Re-conceptualizing Silence

To open up this possibility, we need a new conceptualization of silence that moves away from the notion of silence-as-absence to consider alternative understandings of silence as signal and as practice (Rollo Reference Rollo2019).

Classically conceived as the order of logos, the political order has come down to us as the order of speech, the signifier of meaning, or more broadly, as the order of voice, the signifier of presence. As the effect of voice, political subjecthood is deemed to require “vocal emission” “headed for speech” (Cavarero Reference Cavarero2005, 13). Where this is missing, three related absences are postulated: of the subject of politics, the speaking subject; of political voice as that which gives us a presence/say in politics; and of expression, signification, and communication, as that which makes speech political.

The binary economy of the symbolical order of politics is now easily articulated: on one hand, voice, action, presence; on the other hand, silence, inaction, absence. This is, of course, a problematic binary, leading to the de-politicization, privatization, and marginalization of several political behaviors that cease to register as “political” in representation’s vocal register—for example, abstention, exit, even the expression of popular voice through institutionalized channels when no electoral choice is made (e.g., the blank vote). It is also a disciplining structure: where chosen rather than forced, political silence is said to imply renunciation of the right to a voice (“didn’t vote, cannot contest”).

The enduring legacy of the binary manifests itself in the influential “exit-voice-loyalty” framework (Hirschman Reference Hirschman1970). It postulates two potential responses when we don’t like an institution or community: we can leave it (“exit”) or we can advocate change (“voice”). The exercise of voice is deemed inherently political (Hirschman Reference Hirschman1970, 16), and it is contrasted with exit, characterized as “essentially a private and also typically a silent decision and activity” (Reference Hirschman1993, 194, my emphasis).

Contrary to what the voice-silence binary implies, however, the relationship between voice, speech, and silence is anything but simple.

Voice commonly refers to acts of self-expression. The term is used capaciously, to capture a whole range of human communication, from speech—that is, a vocalization that signifies something (most notably, accounts, articulations, arguments)—to action. People are sometimes said to express their position by doing things like marching together down a street. Their marching may be quiet, like the 1917 Negro Protest Silent Parade, and still be said to express voice. This is not simply because the action may involve verbal elements (e.g., written slogans or phrases), but because it is deemed the potential object of linguistic coding, semantic meaning, or political speech.

One should beware, however, of conflating embodied action with voice. This reflects the dominance of the voice model of politics and does not sufficiently take heed of the distinctiveness of bodily agency, or even silent bodily agency (Rollo Reference Rollo2019). Under the voice model, either silence is like speech or it is not relevant. But this is reductive: it reproduces the authority and hubris of voice, while downgrading silence to its surrogate, disempowered form. It makes silence’s political influence dependent on its translation into voice through representation. While this might sometimes be the case, it is certainly not always so. Deliberate silences often enable agency and empowerment precisely because they protect subjects from assuming a legible position (e.g., refusal to state one’s position before a negotiation; resistance to declaring one’s age to a prospective employer; expression of indigenous sovereignty as refusal to speak). More significantly still, silence opens up to alternative forms of democratic engagement that have productively disruptive potential precisely because they are not like speech (Rollo Reference Rollo2017). Take for instance the Women in Black movement. When collaboratively performing silence in public, its members subvert the gendered figure of the speaking subject and the notion of sovereign voice—a “saying” without listening that is a form of domination and control (Fiumara Reference Fiumara1990, 2). Casting themselves as apprentices of listening, rather than masters of discourse, the Women in Black destabilize the extant political order by staging maieutic silence as a receptive practice giving space to and eliciting the inexpressible (Fiumara Reference Fiumara1990, 98). Their mourning silence is an agonistic bodily sound rather than a nonverbal voicing (Athanasiou Reference Athanasiou2017). To “read it” otherwise forecloses its prefigurative politics.

Now to the relation between speech and silence. Usually viewed as in an “either or relationship,” speech and silence can co-occur. They can be simultaneously present, with one carrying traces of the other. As Foucault reminds us: “there is no binary division to be made between what one says and what one does not say” (1990, 27). Silence can be a form of control within discourse. There is silencing speech just as there are speaking silences—and these can speak multivocally.

The multiple functions and uses of silence are generally recognized (e.g., Jaworski Reference Jaworski1993), and yet two trends persist in the treatment of silence. First, it continues to be presented as internally undifferentiated—with the plural “silences” sliding into one essentialist and totalizing “silence.” Second, attempts to “decode” the nature or meaning of silence tend to fall into a binary approach, taking silence as either linking or separating, active or inactive, assenting or dissenting (e.g., Pettit Reference Pettit and Pettit2002).

One way to avoid these pitfalls is to treat silence as a signal—namely a nonlinguistic form of communication whose meaning is not assumed but interpreted in the broader context of interaction. This approach to silence is strengthened by distinguishing between two ways in which silence may be produced: by passive omission and active commission (Brito Vieira Reference Mónica2019; Scott Reference Scott2017, 5).

