would break a tie, I argue that the individual act of voting, understood as a contribution to a collective communicative act, has a non-transferable capacity to influence how the audience will interpret the election result. Building on the idea that the individual act of voting can have some impact, Section 5 focuses more specifically on how an understanding of the individual act of voting as a form of counterspeech can affect our perception of what counterspeakers can be expected to do. Section 6 concludes. This article contributes to two burgeoning literatures, the ethics of voting and the ethics of counterspeech. I offer the first generalizable framework, and a series of conditions, to assess the effectiveness of a single vote in large-scale democratic elections from a speech-act perspective. Then, in a moment in which social scientists study elections in stable liberal democracies for their long-term impact on the rules of democratic communication, for the first time in the literature, my paper demonstrates that individual electoral behaviors should be objects of concern for scholars of counterspeech, who focus mostly on persuasion and direct responses to bad speech broadly understood (Cepollaro et al. 2022).

Two preliminary remarks are on point. In what follows, I assume that ‘hateful representative claims’ are claims to represent the preferences of an audience through speech acts that explicitly advocate the idea that some groups lack the capacity to become fully cooperating members of society. I also assume that a ‘hateful representative’, for the sake of simplicity, is a member of a mainstream party, an incumbent politician, or a member of the executive who makes hateful representative claims in campaigning for democratic elections.

2 The Democratic Vote as a Speech Act

There seems to be little doubt that voting is an expressive behaviour (Wodak 2019), but there are several ways to capture the expressive capacity of voting. For instance, according to the expressive theory of voting, voting is an action ‘that is undertaken for its own sake rather than to bring about particular consequences’ (Brennan and Lomasky 1993: 25). In this article, I leave aside that theoretical tradition. My aim is not to advocate or question the idea that in voting, we express our feelings. I zoom in on the relationship between the individual act of voting and the collective outcome that can contribute to guiding the political elite in reaching an understanding of what the public deems acceptable. The election result, which includes both the number of votes and the number of abstainers, can be seen as a collectivized speech act (Ludwig 2020) —a speech act that originates in multiple persons— that, depending on its force, can compel or allow to say or do certain things.

This article adopts a social normativist speech-act-theoretical approach (Kukla and Lance 20092013; Kukla 2014) to study the issue of voting as a form of counterspeech. This approach considers speech acts as normatively significant moves that can alter normative status and reasons for action. On this view, the performative force of a speech act is partly constituted by direct addressees and other relevant parties who instigate, facilitate, and take up the speech act as normatively significant (Kukla and Lance 2009).

Social normativism urges us to rethink the centrality of speakers’ intentions for the successful uptake of a speech act (Kukla 2014). Speakers — individuals and collectives — are not full in control of their speech acts’ performative force (Kukla 2014). By taking into account considerations about the conversational context, such as conventional frameworks of interpretation, recognizable authority relationships between speakers and hearers, tones, punctuation, and gestures, the audience interprets whether some words have, for instance, the performative force of an order or request (Kukla and Lance 2013; Kukla 2014). In some cases, the conversational context is so unambiguous that the audience must interpret the performative force of a speech act in a certain way. In most cases, uptake depends on the audience’s interpretation of the governing context (Kukla 2014). Even if a degree of uncertainty seems to be unavoidable, when some frameworks of interpretation are particularly salient, when procedural constraints make listeners recognize speakers’ authority, and when indicators of performative force (such as, tone, punctuations, and gestures) are explicit, members of the audience have less room for interpreting based on their own judgements.

It is against this backdrop that the interpretation of the election result as a speech act having the performative force of an order depends not so much on voters’ intentions as on the context shaping the relationship between would-be representatives and the electorate. Specifically, the purpose of a democratic election compels would-be representative and other relevant members of the audience to recognize the authority of voters. Moreover, there are elections where the salience of some frameworks put further contextual constraints on how the democratic vote can be taken up in practice. Yet, for an election result to have a determined impact on the social space, the audience should be able to grasp its indicators of force. In other words, depending on the degree of intelligibility of its markers of force, the consequence of a vote may be that others are “compelled” or “allowed” or “not allowed” to do certain acts or say certain things (Austin 1962: 55).

In this view, the fact that individual voters have different reasons to vote for a certain candidate does not entail that the election result will necessarily be interpreted as an ambiguous collectivized speech act. This is so because individual votes do not contribute to what a democratic vote can say, but rather, as we will see below, to the intelligibility of its indicators of performative force.

