Abstract
It is often claimed that democratic voters have epistemic responsibilities. However, it is not often specified why voters have such epistemic responsibilities.
In this paper, I contend that voters have epistemic responsibilities because voting is best understood as an act that bears assertoric force. More precisely, voters perform what I call an act of political advocacy whereby, like an asserter who states or affirms that something is the case, they state or affirm that a certain course of political action is the one that should be followed or enacted.
Consequently, the performance of acts of political advocacy such as voting should be understood as bounded by epistemic norms mirroring those binding the act of assertion and yield epistemic responsibilities mirroring the ones required to satisfy these norms.
In* recent years democracy has come under renewed criticism on epistemic grounds, and skepticism concerning its ability to deliver good political decisions has resurfaced. Part of this skepticism focuses on the behavior of the electorate and, more precisely, on the allegedly mediocre epistemic capacities of voters.1 Phenomena such as the spreading of so-called fake news, or findings about the misguided beliefs of electorates on political issues such as migration,2 have breathed new life into the suspicion that the average voter is a mediocre political decision maker, too prone to falling prey to misinformation, intellectual negligence, tribalism, etc.
These concerns rest on the conviction that democratic citizens are not capable of discharging the epistemic responsibilities associated with voting powers. But why should we think that voters have such epistemic responsibilities in the first place? Before we can attribute any epistemic responsibility to voters, we need to explain in what sense voting could be understood as the exercise of an epistemic agency. In this paper, I suggest an answer to this question that I take it to also facilitate a better understanding of the epistemic responsibilities of voters.
The idea central to the account that will be developed in the following pages is that voting can be understood as an act that bears assertoric force. More precisely, I will argue that the role of voters within democratic decision-making practices can be understood as that of performing an act of political advocacy. With this, I mean that their specific task is to contribute to political decision-making by stating or affirming what they take to be the course of political action that should be followed or enacted by the polity. I will then argue that, because this act mirrors that of an asserter who states or affirms that something is the case, it yields epistemic responsibilities similar to the ones required to satisfy the epistemic norms that regulate the act of assertion.
The paper will be structured as follows. In section (1), I present and defend an interpretation of voting as an act of political advocacy and explain why acts of political advocacy should be understood as bearing assertoric force.3 In section (2) I consider potential objections to the account and provide further important clarifications. In section (3), I explain how the account of voting as an act of political advocacy can facilitate a better understanding of the epistemic responsibilities of voters. I argue that much like in the case of ordinary assertions, the performance of acts of political advocacy such as voting is bound by epistemic norms. Consequently, on this view, the epistemic responsibilities of voters consist in doing what is necessary to satisfy the relevant epistemic norm. I will suggest that, whichever epistemic norm we take to apply, voters will be required to master a few basic competences. Section (4) concludes the paper by briefly addressing the implications of the account for the practical normativity of voting.
1 Voting as an Act of Political Advocacy & the Assertoric Force of Votes
Let us define a role as a cluster of specific tasks and powers that applies squarely because – and only insofar as – one acts within the boundaries of a certain practice.4 In discussions concerning the normative requirements associated with participation in democratic practices, it is often claimed that the role of voters entails epistemic responsibilities.5 The goal of this paper is to substantiate and strengthen this common thought. I intend to do so not by focusing on the content of voters’ epistemic responsibilities, but rather by offering a more detailed explanation of what kind of epistemic agency is involved in voting. In other words, the paper seeks to offer a new specification of the broad idea that voting has an epistemic dimension that yields distinctively epistemic responsibilities. I do so out of the conviction that this would also, further down the line, facilitate a more precise outline of the content of voters’ epistemic responsibilities. Hence, much of the work conducted in this paper has exploratory aims. It is meant to lay the groundwork for a potential line of inquiry, rather than to delve deep into it.
The idea central to the account developed in these pages is that the exercise of the tasks and powers constituting the role of voters within democratic decision-making practices is to be understood as the performance of an act that bears assertoric force. As such, the epistemic normativity governing the act of voting can be understood in terms of norms of assertion. Let me emphasize, from the get-go, that the account has normative rather than descriptive ambitions. That is, it is aimed at outlining what an appropriate usage of the prerogatives and powers associated with the role of voters requires, and not so much at providing a description of what real voters do in current democratic practices.6
In this first section, I will focus on outlining why voting is to be construed as an act that bears assertoric force. Before I can move on to that task, let me briefly clarify an important assumption. Determining what is the role of voters within practices such as elections and referenda might look like a trivial issue as, almost tautologically, voters are public decision makers. The specific cluster of tasks and powers associated with being a voter is that of contributing to a political decision-making procedure by providing a personal input that will be later aggregated to the one provided by other voters. The thought is correct but underdetermined. It calls for a more detailed explanation of the nature of the personal input that each voter provides. Now, any further specification which we might give in this regard will be inevitably informed by how we understand the aims of voting practices. What is their underlying rationale or function within democratic societies? Do we think that they are meant to settle a specific issue? And if we do, what kind of issue? Depending on how we answer these questions, we will have a different interpretation of what kind of action voters are meant to perform when they participate in public decision-making and provide their own inputs to it.