Silence as an act of passive omission is a not-doing something, or a negative decision, resulting in a neglect, failure, or inability to act. This inability is no muteness, but a form of locutionary (self-)exclusion motivated by a resignation to power or the belief that no one will listen. An internalized sense of powerlessness, or habitus of resignation, may develop, which translates into a deep-seated dispositional resistance to action, as is typical of states of alienation, experienced at an embodied level as apathy or inertia (McNay Reference McNay, Gonzalez-Arnal, Jagger and Lennon2012). This is how silence is commonly understood, as a silencing or a mark of absence.

Silence as an act of passive omission contrasts with silence as an act of commission—or a deliberate withholding of speech. This silence is a doing: it involves doing a non-something out of a positive decision not to do. It might function like speech, namely, as a non-locutionary speech act whose illocutionary force and perlocutionary effect are gauged against context, and in reference to speech, as “a marked absence occurring in place of an expected [speech] presence.” (Acheson Reference Acheson2008, 537). It can contribute to flows of communication or be used to interrupt and subvert them.

It is often stressed that though silence can act linguistically, as a deliberate withholding of speech, it is the most ambiguous of the linguistic forms (Gray Reference Gray2019). This ambiguity would leave silence particularly vulnerable to neglect and/or misinterpretation—whether unintentional or strategic—and consequently also to misrepresentation. Undoubtedly, specific difficulties surround the interpretation of silence—for example, the difficulty of distinguishing between silence as no sign and silence as sign and for silence as sign to convey propositions or arguments. But the implied contrast between the “transparency” of speech and the “opaqueness” of silence, let alone that between vote and silence, is overdrawn. Speech is not transparent—even for the speaker. Votes don’t speak on their own. They require interpretation. To insist on the “opaqueness” of silence is to place the burden of communication entirely on silence and to discount what else might explain its failed, refused, or distorted uptake.

Moreover, not all silence is ambiguous. Where directives are explicit and signifying conventions strong, the meaning of silence can be distinctive. Think of a wedding ceremony, where the congregation is asked if anyone knows of a lawful impediment to marriage, followed by a “speak now or forever hold your peace.” If one hears silence as the response, it can be taken as an illocution of consent. But the meaning(s) of actively produced silences can also be gathered more generally from context and anticipated speech that does not occur. Take, for instance, the expectations attached to speech-based institutions, such as parliament (from the French word, parler, meaning to discuss things). Take then a party, Sinn Féin, which, since 1917, has adopted a controversial abstentionist policy, whereby it refuses to take seats for which it is elected in Westminster. Sinn Féin’s refusal to lend its voice to adopt or reject policy in the UK parliament is a commissive act—a staged resistance to sit and speak in a foreign legislature, which Sinn Féin sees as an instrument of oppression over the Irish people. Silence at Westminster is an absent presence signified and framed by the empty seat. Non-locutionary acts of considerable perlocutionary force, because “uttered” in institutions where speech is expected, silences of commission such as this show that those who refuse to speak may use their silence to delegitimize extant political configurations and their language games.

Analytically distinct, the two silences—omissive and commissive—are not as sharply divided in practice. Intentional agency is often a response to conditions eluding one’s control. For instance, refusal to speak can be one’s form of resistance to the silencing terms of dominant discourse, as for the person not ticking a box on a form presenting gender as binary. Silence can—and does—often contain elements of both, omission and commission, and these two silences can—and do—fold into one another. Thus, when citizens resent the limited choice of alternatives in an election, they may use not voting as a way of protest. Over time, however, should the situation remain the same, what was once silent resistance can slip into the silence of resignation or neglect.

Another way in which the two silences are related is as signals bearing nonlinguistic communicative content. This may not be immediately apparent. Communication requires publicity and not even all commissive silences seek it. The success of insubordinate silences—of refusal and resistance, for instance—often depends on avoiding exposure (e.g., by not telling, not declaring, not reporting, etc.; see Jungkunz Reference Jungkunz2012). But as deliberate withholdings of speech, commissive silences are rightly taken as signs carrying a variety of meaning. Silences as acts of omission are, however, suffered rather than chosen. Hence, they are frequently deemed no signs. However, I submit that as indicators of socially inflected inabilities (rather than mere muteness) they are signs, and they signal powerfully the interdependence between the experience of being listened to and the ability to express oneself.

This is worth stressing, if only because the production of silence may have little to do with the passivity or activity of the silent. Agency is a socially distributed phenomenon. Whether and to what extent others listen, understand and respond to us determines whether our agency is enabled or frustrated.

This underpins the fragility of silence signals and the oppressive dimension of silencing. Exclusion can, of course, by itself be silencing—think, for instance, of unequal access to spaces in which one might be heard (e.g., media platforms). But silencing can also coexist with voice and can result from perlocutionary frustration (speaking but failing to achieve intended effects) and illocutionary disablement (failing to perform the action one intends with one’s utterance) (Langton Reference Langton1993, 315). For instance, the no-means-yes belief prevents women from refusing sexual contact successfully.