We know that the way elections have been framed impacts on the kind of interpretative judgements members of the audience can make (Shamir, Shamir and Lavi 2015). We know that for each election, there is a limited number of plausible frames that can affect outcome interpretations (Shamir and Shamir 2008). We also know that the media selectively attend to some issues and in this way affect the criteria by which the public interprets the elections results (Price and Tewksbury 1997). Therefore, for an electoral outcome to be interpreted as a way to direct the discursive choices of hateful representatives, it is plausible to assume that the relevant election has been framed as a vote against (or for) hateful representatives (hereby, VAHR). An election framed as a VAHR is an election (a) affected by a media landscape that accentuates conflict and entertainment in political news reporting, (b) in which at least one of the candidates has made hateful representative claims in campaigning, (c) in which the use of hateful representative claims is salient, and (d) in which ‘a vote against hateful representatives’ is among the few plausible interpretations of what voters are saying with their votes.Footnote 3 In both 2017 and 2022, the second round of French Presidential Elections exemplified elections framed as VAHR. The prospect of a Marine Le Pen’s victory over Emmanuel Macron was framed as a ‘legitimation of racist speech’ (Alfon 2022). Donald Trump’s 2016 results were also framed as a victory of his hateful rhetoric (Stafford 2016). In 2019 Indian general election, writers, intellectuals, and political leaders appeal citizens to vote against hate politics and intimidation (The Hindu 2019). In Brazil, the electoral success of Bolsonaro is seen as an indicator that ‘anti-black hatred remains an effective symbolic and practical political tool’ (Alves and Vargas 2018).

In elections framed as VAHR, by the expression ‘an act of voting against hateful representatives’ I mean the act of casting a vote that contributes to make audience members interpret the outcome as signalling that the use of derogatory forms of expression is not allowed in the democratic public sphere. I do not argue that in elections framed as VAHR, the result will be interpreted solely as issuing an authoritative instruction to respect the informal rules of a decent democratic public rhetoric. My idea is that in elections framed as VAHR, the audience (would-be representatives, but also activists, commentators, experts, scholars, and journalists) is in the position to interpret how the same speech act applies onto many, yet limited, salient frameworks of interpretations. Within the limits of the general perception of what the election is about, individual electoral choices can contribute to limiting or expanding audience’s room for deciding whether the electorate is expressing an order or a request relative to the dominant interpretative frameworks.

Even if I cannot list all possible examples of an individual vote against hateful representatives, my argument aims to be valid at least for systems in which the candidate who receives the most votes wins outright, two-round runoff voting systems, and multiparty systems with proportional representation. In systems in which the candidate who receives the most votes wins outright, a vote counts as a vote against hateful representatives, for instance, (I) when there is only one candidate who stands on a nonhateful platform and (II) when there is a tough choice (Guerrero 2010; Wodak 2019) between A, a candidate who uses offensive speech, and B, another candidate who does protest the public rhetoric of A. In two-round runoff voting, which may require citizens to make difficult trade-offs, the choice may be between abstaining, submitting a blank vote, and choosing the lesser of the two evils. In that context, a vote counts as a vote against hateful representatives when, in the second round, the voter’s choice is B, a candidate who does not protest the offensive rhetoric of A but does advocate policies that might be in the interest of targeted minorities.

My argument aims to be valid also in multiparty systems with proportional representation where elections have been framed as VAHR. In such contexts, and on the assumptions that there is a nationwide electoral threshold and that parties can form post-electoral coalitions, different choices can count as a vote against hateful representatives. In some cases, the individual vote should be for the most moderate representative who does not use offensive speech and makes explicit that there will not be any coalition between fringe racist parties and other moderate parties whose candidates have made both explicit and covert hateful representative claims. In other cases, the vote can go for the party in coalition to express a strong signal against hateful representative claims within a certain ideological spectrum.

Democratic elections can shape the quality of political communication for years. Nevertheless, the individual act of voting is neither necessary nor sufficient to achieve the goal of reducing the presence of explicitly derogatory expressions in the democratic public sphere. For this reason, it is important to study the normative connection between the act of voting against hateful representatives and other forms of counterspeech. To this, I turn in the next sessions.