In this paper, I will proceed on the assumption that the aim of voting practices is that of settling a specific political issue, namely what course of political action the polity should undertake. With the term ‘course of political action,’ I mean a broad set of political proposals, priorities, and desired political outcomes that are usually unified within a political project or agenda. A course of political action can be determined directly, via a referendum, or indirectly by delegating or entrusting a selected representative to do so on behalf of the public.7 Now, my point here is surely not that the significance of voting in a democratic society can be entirely reduced to the function of determining a course of political action. My point is rather that it seems implausible to think about voting as a practice that is divorced from the pragmatic goal of determining in which political direction a democratic polity will be steered. If they did not determine the pursuit of one political project rather than another, voting practices would be stripped of a constitutive dimension of their pragmatic rationale. Whatever other aims voting practices are meant to serve, it seems therefore impossible to escape the thought that, at the very least, they serve a ‘settling function’ of this kind: they determine, among a variety of political projects and agendas, which one is to be pursued by the governing bodies of the political community. I take it that to act as a voter generally means to act as a contributor to this endeavor, which is basically that of adjudicating what the polity should do.8
With this background understanding in place, let me illustrate how the personal input that each voter provides to public decision-making practices should be understood. Granted that an assertion is commonly understood as a speech act by which an agent states or affirms that something is the case, my proposal is that a vote should be similarly understood as an act of political advocacy by which the voter states or affirms that they take P rather than Q to be the course of political action that should be followed or enacted. Again, the point is normative: making such a statement or affirmation is what an appropriate usage of voting powers requires. In this section, my focus will be on laying out the core elements of the view. A few relevant objections and issues, including the implications of endorsing a specific understanding of the aims of voting practices, will be discussed in the following section.
The key to construct voting as the performance of an act that bears assertoric force can be found in the idea that votes have an advocative dimension. This is a notion best accounted for by David Estlund. As Estlund explains, if voters are indeed faced with a public decision-making task with a specific objective at stake, then votes are ‘off the fence’ with respect to such objective.9 That is, voters’ inputs to the procedure must be thought of as entailing at the very least some kind of normative support or opposition towards the options that face them. By going to the polls and casting their tally, voters take a position with respect to these options. In this sense, a vote for P implies that the agent supports P and calls for P rather than Q to be the course of action that should be implemented.10 Now, this support or opposition need not be conceived of as one of complete endorsement or alignment with the option. Quite clearly, voters often have to choose between options that do not have their full support or endorsement. But if we hold firm in the background the idea that voting practices settle the issue of what course of political action should be undertaken and that a voter’s role is that of contributing to this endeavor, then the voter’s decision to cast their tally for P rather than Q cannot be considered neutral. It must be thought of as an action that implies taking a stance on the public issue at stake. We must think of it as an action whereby the voter bestows a larger normative support on P rather than on Q. Estlund correctly claims that if votes did not possess such an advocative dimension, the outcome of a voting procedure would be devoid of any indication about what to do next. In other words, it would not signal anything about what should follow from its result. Without recognizing an advocative dimension to votes, even unanimous outcomes would leave the electorate’s decision undetermined, leading to the paradox of a procedure of political decision-making that is not indicative of any underlying collective political choice.11 In other words, if votes for P were taken to be devoid of even a minimal degree of normative support for P and if they were taken not to entail the idea that P has been called for by those who voted for it, then it becomes unclear why we should take P as being any more representative of the public’s decision than Q. This would call into question, if not undermine, the very rationale of having something like a collective decision-making procedure in the first place.
The fact that votes possess an advocative dimension is crucial in constructing a correspondence between the act of voting and the act of asserting, as it makes emerge the reason why the two acts share a crucial epistemic feature. In particular, it showcases why voting can be understood as an act that, like ordinary assertions, bears assertoric force. Generally, an utterance having assertoric force means that the speaker presents the content of the utterance as reflecting how things are or as being worthy of epistemic support, broadly construed.12 The content of the utterance is presented by the speaker as something that should be taken as true, warranted, justified, and to reflect what the speaker believes to be the case with respect to a certain issue. The thought here is that, because of its advocative dimension, a vote functions in analogous ways. Suppose a certain political constituency must decide between two political representatives, P and Q. On this view, the input of a voter counts as an answer to the specific social question of which course of political action that the polity should undertake between the one associated with P and the one associated with Q. The input that the voter gives is not neutral. It is advocative; it calls for the option chosen and entails normative support for it. Thus, the idea central to this account is that a vote for P, being an advocative ‘answer’ to the social question of what the polity should do, conveys that the voter takes it to be the case that P is indeed the course of political action that should be undertaken. The vote conveys what the voter believes to be the right answer to the question about what the polity should do, much in the same way as an assertion conveys what a speaker believes to be case with respect to a certain background issue. Both actions, voting and asserting, entail something analogous to a statement in favor of a certain content. In the case of an assertion, the assertion gives voice to the speakers’ statement in favor of the content of the proposition. In the case of votes, the vote gives voice to the voters’ statement in favor of a specific solution to the public decision-making procedure. The vote thus can be construed as bearing assertoric force because, in the same way as an asserter states or affirms that something is the case, a vote is an act whereby the voter states or affirms that a certain course of political action is the one that should be followed or enacted. Again, my point is not that all acts of voting for P are to be descriptively understood as entailing the assertoric statement that P is the right course of political action to be undertaken. Many voting behaviors do not seem to fit this description. My point is rather that this is what is required in order to exercise the prerogatives associated with the role of voters in a proper fashion. I will strengthen this claim in the next section, addressing also potential leaps in the argument.