These types of silencing are common within our representative systems, and sometimes glaringly so. According to a 2016 study by Pew, 82% of Americans favor paid maternity leave. The policy is now gathering bipartisan support, but for years system biases (including the overrepresentation of business interests and the underrepresentation of women in Congress) kept it off the policy agenda, as if the public did not really care for it. To understand silence, we need not simply look to signals that are not emitted or only weakly so, but to why some voices—however eloquent—do not register.

Equipped with this new understanding of silence, I now explore whether silence can be represented in democratic politics and, if so, how?

Re-Presenting Silence

Voice—I have argued—is integral to how we think of political representation. It is both normatively presupposed by and expected from it. Voice-based communication, understood as an array of acts with linguistic content (expressions, assertions, arguments, opinions, preferences), is traditionally placed at the input and output ends of representational relationships.

This is especially salient in the so-called “transmission belt” view of representation, which treats representation as the medium through which preferences, interests, and values are communicated from a place where they are present (civil society, public opinion) to a place from which they would otherwise be absent (state institutions, policy) (Manza and Cook Reference Manza and Cook2002, 639). From this perspective, representation occurs when communicative responsiveness is observed: that is, where the expressed preexisting preferences of the represented register in and direct the positions and policy decisions (the “authoritative voice”) of their representatives.

As a voice-based model of representation, the “transmission belt” model reads silence as an absence—that is, as the refusal or denial of political self-expression, undercutting representation qua vocal register. Silence gets in the way of representation because it hinders constituency self-organization—by making group coordination impractical, if not impossible—as well as preference formation, expression, and advancement. Representation being synonymous with responsiveness to voice, to represent a silent constituency is a non sequitur: “In a meaningful democracy, the people’s voice must be loud and clear: clear so that policy makers understand citizen concerns and loud so that they have an incentive to pay attention to what is said” (Verba, Schlozman, and Brady Reference Verba, Schlozman and Brady1995, 1). Should voice be replaced with silence, representation would go blank.

The specter of “blankness” activates fears of domination and displaced involvement. Responsiveness to constituents’ interests, I submit, is a key normative commitment of representation, and usurpation is representation’s constitutive attribute. Hence, it is fitting that domination and displaced involvement should be the two normative concerns against which I will be assessing the representation of silence in the remainder of this article. Closely interdependent, they have two different criteria in mind: in the case of domination, that political action is responsive to citizens and their interests (Brito Vieira Reference Mónica and Vieira2017; Fossen Reference Fossen2019); in the case of displaced involvement, that whatever political action is taken it happens (also) through the represented’s involvement or mobilization (Disch Reference Disch2011; Markell Reference Markell2008).

Bearing in mind this distinction, it is easily seen that where representation is taken to mean responsiveness, concerns about displacing involvement are demoted to a secondary level. Concerns about domination remain paramount though, because where too many are silent, non-domination is dependent on voting citizens’ interests being representative of the interests of the silent. This is why the assumption that silence means either indifference or assent is necessary for democratic pluralists to dismiss the necessity of wider citizen involvement to avoid domination through minority rule. Yet, given the structural relationship between economic and political power, such concerns are not so easily dismissed. Empirical studies show that “vocal” citizens are unlike most “silent” counterparts in their socioeconomic status, and this difference is very likely to ground a difference in the interests seeking representation (Scholzman, Verba, and Brady Reference Schlozman, Verba and Brady2010; Verba, Scholzman, and Brady Reference Verba, Schlozman and Brady1995).

But the more fundamental reason why silence is deemed unrepresentable in a transmission-belt model of representation is the model’s overstatement of the communicative quality of voting and the unnecessarily high standard this sets on the possibility of representing silence.

This model regards votes as sufficient to inform and constrain elected representatives to voter preferences. Yet votes in themselves are very poor at telling representatives what voters want, which is why representatives seeking to establish and shape what constituents will reward in the next election gather information by other means: polls, focus groups, etc. That we should continue to refer to the vote as “voice,” when it effectively replaces voice by “compressing multivocality into a single communicative act, designed to exclude qualification or nuance from the expression of preferences,” attests to the normative and ideological appeal of voice models (Coleman Reference Coleman2013, 12). However, just like the silences of omission and commission, votes too are relatively ambiguous. They only explicitly refer to a party or candidate, while being silent about specific policies or pieces of legislation. Votes may not even reflect a positive preference for the “chosen” party or candidate, but apathetic habituation, hegemony, social pressure, resignation to the paucity of electoral choice (i.e., hold your nose and pick), or simply reflect tactical voting, protest voting, or negative partisanship (anyone but x). In sum, votes are signals, and as signals they are far from establishing a mandate or a positive representative relationship.