3 The Place of Voting in the Ethics of Counterspeech

The ethics of counterspeech attempts to answer questions such as these: Who is to undertake counterspeech? What are citizens required to do? And why is it morally justified to demand that citizens do it?Footnote 4 One may resist the very idea of finding a place for voting in an ethics of counterspeech. According to some legal scholars, it is a mistake to treat voting as a communicative act that can appropriately be classified as speech and, therefore, as counterspeech. There is some ground for this objection when we understand the term ‘speech’ solely as a term of art which encompasses only those communicative acts that a plausible argument for free speech should protect.Footnote 5 However, political philosophers and practitioners have already advocated a more expansive conceptualisation of the set of direct or indirect, individual, or collective discursive responses seeking to limit the presence of sanctionable and non-sanctionable offensive or dangerous communicative acts in the public sphere (Lai 2020, Fumagalli 2021). On this understanding, counterspeech can also encompass some conducts with expressive content, such as political vandalism (Lai 2020) and, I suggest, voting. Here, I therefore proceed with the hypothetical claim that if voting is an expressive practice that can count as a speech act, then it can also be studied as a form of counterspeech.

Fumagalli (2021) argues that counterspeech of ordinary citizens can be grouped in two clusters of actions: speaking with intolerant members of society and distancing from hateful representative claims. Counterspeech as speaking with includes forms of continuous discursive engagement with citizens who support hateful viewpoints (Badano and Nuti 2018; Lepoutre 2019). Counterspeech as distancing includes individual and collective attempts at making speech acts misfire, resisting what speakers presuppose, questioning speakers’ authority, and challenging conversational moves (Langton 2018; Maitra 2012). This article investigates the possibility and consequences of considering the act of voting against hateful representatives as one of the actions to be included in counterspeech as distancing.

In studying counterspeech as distancing, the still-nascent literature on counterspeech has either zoomed in on one-on-one forms of discursive distancing aiming at altering the conditions and effects of offensive speech acts (Langton 2018; Maitra 2012) or assessed the justification of non-ordinary and quite vocal forms of collective political participation (Lai 2020; Lim 2020). Specifically, Billingham and Parr (2020) argue that online public shaming might not only make speakers feel ashamed but also uphold valuable social norms, thereby showing solidarity with those who are wronged. Getting up and leaving, Not in My Name campaigns, and boycotts are also popular ways to hinder or disable hateful representative claims (Fumagalli 2021). When tainted symbols are state sponsored, political vandalism, which ranges from putting stickers on statues, to expressing countermessages, to destroying or removing statues, is also a kind of counterspeech as distancing (Lai 2020; Lim 2020). Lai argues that since state-sponsored tainted monuments speak in our name, political vandalism can make it clear that some people disagree with the message (Lai 2020).

To be sure, some types of distancing, such as vandalism and Not in My Name campaigns, aim at impacting primarily the content (what speakers say) of hateful representative claims. Other types of distancing, such as public shaming and voting, aim at challenging the authority of speakers who make hateful representative claims. In a time in which offensive speech permeates the mainstream of democratic campaigning, I contend, it is also important to study ordinary forms of political participation, such as the individual act of voting, as possible ways for counterspeakers to exercise their expressive capacities, distance themselves from hateful representatives, and hinder representative claims.

4 The Effectiveness of the Act of Voting Against Hateful Representatives

A good ethics of counterspeech should advocate actions and behaviours that are likely to have some impact (Howard 2021: 934). If individual voting is sure to change nothing, counterspeakers ‘can reasonably ask why they should bear the costs of counterspeech if there is no anticipated moral pay-off’ (Howard 2021: 934).

The effectiveness of the individual act of voting is a particularly pressing issue, as a rich literature demonstrates that even if a remarkably high number of citizens vote, it’s extremely unlikely that any one vote would break a tie (Downs 1957: 244–46). Against this backdrop, and assuming that voting is not compulsory, counterspeakers may wonder whether casting a vote against hateful representatives is effective enough a form of counterspeech to be worth the effort.

In this section, I claim that a new conception of effectiveness—prospective effectiveness—can help us to rebut scepticism about the effectiveness of an individual vote against hateful representatives. I use the word ‘prospective’ to draw attention to the idea that this conception aims at describing the degree of effectiveness that an individual action can have, given its most fundamental properties. This conception is built on the social normativist idea that acts of distancing can be speech acts enacting new (or reaffirming old) conversational norms. For instance, Lai (2020) demonstrates that vandalising tainted monuments constitutes an enduring way to establish what ought not to be said. Against this backdrop, an individual act, I contend, is prospectively effective when it has the capacity to alter how a collective action is interpreted and such a capacity cannot be passed from one agent to another.