With the key elements of the view on the table, let me provide a couple of important clarifications. The first pertains to metaethical issues that the argument might raise. An assertion is an utterance with truth value. Its content can either reflect how things are or not. By construing voting as the act of stating or affirming that such-and-such is the course of political action that the polity should undertake, I cannot avoid the implication that votes have truth value as well. That is, I cannot avoid the implication that there is a truth of some sort to the question about what a political community should do, and a vote can either reflect it or not. Thus, I am working on the cognitivist assumption that normative statements, including statements about politics, admit distinction along the lines of truth or falseness. It is important to clarify, however, that this need not entail any thick realist commitment or robust notion of truth. The idea that acts of political advocacy such as voting can be truth-apt normative statements can be understood in less burdensome ways. It can be understood in a minimal deflationary sense and without reflecting any metaphysically heavyweight conception of truth.13 It can be understood in broad constructivist terms, by depicting normative truths about politics as reflecting what rational agents would agree upon in properly constrained circumstances rather than mind-independent moral facts. Therefore, the argument outlined in the pages of this paper succeeds as long as normative statements about politics can be regarded as more than taste judgements, without any further metaethical commitment.
The second clarification concerns the correspondence between the act of asserting and the act of voting on which my argument hinges. In the previous paragraphs, I tried to highlight how voting and asserting are acts that share a key epistemic feature, namely assertoric force. They are both acts whereby a subject presents a certain content as reflecting what they take to be the case with respect to a certain issue. What if assertions have properties that do not perfectly translate to votes? Take, for example, the informative function of assertions. An assertion is a speech act whereby a speaker states something to a hearer. Assertions thus convey information, play a role in the transfer of knowledge, and can represent sources of testimony within public conversational practices. Votes are mostly a private act that individuals perform in the secrecy of the ballot box. It could then be argued that they do not share the informative function of ordinary assertions, at least not straightforwardly.14 There might be more examples of properties of assertions that do not translate to voting. However, I do not take these to present any crucial threat to my account. My purpose here is to support the conclusion that we can understand the epistemic norms binding the act of voting in terms of the epistemic norms that bind the act of assertion. For this purpose, voting and asserting need not share each and every epistemic property. If they do share the same force as I have argued, such that votes can be plausibly treated as statements concerning what a voter takes to be the case with respect to what the polity should do, this seems to provide votes with enough ‘epistemic edge’ to sustain the normative claims that will be spelled out in the rest of the paper.
For the sake of simplicity, throughout the rest of the paper I will refer to this specific interpretation of voting as the account of voting as an act of political advocacy. Recall that the ultimate purpose of the account is twofold. On one hand, I want to offer a reconstruction and specification of the common thought that certain epistemic responsibilities attach to voting powers. On the other, I want to pave the way for a reconstruction of what these epistemic responsibilities might be. Before I can accomplish any of these things, it is perhaps worth spending a few words to strengthen the plausibility of the account by dispelling some of the perplexities that it might raise. I turn to this task in the next section.
2 Objections & Clarifications
The central feature of the account of voting as an act of political advocacy that I presented in the previous section is the idea that voting can be construed as an act that bears assertoric force. In this section, I will address two perplexities. Both target the idea that the advocative dimension of votes supports treating them as assertions concerning what the voter takes to be the course of political action that the polity should undertake. Neither gives us reason to reject the account.
The first perplexity concerns a potential leap in the argument.15 Suppose we accept that a vote for P has an advocative dimension in the sense that it entails normative support for P. From this, it does not necessarily follow that the vote is to be treated as a statement expressing that P is the option that is worthier of normative support and that should be enacted by the polity. Take the case of someone who advocates the course of political action associated with option P because it entails a significant tax cut that would benefit the voter’s business. In this case, the vote would not entail a statement that P is the best course of action for the polity to take but rather merely that P is the course of action preferred by the voter in light of their own interests. In other words, the worry here is that a vote for P can have an advocative dimension and yet that this need not translate into the vote bearing any assertoric force.
How are we to address this potential leap in the argument? I concede that, from a descriptive standpoint, votes can have an advocative dimension and yet not bear any assertoric force nor entail any statement presenting the option chosen as the right one. The point, however, is that such a usage of voting powers would be somewhat improper in my view. The assumption that voting practices are aimed at the determination of a course of political action plays a crucial role in sustaining this claim. Such an understanding is predicated upon the idea that voting practices are inescapably underpinned by a rationale of settling a specific political issue. On this assumption, voters are meant to employ the decision-making prerogatives associated with their role to address the issue at stake in the ballot box. And there is reason to hold that a voter can properly address the issue at stake only by means of a statement or affirmation of their beliefs on the matter. Estlund himself explicates this point while discussing issues of aggregation. Estlund thinks that we can provide a convincing account of the aggregation of different voting inputs only insofar as we can understand them as being in a relevantly similar relation to the issue they address. Consequently, while votes for P and votes for Q will point to different choices, they will be relevantly similar insofar as they are construed as inputs that have the same referent to begin with: the choice between P and Q. That is, votes must be construed as inputs that have a common objective or, to repeat Estlund’s analogy, as ‘answers to the same social question.’ Without any such common objective to which they can be referred to, it becomes extremely difficult to provide an account of what the votes’ aggregation represents and hence to identify what it is that a voting procedure has effectively selected. And while Estlund grants that votes that report desires or preferences might be advocative and ‘off the fence’ with respect to the options on the table, he argues that it is not possible to provide a convincing account of their aggregation precisely because they fail to properly refer to the same object or issue. If X reports on their preferences and Y reports on their preferences, X and Y are addressing the same issue only in a formal sense. In terms of content, each of them is addressing the ‘agent-relative’ issue of their own inclinations with respect to the political options on the table rather than the ‘agent-neutral’ issue of which of these options is the one that should be undertaken.16