Enfranchising Silence

Once one brings both vote and silence under the category of “signal,” the difference between the two becomes one of degree, rather than kind, as do the challenges they pose to representation. Their treatment as “signals” also highlights that the purported weakness of silence-as-signal owes as much to the ambiguity of silence as to the cost-free manner in which “silence” can be ignored by a representative system designed to respond to those who are more likely to vote and/or more likely to voice, by organizing and contacting representatives directly to strengthen signals about the content and intensity of their views (Bartels Reference Bartels2016).

Hence, the question follows: can silent signals be enfranchised in such a system? With its focus on vote, the transmission-belt model fails to register silent signals, most notably those of abstention. For silence to pull levers within it, the model needs to integrate silence through the only register it recognizes: vote. The blank and the compulsory vote are two ways in which this might be secured. I take them in turn.

The Blank Vote

One way of enfranchising silence is to separate out silence as an act of commission and channel it into the electoral system by institutionalizing the blank vote on the ballot. Giving voters a “none of the above” option might empower what would otherwise be a relatively powerless exit into abstention by rendering silence’s communicative content explicit and allowing protest to register within the electoral system. Furthermore, in showing the level of political discontent, the blank vote may increase the representative system’s proactive responsiveness in order to forestall exit and incentivize new political forces to emerge, seeking out those who feel unrepresented.

Implicitly relying on the distinction between commission and omission, campaigners for the blank vote compare it positively with abstention. Abstention is easily depicted as political inaction and demeaned as free-riding on the representative system by lazy, disengaged, or apathetic voters. But the blank vote should not be discounted as political inaction, nor regarded as action whose communicative content it is up to the representative system to interpret. The blank vote is not a silent vote, as it is also sometimes deceptively known. One can see why it may be deemed “silent” though. Where voting is regarded as the key form of political voice, and political voice is equated with a positive choice for one of the political forces competing for our representation, the blank vote emerges as silent because it refuses to speak in the ways anticipated. The blank vote is a vote not to vote, so as not to exercise a voice binding one to any of the representative relations on offer. But precisely because it voices disidentification with the options provided—and perhaps even with the system providing them—its expressive nature is conspicuous.

Yet, even if the blank vote helps the silence of commission to be separated out from the silence of omission, and even if it enables dissatisfaction with the representative system to be signaled clearly, to count as empowerment it must not simply be expressive but linked to consequences. Hence, the blank vote, which is normally deemed invalid, must be formally admitted, reported, and taken into account when determining vote shares and the allocation of seats. Real representation for blank votes is a growing demand. After the financial crisis, for instance, “none of the above” parties formed in Spain (Escaños en Blanco/Ciudadanos en Blanco), campaigning for changes to the electoral law linking blank votes to electoral results. They also proposed to symbolically represent those ignored or frustrated by the representative system through the election of blank/empty seats (which they achieved, at local level, in Catalonia), marking citizen voice exclusion beyond elections.

If the blank vote ceases to be simply expressive or its vote share increases, then difficult questions arise: from questions about trade-offs between voting blank and risking electing one’s least favorite option to what happens in the unlikely scenario that the blank vote gets the greatest share of votes or becomes the main opposition party. The second round of France’s 2017 presidential election indicated the risks presented by such scenarios, with an estimated 9% of votes cast being blank or spoiled, but only counting towards the turnout. Elsewhere, as in Colombia, however, the blank vote can be of great consequence: a win for the blank vote in two-round elections triggers a re-election, where none of the candidates who initially ran is allowed to participate. This can wipe out the entire political status quo. Yet the gamble is high-risk: should the blank vote come second, it means a win for the first candidate (in congressional elections); should it come first, the “blank vote” camp needs to create lists and choose its own candidate (in presidential elections). Yet this presents a challenge: the blank vote is a negative vote signaling dissatisfaction with the choices available, but not necessarily pointing towards choices of its own.

Radical scenarios aside, the question of whether the blank vote can empower the silence of abstention by disambiguating it calls for further consideration. The blank vote is more determinate than abstention in that abstainers might not think all parties are equally bad. If, for instance, the incumbent party is likely to be re-elected, abstention can express satisfaction, or, where the electoral result is open or perceived as such, abstention can express the understanding that all options are equally acceptable. By contrast the “none of the above” vote is a protest vote, showing frustration with the available options.

The “none of the above” vote can thus partly disambiguate abstention and give more teeth to systemic disaffection, but it cannot represent all abstainers, not even amongst the disaffected. Survey data shows that abstainers are more likely to be dissatisfied with the state of democracy than voters. Disadvantage is a major predictor of abstention: the more socially and economically marginalized, the less likely one is to vote or to participate in other ways (Brennan and Hill Reference Brennan and Hill2014, 146–47). The profile of blank voters differs, however, from the profile of abstainers: blank voters are well-educated and politically sophisticated (Superti Reference Superti2015). It is thus unlikely that the two constituencies should converge. They may, of course, occasionally align over protest about problems cutting across the representative system—for example, perceived widespread corruption. While neither might feel represented by the options on the ballot, they are unlikely to be represented by the same alternative option, should this emerge from the pressure of the blank vote. Hence, the blank vote might clarify, confer visibility, and demand response to some of the silence generated by the representative system. But if incentivized to seek out blank voters, the system might reinforce rather than erase preexisting biases in favor of the educated and the politically knowledgeable.