Even accepting two very plausible assumptions—that it takes much more work to rebut offensive speech than to engage in it (Lim 2020: 206), and that it is almost impossible for one person alone to change the communicative environment in which hateful representative claims proliferate—political philosophers can identify the characteristics of different acts of distancing and study whether such acts have in themselves the capacity to translate individual support into a non-transferable contribution to limiting audience’s room for interpretation.

Speech-act theory includes ‘a vote for’ in his list of exercitives (Austin 1962: 155). An exercitive is a type of speech act through which a single speaker or a group of likeminded individuals compel, allow, or do not allow certain acts (Austin 1962: 119, 154; Kukla 2014). In approaching election results as exercitives, one should keep in mind that interpreting the performative force does not necessarily entail recognising single intentions behind the speech act being performed. In the case of a vote for a certain candidate, voters and abstainers perform an action that contributes to a single output, where the output expresses something with a certain force because of all the actions taken by single agents in their individual capacities (Ludwig 2020). For the election result to secure uptake as an exercitive speech acts, the audience should be able to understand the indicators of force unquestionably.

On the assumption that both voters and hateful representatives accept the electoral procedure, the number of votes does not influence the set of available interpretative frameworks. The number of votes, as the main indicator of force, contributes to reducing audience members’ capacity to alter the performative force of the election result. If only a small number of citizens vote against a hateful representative, the resulting vote may be too elusive to secure uptake. In this case, a vote may be too ambiguous to even signal irritation with the proliferation of hateful representative claims. In contrast, in an election framed as a VAHR, when many citizens cast a vote that can count as a vote against hateful representatives, it is more likely that the resulting vote will be perceived as a collectivized speech act signalling that certain actions are prohibited. In other words, the clarity of the indicator of force of an election result links with the number of citizens contributing to the collectivized speech act. In this view, the electoral outcome concerns, in Guerrero’s (2010: 274) words, ‘not just who wins and loses, but how the winner wins and how the losers lose’.

The last point is particularly relevant for this analysis. In interpersonal communication, exclamation marks, performative verbs, and the rise and fall of a voice are indicators of performative force. All such markers are absent in the case of secret voting. Yet in a procedure in which hateful representatives must take seriously the number of votes they collect, the sheer number of votes for a candidate, and, connected with this, the relative gap between candidates, can be understood in the same way as other markers of force. Even in secret ballots votes are communicated to someone who is in the capacity to count them (Ludwig 2020: 49). Numbers make a vote understandable as merely expressing a feeling, as an ambiguous signal, as an exhortation, as an imperative, or as enacting new conversational norms.

If one accepts that the election result is a collectivized exercitive speech act, new norms can come into being when many voters contribute to increasing the clarity of the indicators of performative force. The larger the number of votes against hateful representatives, the more likely the election result will receive uptake as an exercitive speech act. If a small number of voters contribute to enacting a collectivized exercitive speech act, the audience will be more likely to interpret the results according to their own judgment. This point suggests that each vote has a non-transferable capacity to increase the clarity of the indicators of force of a collectivized speech act whose performative force connects with the number of voters, and relative distribution of votes.

The capacity to increase the clarity of the indicators of force is non-transferable because the set of available votes is finite, and voters can vote only once. Footnote 6 Therefore, abstaining and voting for another candidate imply that the collectivized speech act has less force than it would otherwise have. When votes are secret and tallied collectively, and certain frames motivate interpretative judgements about the electoral outcomes, voters taken individually cannot speak loud enough to challenge the interpretation of votes (Wodak 2019: 380). Moreover, in the electoral context, voters taken individually cannot substitute their potential contribution to the aggregate speech act with another form of political participation. Even if they engage in other meaningful forms of counterspeech, their choice not to vote implies that the indicator of force of a democratic vote are less clear than it would have had, had voters in their individual capacities done otherwise.

Even if the observation that most votes are irrelevant to the outcome remains valid, this section has demonstrated that single votes influence the indicators of force of a collectivized speech act. All voters, therefore, have a distinctive and individual capacity to affect how the audience will receive the outcome of a democratic election. Such a capacity, I have said, is small but non-transferable to other citizens. My argument so far has studied the relationship between all possible voters, their decision to cast a vote against hateful representatives, and the democratic outcome understood as a collectivized speech act that can enact some conversational norms. In what follows, I address the case of counterspeakers from a normative point of view. If the individual act of voting can influence in a non-transferable way a speech act that can enact some conversational norms, it is important to investigate the relationship between the individual act of voting and other forms of counterspeech.