To simplify matters, the important thought is that the nature of the ‘social question’ underlying voting practices restricts the range of what may or may not count as a proper answer to it. There is a difference between asking me what I want to eat for dinner and asking me what it is better for me to eat in virtue of my current diet. The former question refers to an agent-relative issue. The latter refers to an agent-neutral one. And whereas there would be no problem in answering “I would like to eat pasta” to the former, the same cannot be said in the latter case. Epistemically speaking, such an answer would not address the question I was asked, as it would not be explicating what I think is best but rather what I would be inclined to do. The same applies to questions that pertain to a collectively relevant issue. What I think that we should do on a given issue and what I would prefer that we do may coincide at times, but they are indeed two different matters. Consequently, as long as we hold on to the idea that voting procedures serve the function of settling the agent-neutral ‘social question’ pertaining to what course of political action the polity should follow, someone who answers it by reporting something like their desires or preferences is not properly addressing the social question at hand, in that their input is only tangentially referring to the issue that the question is meant to settle. So, when I say that a proper usage of voting powers should be construed as the exercise of the act of political advocacy bearing assertoric force, what I mean is that only votes in which the advocative dimension of the input translates into an act with assertoric force are truly in keeping with the underlying rationale of the voting practice and represent a proper fulfillment of the decision-making task associated with the institutional role of a voter.17
Now, this response is admittedly only as strong as the assumption from which it stems. While I did not make appeal to loaded notions such as the common good or the general will of society, I nevertheless proceeded on the assumption that the aim of adjudicating a specific political issue represents an important normative dimension of the practice of voting, a dimension that inevitably informs how we understand the role of voters and what they may or may not do in the exercise of their prerogatives. I did so because, whatever else we may think about the function of voting in democratic societies, it seems to me difficult to dispense with this dimension and more generally with how voting establishes that certain policies, aims, programs, and values will be pursued over others. Obviously, if we instead hold that something like the equal accommodation and representation of the diverging political interests that inhabit a certain society is all that is normatively relevant to voting practices or what is most important anyway, then the response offered above will not be palatable.18 The view that I outline in these pages has the inevitable implication of leaving a narrower normative margin for voting behaviors that fail to bear assertoric force (i.a., self-interested voting; preference-based voting; etc.). This is not to say that citizens of a democratic polity are never permitted to vote in these ways. Again, my ambition here is to explain the epistemic dimension of the responsibilities of voters, not to give a full account of all the normative considerations that apply to voting in a democracy. But it is true that, in order to override the task of addressing the question of what course of political action the polity is to follow and thus legitimize these forms of voting, my theory would undoubtedly require comparatively weightier moral reasons than alternative theories would.
Let me now turn to a second perplexity targeting the idea that voting can be construed as an act that bears assertoric force. Treating a vote for P like an assertion presenting P as the ‘right’ course of political action is implausible, the perplexity goes, because whereas assertions have a determinate content that the asserter presumably intends to convey, votes do not have such a determinate content or at least not an easily discernible one. From this standpoint, even ascribing an advocative dimension to votes is often an unwarranted move that attributes to votes an implication, i.e., that the voter takes a certain political option as comparatively worthier of normative support, that might not be there at all.19
Protest voting and strategic voting provide the standard examples in this regard. For what concerns the former, think about Brexit. Interpreting voting as the performance of an act of political advocacy would require us to attribute to any vote for Leave an advocacy in support of the withdrawal of the United Kingdom from the European Union. But it could be argued that many of those who voted to leave the EU did so in order to convey a far more generic sense of dissatisfaction with the political establishment rather than their support for Brexit. For what concerns strategic voting, take the case in which a voter wants to vote for R but ends up voting for P because P is deemed most likely to hinder Q’s victory. In this case, the vote for P does not seem to entail any actual support or advocacy for P.
The objection, in other words, points to a discrepancy between the chosen political option (and the proposals and policies associated with it) and what the voter intends to advocate through her tally. I will not deny that such a discrepancy might occur. The examples mentioned above make it quite clear that, on the contrary, it is a common phenomenon. We should be cautious, however, not to overestimate the significance of this discrepancy and the force of the objection associated with it. The account of voting as an act of political advocacy is predicated on what I take to be a plausible interpretation of the epistemic dimension of voters’ role and of the powers associated with it. I have argued that the role of voters inevitably entails acting as public decision makers who, through their inputs, are called upon to settle which course of political action will be pursued by the community. If this understanding, which links the role of voters with a ‘settling function’ that is taken to underlie voting practices, is plausible, then whether some further or different intention, i.e., strategy or protest, underlies the vote does not change the fact that, within the functioning of public decision-making procedures, and in light of the tasks and powers associated with the role of voters within them, a vote for P amounts to an act that implies a certain degree of normative support for P. On what kind of intentions this support is predicated does not matter. Suppose that I choose to vote for Brexit because I want to ‘shake up things’ even if I do not necessarily endorse the course of action of withdrawing from the EU. There is a discrepancy between what I am lending support to and what I take myself to be advocating. But this does not change the fact that, within the functioning of the practice, my vote will represent ‘a voice’ in support of the withdrawal of the UK from the EU. The fact that my exercise of decision-making powers is predicated on protest intentions is my own responsibility.