Compulsory VotingFootnote ¹

This skewing effect might seem easily solved by integrating abstention into the electoral system through compulsory voting. After all, domination haunts those silently exiting the system: nonvoters are typically among the less well-off and, in renouncing their modicum of voting power, they risk having their interests overridden by those voting. Could compelling them to vote defend them from domination through indifference to unwarranted interpretation or even manipulation of the weak signals provided by their silence?

Before addressing these questions, a preliminary note is necessary. Compulsory voting is sometimes said to compel voice by denying the possibility of exit. However, while compulsory voting seeks to draw positive choices from voters, it need not violate the right to silence entirely. All that compulsory voting requires is turnout. Once voters turn up, they are free to spoil the vote, abstain, or choose the “none of the above option,” if given.

What changes then? If voters are forced to go to the polls, but do not express a partisan preference, the silence of abstention would seem to become commissive: deliberate, more informative, and harder to dismiss. But this is too quick. A failure to express a preference can then still be omissive, not a deliberate expression of dissatisfaction with the options but a reflection of voter limitations in accessing or processing electoral information (Driscoll and Nelson Reference Driscoll and Nelson2014). It is sometimes claimed that even a registered preference not to voice conveys “vitally important information to politicians, and other voters,” since “it communicates that there is a constituency whose votes are up for grabs and that unavailable alternatives need forming” (Brennan and Hill Reference Brennan and Hill2014, 142). Yet given that the non-registration of preference may be omissive or commissive, it remains uncertain how much more information one is effectively getting. This is not to say that things are not further disambiguated. In place of abstention signaling approval, we might now have a formal vote. Some abstention might turn into explicit protest, for instance, through intentional spoiling of the vote or “none of the above” votes (if the option exists). But invalid votes—blank or accidently spoilt—are still likely to be cast by the politically unknowledgeable, uninterested, untrusting, and disaffected (Singh Reference Singh2019). Their votes will mix elements of commission and omission, and where they are meant as protest, this might now—confusingly—include protest against compulsory voting itself.

There are, however, potential benefits in terms of domination and even involvement. As citizens with lower incomes and education levels are required to vote in higher numbers, inequities in turnout are ironed out, and parties may have incentives to reach out and invest in their political socialization. Additionally, as the voting population becomes more representative of the general population, the probability of the electoral signal being distorted to the advantage of the well-off is likely to decrease. But the benefits may prove elusive. If many amongst the forced voters lack interest, knowledge, and sophistication, and if political sophistication is a precondition for choosing the party closest to one’s political preferences and holding it to account, then forcing non-sophisticated voters to cast a vote may undermine the representative function of elections and leave them as unrepresented as before (Jakee and Sun 2016).

It could be objected that political sophistication is endogenous and will increase with the incentives created by repeated mandatory electoral participation. Yet the evidence on this is inconclusive (Carreras Reference Carreras2016). What is more, compulsory voting has been found to have no effect on accountability and a negative effect on preference correspondence (Dassonneville, Hooghe, and Miller Reference Dassonneville, Hooghe and Miller2017). This finding may have less to do with voters’ choice than with the supply side of representation. Compulsory voting increases the adoption of programmatic vote-seeking strategies, involving non-contingent policy bundles aimed at large groups rather than subgroups (Singh Reference Singh2019). It may also encourage a focus on the politically sophisticated—seen as more likely to switch their vote in response to policy changes—and to persuasion being pursued over mobilization (Singh Reference Singh2019). Thus for most of those previously exiting silently into abstention who are now forced to turn out, compulsory voting may produce some indirect responsiveness by preventing wider distortions of the electoral signal, but not necessarily foster the formation of and mobilization around their interests. In sum, while domination may decrease, involvement may remain largely untouched.

This takes me to a final and more fundamental objection to compulsory voting as a way to avoid powerless exit into silence: in order for voting to empower, the option of exit must be real and effective. Ceteris paribus, silently exiting to another party offers a stronger signal than silent exit to abstention, since the latter reduces one’s customary party’s vote share without increasing the voter share of a competing party. It also offers—especially where the party exited in protest becomes vulnerable to electoral competitors—a better chance that the party will reach out to seek information and enhance responsiveness by changing policy. But protest voting is demanding, strategic, and risky. Hence, abstention may prove a more palatable if less effective strategy for many dissatisfied voters. What is more, for voters on either extreme of the political spectrum, with a strong distaste for centrist parties but wanting to signal dissatisfaction to their preferred party, or for voters wanting to send a signal of dissatisfaction to the entire political-party system, abstention may well be the only adequate option.