5 Forms of Counterspeech, Counterspeakers, and Voting Against Hateful Representatives

In this section, I first compare the morality of different forms of counterspeech as distancing. Building on this comparative analysis, I then investigate how understanding the act of voting against hateful representatives as a part of a broad counterspeech strategy can influence what counterspeakers should expect one another to do in elections framed as VAHR. Counterspeakers may be led out of frustration to prioritise nonelectoral acts of distancing—such as donating to charity, public demonstrations, rioting, vandalism, Not in My Name campaigns, and other forms of online activism. It is, however, misleading to frame counterspeech as necessarily involving a choice between different acts of distancing. Joining a collective act of distancing does not prevent us from engaging in other acts of distancing.

5.1 Voting Against Hateful Representatives and Other Forms of Counterspeech as Distancing

Most of the work on counterspeech as distancing revolves around the justifiability of different individual and collective actions (Billingham and Parr 2020; Lai 2020; Lim 2020). In what follows, I compare the act of voting against hateful representatives with other forms of counterspeech as distancing in light of three criteria—representativeness, non-evasiveness, and necessity. This short list builds on the results of multiple theoretical approaches to counterspeech (Howard 2021; Lai 2020; Lim 2020).Footnote 7 First, when visible acts of distancing express an attempt to speak on behalf of the community of counterspeakers as a whole or in support of a targeted group, justifiable forms of counterspeech as distancing should represent the feeling of a larger group accurately. A misrepresenting act of counterspeech gives a false or misleading account of what the group of counterspeakers really stands for. Second, a justifiable form of counterspeech should be communicative and enable agents to articulate reasons for actions. A non-evasive act of counterspeech puts the agent in a position ‘to articulate her commitments and reasons for her actions to others within her community’ (Lim 2020: 209). Third, justifiable forms of counterspeech should avoid using more force than necessary to avert an evil. Necessity demands that one chooses the least morally controversial of all available alternative means (Steinhoff 2019). Taken together, representativeness, non-evasiveness, and necessity shed light on the specific criteria counterspeakers should consider when they deliberate on the moral justification of their contribution to the collective endeavour of countering offensive speech.

Representativeness. A single act of voting against hateful representatives exhibits one’s commitment to certain political ideals. Unlike other acts of distancing, the act of voting does not express a public claim to speak on behalf of the group of counterspeakers. Through acts of political vandalism, a single counterspeaker can claim to speak on behalf of a whole group (Lai 2020; Lim 2020). Yet the representativeness of acts of political vandalism depends on whether the acts follow deliberation among counterspeakers and on whether the propositional content of the acts is in line with the sort of message the community of counterspeakers is fighting against (Lim 2020: 209–10). Such an alignment cannot be taken for granted. As Marschall (2017: 210) writes with regard to the Rhodes Must Fall campaign in South Africa, it can be difficult ‘to determine whether observable traces of defacement signify a message of socio-political protests or simply an act of malicious damage’. In this sense, the act of voting against hateful representatives is closer to boycotting, online public shaming, and Not in My Name campaigns than to political vandalism. If considered in isolation, a single act of voting cannot be said to be misrepresenting. If considered collectively, voting, boycotting, and Not in My Name campaigns can contribute to sending a message loud and clear. In elections framed as VAHR, such a message can give an idea of what the community of counterspeakers considers as an appropriate way to make representative claims.

Non-evasiveness. Boycotting and Not in My Name campaigns are engaged in by a subgroup of citizens who reject a claim made by hateful representatives. They are communicative in character and put individual citizens in a position to explain to other citizens their reasons for actions. In this respect, voting, which, at least in large-scale democratic elections, is secret, share some features with online public shaming. However, it would be misleading to conclude that online public shaming and secret voting are evasive to the same extent. On the one hand, because some social media platforms and email groups ensure relative anonymity, the online dimension of public shaming can shield shamers from the demand that they articulate reasons for their actions (Billingham and Parr 2020: 380). For instance, one may be unable to know that her friends and family members join online public shaming campaigns. On the other hand, the secrecy of voting entitles voters not to reveal the content of their votes to anyone. This entitlement, however, does not exclude the fact that other citizens can know whether I go to the polling station. The secrecy of voting, in other words, does not exempt voters from pressures to justify their choices. It influences the way voters exchange their reasons for action. Because of the secrecy of voting, voters must take what other voters say about their votes at face value. In exchanging reasons for actions, except for those cases in which an agent publicly supports a candidate or is vocal in advocating abstention, the secrecy of voting forces speakers and hearers to accept that their interlocutors are telling the truth about the choice they have made. The act of voting against hateful representative, therefore, is more evasive than other forms of counterspeech, such as boycotting and Not in My Name campaigns, but it may require counterspeakers to explain to other voters their reasons for action.