The implication for the present account is as follows. In this case, the fact that my vote in favor of leaving the EU does not correspond to my actual ‘advocative intentions’ warrants treating the vote as an insincere assertion, but an assertion nonetheless. As much as someone can utter a statement whose content does not really correspond to what the speaker actually takes to be the case, a voter can voice support for an option that does not correspond to what they actually take to be the course of political action that should be followed for reasons of strategy, protest, and so forth. But nothing in the example undermines the idea that the vote to leave the EU should count, with respect to the question that the referendum is meant to settle, as a statement in favor of the UK withdrawing from the EU. Now, whether it represents a proper statement is a different question. And, at least from the standpoint of the epistemic normativity of voting that I am trying to spell out in these pages, it is quite clear that citizens who engage in strategic and protest voting engage in a voting behavior that, by virtue of not reflecting what they actually take to be the course of political action that ought to be followed, is at odds with the epistemic dimension of their responsibilities and thus somewhat improper. Again, this does not necessarily preclude the possibility that normative considerations of a different kind may justify insincere acts of political advocacy despite their epistemic impropriety. Indeed, I suspect that several normative considerations may be appealed to in this regard but since spelling them out is a task that cannot be accomplished without reference to a broader conception of the value of democracy, I set this aside for the time being. The main point of interest here is that discrepancies between the ‘true’ advocative intentions of voters and the options they choose do not seem to undermine the interpretation of voting as an act of political advocacy.
If the considerations mentioned in these two sections hold, we should have reasons to accept that a proper exercise of the prerogatives associated with the role of voters requires the performance of an act that bears assertoric force. As members of a community involved in the endeavor of trying to settle a certain public matter, we ought to contribute to this goal by doing something analogous to ‘declaring’ what we take to be the solution to the matter.
3 Voting, Political Advocacy, and the Epistemic Responsibilities of Voters
In the previous sections I discussed the first part of my account and argued that, if votes have an advocative dimension, then voters perform their role by means of an act of political advocacy that bears assertoric force. Like an asserter states that something is the case, a vote is an act whereby the voter states that P rather than Q is the course of political action that should be followed.
In this section, I will build upon this interpretation of voting and explain how the account of voting as an act of political advocacy not only substantiates the thought that voters have epistemic responsibilities, but also paves the way for an understanding of what these responsibilities might be.
I will do so through the following argument. If the act of voting can be indeed understood as the performance of an act that bears assertoric force, it is an act bound by the epistemic norms regulating the act of assertion and warranted only insofar as these norms are met. Whatever norm of assertion we take to apply to voting, the epistemic responsibilities of voters will consist in doing what is necessary to be in the position to meet the requirements of the norm and vote in a warranted way. At face value, this will entail at least mastering a basic set of crucial competences.
The first point – voting is an act bound by epistemic norms – follows as an implication of the correspondence between voting and assertions that underlies the account of voting as an act of political advocacy. The epistemic norms that regulate acts like the uttering of assertions are the subject of significant debate. Part of this debate focuses on fleshing out the grounds on which assertions are warranted. This is a significant issue for us because it yields the criteria that allow us to undertake an epistemic evaluation of asserters. Some epistemologists argue that only knowledge warrants assertion: you can legitimately assert P only if you know that P.20 Some others have contested this norm and argued that it should be replaced with a justification norm: an assertion is warranted only if the agent uttering it can provide some form of justified belief or epistemic support for it.21 Some others have argued for a norm of safety: an assertion is warranted only if the agent utters it on an epistemic basis that not too easily could have led to a false assertion.22
If I have been indeed correct in suggesting that voting for P is to be constructed as an act bearing the force of an assertion such as “P is what should be done,” then voting triggers epistemic norms much in the same way as ordinary assertion-making does. I will not take sides in the debate on which epistemic norm regulates the act of assertion, nor try to settle which norm of assertion translates best to the case of voting. For the time being I will proceed ecumenically, as this is enough to make the crucial point emerge. The crucial point is that the uttering of assertions is something that is the object of epistemic evaluation and that there is consensus on the idea that some assertions are unwarranted and have insufficient grounds to be uttered. If voting, as argued above, is an act that bears assertoric force, whereby we contribute to a public decision through the affirmation of what we take to be the solution to the matter, then a similar point applies: there will be requirements that, if not met, would make voting for P an epistemically unwarranted act. For the sake of parsimony, suppose that the right norm is a relatively permissive one, such as justification.23 From this standpoint, only votes that are backed up by sufficient epistemic support would be warranted. More precisely, the epistemic support mustered would have to be strong enough to give the voter good reason to believe that P is the course of political action to be enacted.
On these bases, we can also draw a general framework for specifying the content of the epistemic responsibilities of voters, namely if the act of voting for P is warranted only insofar as the requirements associated with an epistemic norm are met, the epistemic responsibilities of a voter will consist in doing what is necessary to put themselves in the position to meet such requirements. And whatever equivalent of the norm of assertion that we think applies, it seems clear that a voter will be more likely to meet the requirements associated with the norm (and hence vote in a warranted way) only insofar as they master certain basic competences. For instance, it appears clear that the voter will have to gather correct (or not obviously false) information about relevant political facts and of the disputes concerning their interpretation. Furthermore, the voter will be required to master the tools necessary to make a comparative evaluation of political options, such as having a certain degree of understanding of what these political options entail and of their likely impact on the community. Finally, the voter will be required to be receptive to the issues faced by the political community, as well as to reflect on what is at stake in them and how they relate to previous states of affairs.