A compulsory system will push for the normalization of oppositional silence: invalid votes still contribute to voter turnout, while poll abstention adversely affects it and—potentially—the legitimacy of an election. When “internalized,” the act whereby one suspends one’s cooperation with the political system (by refusing to register a choice or spoiling the vote) is also the act whereby one complies with it (by going to the polls). Refusal to choose lends itself to being interpreted as a failure to act or as a reformist act making demands from within the representative system rather than repudiating its authority and calling for its serious restructuring.

The right to exit the formal representative system constitutes thus not only a freedom from but also a freedom to make politics by removing oneself from politics as is. This removal need not denote unreflective political inaction just as voting need not denote reflective participation. It can be a reflexive reaction to concerns of private and general interest, such as a broadly uncompetitive and unrepresentative political system in which voting might have ceased to be empowering for many.

To conclude, the transmission-belt model of representation is constitutively incapable of accommodating silence, especially omissive silence. Because it makes presence primary and representation derivative, it formulates the problem posited by silence as how to reflect within representative institutions a presence that may otherwise be obscured by abstention. What it misses is that the question is not simply one of disambiguating a preexisting presence, but how to produce that presence (i.e., preferences, interests, identities, constituencies) in the first instance (Disch Reference Disch2011). Some problems of information and ambiguity affecting commissive silence may be mitigated by implementing positive abstention mechanisms, such as the blank vote, and providing political clout to “easy, low-cost and actionable” options for silent exit (Warren Reference Warren2011, 699), such as protest voting. But to mandate a 100% voter turnout fails to produce the incentives to constitute and mobilize constituencies amongst those trapped in omissive silence while potentially forcing more distorting signs into the representative system.

Enacting Silence

With its normative and ideological emphasis on representing voice through the ballot, the transmission-belt model of representation struggles with silence: it overstates the ambiguity of silent signals while understating the ambiguity of voting in asserting a coherent mandate. What the model misses is that just as silence is never just a cause but also an effect of representative exclusions, voice is not just echoed in the representative system but primarily enacted and spoken for by it.

In this respect, the issues surrounding the representation of silence are not so different from those surrounding the representation of expressed votes or indeed political representation tout court. However, focusing on silence brings these issues to the fore while it also uncovers silencing effects owed to the discursive bias in representation. In this section, I show how the latter manifests itself in constructivist views of representation. While I argue that they offer a more promising way to empower citizens from their silent positions, I reconstruct and expand criteria of legitimate representation to more effectively ward off dangers of domination and displaced involvement.

On a constructivist view, preferences, interests and identities are constituted through processes of representation rather than objects of representation prior to such processes (Saward Reference Saward2010; see also Brito Vieira and Runciman Reference Mónica and Runciman2008; Disch Reference Disch2011; Reference Disch2015; Montanaro, Reference Montanaro2012). What is true of preferences, interests, and identities is also true of the groups sustaining them. Groups are not simply given, prior to or independently of, representation, but (at least partly) brought about by being represented as such (Brito Vieira Reference Mónica2015). In bringing group members into a meaningful relationship with one another and with out-groups, representative claims generate a distinctive self- and collective understanding of who they are and what they want. This forms them into and mobilizes them as a group. As Disch puts it, political representation “aims, then, not to reproduce a state of affairs but to produce an effect: to call forth a constituency by depicting it as a collective with a shared aim” (Reference Disch2011, 207).

If representation is constitutive then there is necessarily an element of usurpation involved in it (Brito Vieira Reference Mónica2015). This is not simply because authority to enact a group must be seized before it can be gained, as the authorizing subject—the group—might be non-existing, non-defined, or non-capable prior to its representation. It is also, more generally, because a representative claim always appropriates the voice—or, indeed, the silence—of the represented.

To speak of appropriated voice is not to posit an original—immediate, authentic, self-created—voice that is then appropriated. It is rather to argue that any voice arising from representation is produced—a precarious, contested appropriation, calling for recognition and validation by the represented. It would thus be a mistake to equate this constitutive usurpation—a necessary and productive element of representation, inviting the represented into a relationship—with displaced involvement, the deviant pathology identified by my normative criteria for assessing representation. While one can evolve into the other, the latter occurs when representation loses its necessary relational quality and slides into a mere speaking and acting in the place of.

This danger is more pervasive than is acknowledged (Fossen Reference Fossen2019, 834). For while it may seem that speaking about—that is, representing or portraying the group as being like this or that—raises fewer concerns of displaced involvement than speaking and acting for, in practice, the distinction is hard to draw, with both types of “speaking” intertwining closely and widely in the construction of representative claims (Alcoff Reference Alcoff1991, 9). Furthermore, to speak about is to create for the represented a public self in the presence of others. This affects how audiences engage the represented and provides the latter with a means of self-understanding and self-description that may empower or indeed disempower reflexive self-constitution and effective talk-back.