Necessity. Lai (2020: 609) and Lim (2020: 186) agree that when no other responses to tainted commemorations are possible, vandalism is morally permissible. Despite being nonviolent and peaceful, online public shaming can also impose significant burdens on individuals who are exposed to organised online campaigns (Billingham and Parr 2020: 998–1000). For this reason, online public shaming meets the necessity criterion only when less damaging means have proven ineffective. Moreover, what can meet the necessity criterion in the here and now seems also to relate to the community in which one operates. If the community is prone to accept strong forms of social control, online public shaming can be very intimidating and become a form of vigilantism with offline elements. Like boycotting and Not in My Name campaigns, the act of voting against hateful representatives is an example of disputing representative claims in a way that helps make hateful representatives accountable without inflicting damage on property and people. All three means are nonviolent and peaceful demonstrations through which citizens of liberal and democratic regimes exercise their fundamental democratic rights.

If this comparative reading is correct, single acts of voting against hateful representatives, as opposed to acts of online public shaming and political vandalism, cannot be misrepresenting, as they are private ways to contribute to an outcome. Even if in an election framed as a VAHR, the individual act of voting against hateful representatives, unlike boycotting and Not in My Name campaigns, is protected by the right to secrecy, voters, unlike online shamers, are not fully shielded from the demand that they articulate reasons for an electoral choice. Moreover, the act of voting against hateful representatives, like boycotting and Not in My Name campaigns, is a nonviolent way to express a fundamental democratic right. In other words, the act of voting against hateful representatives might be less irreprehensible than highly communicative actions, such as boycotting and Not in My Name campaigns, but it is not harmful and, therefore, should be considered as one of the first acts of distancing that must be taken so that other actions, such as public shaming and vandalism, can meet the necessity requirement.

As I shall say in what follows, the inclusion of a morally uncontroversial action in the set of possible justified responses to the deterioration of political communication can influence mechanisms of mutual accountability between citizens who recognize themselves (and want to be recognized) as counterspeakers. Even if the action under scrutiny is generally considered as morally uncontroversial, or at least one of the least controversial forms of citizens’ engagement in improving public political communication, one should keep in mind that the addition of an action to the set of available controversial and noncontroversial means to an end, is not a neutral intellectual operation.Footnote 8 It seems indeed problematic to undertake more controversial counterspeech actions, such as vandalising and online public shaming, without having cast a vote against hateful representatives in elections framed as VAHR. Seen from this perspective, if some counterspeakers think that at a certain time, it is appropriate to engage in more controversial forms of counterspeech, then other counterspeakers can now expect them to vote against hateful representatives in elections framed as VAHR.

The present argument should not be understood in strictly sequential terms. Even if the act of voting against hateful representatives is one of the least disruptive forms of counterspeech, it does not have to be the first act of counterspeech. The idea that voting is one of the least disruptive forms of counterspeech can also have a retrospective normative force. When an election is framed as a VAHR, counterspeakers, being conscious of what happened previously, can hold one another to certain standards.Footnote 9 To this, I turn in the remaining of this section.

5.2 Counterspeakers and Elections Framed as VAHR

It is important not only to evaluate acts of distancing comparatively, but also to study what counterspeakers can be expected to do in elections framed as VAHR, given that they have already engaged in counterspeech and given that they may engage in counterspeech in the future.

Counterspeakers may back a candidate for different reasons, but, just alike other voters, they are not fully in control of how the audience will interpret the election result. In elections framed as VAHR, individual votes, even if they do not originate in a commitment to oppose hateful representative claims, can contribute to an outcome that the audience will interpret as a vote affirming or licensing some rhetorical norms. Therefore, my argument suggests that in elections framed as VAHR, counterspeakers should consider prima facie reasons to prioritise their interest in undermining offensive communicative acts.