The reasons in support of this last set of claims should be quite straightforward. Ordinary assertion-making is the exercise of an epistemic agency. Regardless of what specific epistemic norm we deem to apply, it seems quite clear that for that agency to be properly conducted, the subject performing it needs to be at least within a range of minimal agential capacities or competences. For what concerns assertion-making, any plausible epistemic theory would consider an utterance concerning what ought to be done that has been formulated without knowledge of the relevant facts, or without weighing the options at disposal, as an utterance that falls below conditions of sufficient epistemic support and hence of warranty, perhaps to the point of making the asserter epistemically negligent.24 If this is true for epistemic agency in general, and if voting is indeed an act of political advocacy that bears assertoric force, any conception of the epistemic support that voters ought to muster in order to perform this agency properly will require such basic competences.
This is not to say that this will constitute everything that is relevant for an epistemically responsible performance of the role of voter nor that mastering such competences guarantees that voters will choose the best political option available. Recall that my purpose here is not to build a full account of the epistemic responsibilities of voters, but rather to lay the grounds for such an account. Hence, for the time being my point is just that basic competences such as the ones mentioned above will be the minimal starting point of any account of this kind. I am by no means excluding that the epistemic responsibilities of voters might extend further. I suspect that, among the various norms that might apply to voting, some of them might encourage more stringent accounts in this regard. But whether this is truly the case or not is something that I will set aside for other inquiries.
4 The Epistemic Responsibilities of Voters and the Practical Normativity of Voting
The aim of the analysis conducted in this paper was to substantiate and strengthen the common thought that, as part of their role within democratic decision-making practices, voters have epistemic responsibilities. I have done so by showing how the role of voters requires the performance of an act of political advocacy, an act that bears the force of an assertion. I have then argued that such an account not only supports the conclusion that voters do have epistemic responsibilities, but that such responsibilities are to be understood in terms of the requirements which are necessary to meet whatever norm of assertion that regulates the act of voting.
Now, in substantiating and strengthening the common conclusion that voters have epistemic responsibilities, I have borrowed from considerations that relate to the epistemic normativity that regulates the uttering of assertions. But unwarranted assertions are not necessarily morally problematic and the fact that an assertion is bound by a certain epistemic norm does not necessarily entail that an asserter has anything like a moral duty to abide by that norm.25 Similarly, the fact that voting is an act bound by a certain epistemic norm does not necessarily establish that there is anything like a practical normative requirement that binds voters to meet that norm. We may then have reason to ask what, if anything, gives practical normative relevance to the idea that the role of voters requires an act of political advocacy and thus compliance with the ensuing epistemic responsibilities. In this concluding section, let me briefly turn to this issue and spell out how the proposed account of voting can contribute to discussions concerning the practical normativity of voting.
The account substantiates the thought that certain epistemic responsibilities follow from the tasks and powers associated with the role of voters in public decision-making practices by means of an epistemic interpretation of the kind of agency that the role requires from its occupants. As such, once coupled with a plausible defense of the idea that voters have something like an individual moral obligation to uphold proper conduct whenever they act in their institutional capacity, the account could serve as a premise within a larger argument aimed at showing that certain epistemic responsibilities act as moral constraints on the behavior of voters. How might we go about defending the idea that voters have a role-based moral obligation of this kind? A natural way to proceed is by drawing from a broader conception of the value of democracy. The commitment to uphold the value of democracy, however understood, is what could give practical relevance to the requirements associated with the role of voters and, consequently, give moral force to the idea that a proper usage of voting powers calls for the performance of an act of political advocacy entailing epistemic responsibilities.
Now, in light of the theoretical commitments incurred along the course of these pages, it is tempting to infer that the present view is bound to align with standard epistemic conceptions of democracy. That is, it is tempting to infer that it is bound to articulate the value of democracy instrumentally, exclusively in terms of a tendency to make good political decisions, and thus that the goal of getting political decisions right is what gives moral force to the account of voting as an act of political advocacy. This inference should be resisted, at least partially. On one hand, standard epistemic conceptions of democracy are undoubtedly a natural outlet for the account of voting presented in these pages. Theories of this kind rely explicitly on assumptions of minimal competence on the part of the electorate, and the account can find a place within them by specifying the epistemic requirements by which a minimally competent voter would have to abide for the epistemic qualities of democracy to be reaped. That said, the account is by no means committed to following this approach and is indeed compatible with different theories.
It is important not to overstate my claims here. Having worked under the explicit assumption that the aims of voting practices are to be understood in terms of adjudicating a specific political issue, rather than merely in terms of guaranteeing a fair and equal accommodation of diverging interests, the account can be ecumenical only to a certain extent. For instance, it is obviously not compatible with conceptions in which the value of democracy is spelled out exclusively in terms of political equality, just to give an example. But the assumption that voting practices are meant to serve the purpose of adjudicating what course of political action should be undertaken ‒ and thus the ensuing argument that the role of voters is to state what they take to be the right option in this regard ‒ does not necessarily commit us to claim that the value of democracy is exhausted by whether the practice of voting gets this judgement right. The account can be employed as part of mixed approaches, in which epistemic and practical considerations concur in determining the value of democracy.26 Or it can be employed as part of conceptions in which epistemic considerations play a role in establishing the value of democracy, but for reasons that are ultimately derivative of broader non-epistemic commitments. For instance, the account is compatible with conceptions that understand the value of democracy in terms of bringing about collective autonomy or guaranteeing reciprocal civic accountability, and yet recognize how these non-epistemic values can be realized only on certain epistemic preconditions.27 The account could feed into these conceptions by specifying the conduct that voters ought to uphold for these preconditions to obtain. On this non-standard approach, the account of voting as an act of political advocacy acquires moral force as part of a broader commitment to realize non-epistemic values, and not in the name of getting things right.