The identification of domination as my second criterion for assessing the legitimacy of representation may seem inadequate—if not wholly contradictory—within a constructivist view of representation (Disch Reference Disch2011). I defined domination as lack of responsiveness to the represented and their interests. Responsiveness, authorization, and accountability constitute the normative triumvirate of the transmission-belt model of representation, with its focus on questions of control rather than involvement. Responsiveness, in particular, is its nub, as the model hinges on representatives being responsive to the represented and on the represented being able to hold representatives accountable for what the latter say or do in their name (Dahl Reference Dahl1972, 1). This, in turn, requires previous authorization, insofar as the most fundamental form of accountability—“the possibility that authorization may be withdrawn or contested” (Schweber Reference Schweber2016, 393)—is taken to depend on it.

Taking a constructivist position, I submit, involves redefining rather than abandoning these criteria. In the transmission-belt model responsiveness implies correspondence between citizens’ preferences—given prior to, and independently of representation—and representatives’ actions or policies (Sabl Reference Sabl2015). But if voters’ preferences are endogenous to political representation, responsiveness to existing preferences is a circular criterion of representation, assessing representatives by criteria partly of their own making (Disch Reference Disch2011).

While the objection holds, it does not follow we can do without the criterion of responsiveness, if taken in the reconstructed sense I advance here. Domination, I argue, occurs when the normative (not ontological and temporal, as is assumed in the transmission-belt model) priority of the represented and their interests is violated. To represent is to act on behalf of the represented, and in their name—not the representative’s. This does not imply that the interests of the represented precede representation or that the representative must follow the represented’s views of their interests. It does imply, however, that representatives must orient themselves to the represented and their interests, and that where they deviate from what the represented believe to be in their interest, they must engage the challenge and give reasons to the represented to broker their interests in what is proposed (Brito Vieira Reference Mónica and Vieira2017, 28–29; Pitkin Reference Pitkin1967, 213; cf. Fossen Reference Fossen2019, 834–35). This requirement of “response” demands an “openness to being interpellated by another’s address” (interpellation being here irreducible to voice) (Athanasiou Reference Athanasiou2017, 244), and it is predicated on the acknowledgement of the contestability of claims to represent.

By the same token, authorization, understood as process, has an important place in constructivist accounts. If representation is “not just there, a thing” (Saward Reference Saward2010, 13), but a relationship of giving-voice-to whereby represented and representative are mutually constituted, the emphasis must move, however, from authorization as the act establishing representation to uptake and acceptance by those subject to representative claims (Saward Reference Saward2010, 151–53). Authorizing becomes thus a retrospective and ongoing process, so that even where authorization is formally or informally pre-given, what remains critical from a constructivist perspective is that the procurement of such authorization “does not render null and void all the attendant problems of speaking for others” (Alcoff Reference Alcoff1991, 10). Equally, and of salience to my argument, “the power to confer such authorization, and to have power over the designated representative,” may be “rarely present in instances where one is spoken for” (1991, 10), demonstrating the problems of inferring an assenting constituency from silence.Footnote ²

Before I examine claims to represent silent constituencies using these criteria, it is important to stress the discursive bias of the particular constructivist understanding of representation I adopt in my analysis of such claims, representation as a matter of claims-making (Saward Reference Saward2010). In the claims-making view, claims are constitutive of representation—that is, that someone is a representative, or a constituent, is a consequence of claims-making. Makers of representations “call forth” constituencies by making claims about themselves and their constituencies and offering these claims to the would-be audiences or constituencies for approval or disapproval. In what follows, my analysis focuses on two distinct aspects of claims-making: claims about silent constituencies, casting them in particular ways, and claims that someone should be regarded as a representative of a silent constituency. These claims can, of course, be separated: one can make representations of someone or something without claiming to be their representative. But since constituency and representative are mutually constituted, how one constitutes a constituency and how one constitutes oneself as its representative are closely interdependent.

Adopting a claims-making framework is not without challenges given the framework’s discursive bias.Footnote ³ To focus on claims is to focus on speech acts as the way to represent and create political presence. Here, I want to qualify this assumption. Silence too can be a mode of representation rather than just something to be represented. As demonstrated by Sinn Féin, it is possible to make representational claims silently. That a representational claim need not be voiced to be articulated is, again, shown by athletes refusing to sing national anthems because they do not believe them to represent their communities. But their silence might still look like a signal, or way of speaking, reliant on language for the translation of representations into propositions or claims which constituencies may approve or disapprove. There is, however, a more distinctive way in which silence can act as a mode of representation: where it refuses to act like speech, and explicitly departs from claim-centered modes of representation.