I do not think that even from the perspective of counterspeakers, counterspeech-based reasons should be the only reasons that matter in choosing a candidate. While each counterspeaker may share a common goal with other counterspeakers, each may also have conflicting commitments that speak against the otherwise preferred candidate. Yet, in balancing their commitments, counterspeakers should consider that while one’s distinctive contribution blends into the crowd, the performance of other visible forms of counterspeech is common knowledge and can assure others about—and justify mutual expectations about—what they and other counterspeakers will do. Provided that the goal of containing hateful representative claims is a legitimate moral goal, the existence of such an expectation can be a strong reason for prioritising a course of action over its alternatives. Therefore, in acting otherwise without providing a sufficient countervailing consideration, counterspeakers violate a non-neutral expectation and open themselves to adverse moral judgement.

Adverse moral judgements admit degrees, and countervailing reasons determine to what degree counterspeakers are subject to such judgements. In holding other counterspeakers accountable, one should at least consider the probability that hateful candidates will win a majority of the votes as well as facing other personal costs, such as limited voting hours, polling-place closures, and lack of language access. Moreover, counterspeakers may have a plurality of other commitments that generate problems of competing motivations. For instance, counterspeakers with multiple identities may be held accountable for an ever-expanding constellation of duties, such as the duty to pursue justice, the duty to promote a shared interest, the duty to advance their conception of the good, and the duty to act consistently with partisan or religious affiliations (Fumagalli et al. 2021). The moral costs of choosing between competing motivations suggest that there should be some latitude in assessing the normative demand of the expectation to vote against hateful representatives in elections framed as VAHR.

Counterspeakers, by weighing countervailing normative demands against each other, should be able to decide how best to fulfil all their duties (or most of them) and, therefore, what is for them the most appropriate means of contributing to the goal of containing the proliferation of hateful representative claims. On this view, the normative demand to vote against hateful representatives does not always require the same pattern of action; it just requires that counterspeakers vote in a way that, under those specific electoral circumstances, others can take as a vote against a hateful representative.Footnote 10 Even supposing that counterspeakers have ample latitude to decide among various courses of action in a particular situation, mutual expectations still contribute to holding one another to certain standards. When elections are framed as VAHR, such expectations entitle counterspeakers to ask other counterspeakers to justify their electoral choices.

One last remark is in order. Even if counterspeakers make the right electoral choice, there might be other contextual variables that obstruct successful communication processes. In such cases, despite the large number of votes, the result may not receive uptake as an exercitive speech act. When we lack the procedural conditions for preventing members of the audience to interpret the election results according to their own judgments, counterspeakers can fulfil their pro tanto duty also by performing forms of silent agency (abstention, blank ballots, null ballots, spoiled ballots, and none of the above votes). Yet, in opting for forms of silent agency, it is important to consider that there might be a significant shift in the recognition of authority relations. Specifically, even if counterspeakers explain that their forms of silent agency are intended to sanction hateful representatives for their representative claims, and even if a large number of counterspeakers join the collective action project, it is not granted that the audience will act as if counterspeakers are entitled to enforce conversational norms in that way.

6 Conclusion

I have argued that in an election framed as a VAHR, the election result, provided that its indicator of force is sufficiently explicit, can be interpreted as an exercitive speech act establishing what ought to be said in that context. In view of the fact that each voter can contribute in a non-transferable way to the clarity of such an indicator of force, counterspeakers may expect a justification from other counterspeakers who abstained or voted for a less-than-the-best available candidate and can legitimately question them on their voting habits. This article can be read both as an argument against abstention and as an argument against purism. When in multiparty systems elections are framed as VAHR, the account proposed requires counterspeakers to consider that their single vote has a non-transferable capacity to contribute to a speech act and can reinforce what the public considers an acceptable communication in democratic campaigning. Therefore, in elections framed as VAHR, if counterspeakers aim at unquestionably expressing what they deem an unacceptable form of public political communication, they should contribute to the collectivised speech act that is more likely not to misfire, even if this means voting for a second-best candidate. Seen through these lenses, the act of voting against a hateful representative requires counterspeakers to consider at least how the media and politicians have framed the elections, the relationships between candidates, the possible gap between candidates, and the candidate who is more likely to generate an outcome that the audience can treat as a speech act challenging the choice of making hateful representative claims.Footnote 11