A more thorough discussion of these issues is needed before we can draw further conclusions. But what this paper has hopefully confirmed is that proper conduct as voters will inevitably require the fulfillment of some epistemic responsibilities and thus that no plausible ethics of voting can be successfully constructed if we turn a blind eye to the epistemic dimension of voting practices.
Acknowledgements
This paper draws on my doctoral dissertation, defended at the University of Warwick. For helpful discussions and comments on various drafts of this paper, the author wishes to thank Alfonso Anaya, Sameer Bajaj, Matthew Clayton, Michel Croce, David Estlund, Maria Paola Ferretti, Corrado Fumagalli, Samuel Honsbeek, Carline Klijnman, Zsolt Kapelner, Thomas Mulligan, Chris Noonan, Valeria Ottonelli, Fabienne Peter, Patrick Tomlin as well as two anonymous referees. The paper has also greatly benefitted from the feedback of audiences at Brown University, Central European University, the University of Pavia and the University of Warwick. Part of this work was generously supported by the H2020 research project redem ‘Reconstructing Democracy in Times of Crisis: A Voter-Centred Perspective’ (grant n° 870996).
Biographical Note
Michele Giavazzi is a Postdoctoral Research Fellow at the University of Genoa, Italy. Before joining Genoa, Michele was a PhD Student in philosophy at the University of Warwick and a Visiting Research Student at Brown University. His research interests lie at the intersection of democratic theory and political epistemology with a particular focus on epistemic challenges to democratic legitimacy.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 870996. The article reflects only the author’s view, and the Agency is not responsible for any use that may be made of the information it contains.
Jason Brennan, Against Democracy (Princeton: Princeton University Press 2016); Ilya Somin, Democracy and Political Ignorance: Why Smaller Government Is Smarter (Stanford: Stanford University Press, 2013).
See the regularly conducted survey ‘Perils of Perception’ by Ipsos Mori.
As I explain below, the notion of advocacy is borrowed from David Estlund, ‘Democracy without Preference,’ The Philosophical Review 99 (1990), pp. 397–423. Fabienne Peter’s work on the relevance of the literature on practical reasoning for political normativity provided another important source of inspiration. See Fabienne Peter, ‘Epistemic Norms of Political Deliberation,’ in M. Hannon and J. de Ridder (eds.), The Routledge Handbook of Political Epistemology (New York: Routledge, 2021), pp. 395–406.
Arthur I. Applbaum, Ethics for Adversaries. The Morality of Roles in Public and Professional Life (Princeton: Princeton University Press, 1999). Michael Hardimon, ‘Role Obligations,’ Journal of Philosophy 91 (1994), pp. 333–363.
For an overview of how the current literature conceptualizes the broad issue of the epistemic responsibilities of democratic citizens, see Cameron Boult, ‘The Epistemic Responsibilities of Citizens in a Democracy,’ in M. Hannon and J. de Ridder (eds.), The Routledge Handbook of Political Epistemology (New York: Routledge, 2021), pp. 407–418.
Let me also state that, while the paper has normative ambitions, these are restricted to explicating the epistemic normativity of voting. That is, my aim here is not to give a complete overview of all the normative considerations that apply to voting nor, as it will emerge in various passages of the text, to argue that epistemic considerations are all that matters for voting. Consequently, I will leave open how conflicts between epistemic and practical normativity in voting should be settled.
That the account applies to both referenda and representative elections might not be evident at first glance. While direct voting procedures are indeed meant to settle a specific issue, during general elections voters delegate or entrust a representative with the powers to make such choices on their behalf. Hence it might be argued that the concept of a course of political action (and with it the idea that voting practices are meant to settle a specific political issue) is not suitable for the case of general elections. Now, while I am happy to concede that the present account applies more intuitively to direct decision-making procedures, I do not think this is a major problem for my argument. Even if democracy usually takes the form of a representative political system, electoral choices nevertheless take place against a background of differential projects, proposals, and desired outcomes. Unless it can be shown that the choice of a representative can be disconnected from the choice of the proposed course of action in the name of which the representative supposedly stands – an idea that I frankly cannot see how to plausibly support – the distinction between direct and indirect voting procedures is not relevant for my purposes.
It is important to clarify that I conceive of the role of voters as being restricted to choosing among the agendas that are available to them. Even though citizens of a democratic polity might contribute quite significantly to setting political agendas, I do not take this to be something that they do when they act as voters. The process by which political agendas are set is antecedent to the voting procedure and sees citizens involved through various forms of political activism. However, once they are in the ballot box, citizens can only focus on the options available to them. I thank Carline Klijnman for drawing my attention to this issue.
David Estlund, ‘Democracy without Preference,’ p. 404.
David Estlund, ‘Democracy without Preference,’ p. 419.
David Estlund, ‘Democracy without Preference,’ p. 404.
Sanford Goldberg, Assertion. On the Philosophical Significance of Assertoric Speech (Oxford: Oxford University Press, 2015). Peter Pagin, ‘Information and Assertoric Force,’ in J. Brown & H. Cappelen (eds.), Assertion: New Philosophical Essays (Oxford: Oxford University Press, 2011), pp. 97–136. François Recanati, Meaning and Force. The Pragmatics of Performative Utterances (Cambridge: Cambridge University Press, 1987); Robert Stalnaker, Context and Content (Oxford: Oxford University Press, 1999).
See also David Estlund, Democratic Authority (Princeton: Princeton University Press, 2008), pp. 24–28. I am indebted to David Estlund and Chris Noonan for helpful discussions on these issues.
While I will not pursue this line of argument, I do think that voting can have a similar informative function precisely in virtue of its advocative dimension. On this view, a vote can be seen as providing information or even a testimony, in that it represents a tool through which participants in the ballot box let it be known or make it explicit that their stance over the matter is such-and-such.
I thank two anonymous reviewers for pushing me to clarify my thoughts on this issue.
David Estlund, ‘Democracy without Preference,’ pp. 411–416. Estlund labels the aggregation issue faced by preference-based interpretations of voting as ‘the Indexical Problem’ (413). From this standpoint, votes that report individual preferences equate to statements such as “I prefer the course of political action P” or “I prefer the course of political action Q.” Because of the indexical “I” that characterizes them, these statements have different referents and thus fail to address the same issue. Notice how the requirement that votes have a common referent and address the same issue is present also in other epistemic theories of democracy, such as those modeled after the Condorcet Jury Theorem.
Perhaps the criteria for an appropriate voting behaviour may vary depending on the kind of political decision at stake, as explained in Steven Wall, ‘Democracy and Equality,’ The Philosophical Quarterly 57 (2007), pp. 427–428. Votes that lack assertoric force may be appropriate for political decisions in which the course of action that should be followed is determined by no other consideration than what the majority wants, as in such case there would be no ‘answer to the social question’ to be figured out that is independent of the preferences of the electorate. Like Wall, however, I doubt that most political decisions are of this kind. More often than not, political decisions take place in circumstances in which a wide variety of normative as well as empirical considerations weigh on what the polity should do. This constrains the kind of conduct that voters should uphold, at least on the present view. It remains to be seen how the interpretation of voting as an act of political advocacy would relate to more sophisticated accounts such as, for instance, the interpretation of the common good defended in Eric Beerbohm & Ryan W. Davis, ‘The Common Good: a Buck-Passing Account,’ The Journal of Political Philosophy 25 (2017), pp. e60-e79. On this view, what the polity should do is understood as a state of affairs that members of the polity have reason to act together in bringing about and thus as something that is relative to the members and yet not reducible to a mere function of their own individual preferences.
This position, understood in a very broad sense, can be attributed to a variety of views in the proceduralist tradition in democratic theory. For a classic statement, see Robert A. Dahl, Democracy and its Critics (New Haven: Yale University Press, 1989).
Thomas Christiano, ‘Voting and Democracy,’ Canadian Journal of Philosophy 25 (1995), pp. 404–410.
Timothy Williamson, ‘Knowing and Asserting,’ Philosophical Review 105 (1996), pp. 489–523.
Jennifer Lackey, ‘Norms of Assertion,’ Noûs 41 (2007), pp. 594–626. Jonathan Kvanvig, ‘Norms of Assertion,’ in J. Brown & H. Cappelen (eds.) Assertion: New Philosophical Essays (Oxford: Oxford University Press, 2011), pp. 233–250.
Duncan Pritchard, ‘Epistemic Luck, Safety, and Assertion,’ in C. Littlejohn & J. Turri (eds.) Epistemic Norms: New Essays on Action, Belief, and Assertion (Oxford: Oxford University Press, 2014), pp. 155–172.
More stringent norms would yield very demanding requirements with, potentially, momentous implications for the ethics of voting. For instance, suppose that the epistemic responsibilities of voters were to be understood in terms of a norm of assertion such as knowledge. Such a norm would probably entail the striking implication that most votes are unwarranted and, thus, that almost all voters violate their epistemic responsibilities. Most notably, if votes are truth-apt statements concerning the course of political action to be undertaken, on this norm all those who vote for the ‘wrong’ option would fail to meet the appropriate norm of voting and thus be in violation of their epistemic responsibilities. Hence, a knowledge norm would rule out as unwarranted even political decision-making inputs that are predicated upon reliable but not necessarily correct considerations (on this see Fabienne Peter, ‘Epistemic Norms of Political Deliberation,’ pp. 399–402). While the problematic nature of this implication is not by itself a reason to reject its validity, at least not without having conducted a more extensive inquiry, such a task exceeds the purposes of this paper. At least for the purposes of this work I think it best to proceed with parsimony and work with more permissive epistemic norms, as this is enough to show how voting yields significant epistemic responsibilities. I thank an anonymous referee for pressing me on this issue.
The idea that those who perform an epistemic agency without taking care to have the appropriate agential capacities are negligent is well explained in Ernest Sosa, Judgment and Agency (Oxford: Oxford University Press, 2015), pp. 69–73.
Sanford Goldberg, Assertion. On the Philosophical Significance of Assertoric Speech, pp. 175–176.
Among others, see David Estlund, Democratic Authority and Fabienne Peter, ‘The Grounds of Political Legitimacy,’ Journal of the American Philosophical Association 6 (2020), pp. 372–390.
For an autonomy-based view that seems compatible with these remarks, see Adam Lovett, ‘Democratic Autonomy and the Shortcomings of Citizens,’ Journal of Moral Philosophy 18 (2020), pp. 363–386. I offer an accountability-based view in unpublished material.