To appreciate this possibility properly, it is important to note that the notion of representative claim presupposes a particular political subject—a public speaker, a rhetor—and a particular form of political subjectivity or agency—discursive. When feminist groups like the Women in Black perform silence as their mode of dissent, they contest a gendered order identifying political action and political agonism with speech, and speech, particularly public speaking, with power, citizenship, and authority. Their use of silence is thus no mere “background” to claim-making or even mere bodily speech act. It represents—not by saying, but by demonstrating—war victims and women as silenced. It refuses complicity in the discourses responsible for their depoliticization and subordination and reclaims the agency of silence to interpellate them (Athanasiou Reference Athanasiou2017, 231). It also prefigures a new politics, “breaking the linguistic and representational structure” of politics as we know it (Athanasiou Reference Athanasiou2017, 245). One might see a radical departure from representational politics in their enactment of a political subjecthood not exclusively reliant on transforming representations into claims, propositions, or demands (Athanasiou Reference Athanasiou2017). But there is another way to see it: as a critique of the treatment of the represented as essentialized objects of representation within a one-sided model of representational politics focused on claiming, but hardly suited for genuine responsiveness. Yet listening out for and being responsive to citizen silences pregnant with political meaning is one, perhaps the main, task of representation.

I can now turn to the analysis of a specific and widespread linguistic claim about silence: the claim to represent an alleged “silent majority.”Footnote ⁴ Silent majority claims have a long history, starting with Nixon, who famously used it to co-opt a socially conservative electorate in silencing anti-war protesters and the 1960s civil rights movement. Thus, in some contexts, the phrase works as a shorthand for a racialized identity embraced by working-class whites, especially Republican voters. But claims to represent the silent majority have been made on both the right and the left of the political spectrum, and the claim’s identity has been both produced and re-signified in the process. The claim has been used to mobilize racial minorities around exclusion from majority rule; for example, during the 1960s, Black Power groups presented themselves as the real silenced majority in major cities like Chicago. Conversely, it was used to resist the protagonism of these black militants: the NAACP leader, Roy Wilkins, urged the true black majority, “the silent middle,” to resist the capture of the African American political agenda by black militants prioritizing racial issues over crime (Fortner Reference Fortner2015). After the 2008 global financial crisis deepened inequality, Ed Miliband, UK Labour party leader, positioned himself as the spokesman for a law-abiding silent majority. Populists on both right and left have followed suit (Mudde and Kaltwasser Reference Mudde and Rovira Kaltwasser2017).

There is a genealogy to be teased out of these multiple appeals to the “silent majority” and their changing contextual significance. But here I want to focus on why claims about a silent majority enjoy such political purchase in democracies and examine possible implications of these claims for representational dynamics.

The political appeal of the claim may be obvious: it raises concerns about minority tyranny. But there is more to its appeal than first meets the eye. First is the moral pull of majoritanianism in a context where representative democracy has become associated with rule of the majority. Second is a culturally specific understanding of silence that bolsters the fiction of the majority as the people, understood as an entity with one voice, and at one with itself. I take these in turn.

Majoritarianism’s moral pull hinges on two premises: the equality of voters qua speaking citizens who must have an equal “say” in decisions affecting them and the fiction that the majority is the people (Rosenblum Reference Rosenblum2008, 52). The latter, in particular, and as Tocqueville predicted, foregrounds the recurrence of the silent majority trope in democratic politics: “When [parties competing for our representation] lack [a majority] among those who have voted, they place it amongst those who have abstained from voting, and when it still happens to escape them there, they find it among those who did not have the right to vote” (Reference Tocqueville2000, 230).

In speaking about a “silent majority,” one is not, however, only representing the constituency as a numerical majority, but also, and critically, as silent. There are two distinct ways in which the silent majority can be deemed a majority and in which silence might be deemed its expression. Put in broadly Hobbesian terms, the claim might be that it is one voice that is being silenced or that a multitude of different voices are being silenced, each (individual or group voice) in its own way, adding up to a numerical majority. The distinction underwrites contrasting conceptions of the represented, as a unity without parts or as a whole comprised of parts, respectively.

Much of the clout of the “silent majority” claim lies precisely in projecting its silence as a collective silence, muting the sovereign will and thus sustaining the fiction of the people-as-one. This is a fiction fraught with dangers of holism: totalizing claims to a “silent majority” are no different from totalizing claims to the “voice of the majority.” Just as the mandated voice of the people can be constructed out of an aggregation of information-poor ballots, so can the ambiguity of silent signals be made to be heard as a unison, blank authorization.

This is not to say, however, that the claims are entirely interchangeable. There is a specificity to claims about a “silent” majority, immersed as they are in cultural norms and givens, that strengthens the fiction of the people-as-one. This is most notably the case with the understanding of silence as the “Other” of voice, with this “Other” referring to either a lack (of speech, constituting absence) or a beyond (speech, conveying full presence).

It is as a figure of absence—and thus powerlessness—that silence is first heard in claims about a “silent majority.” This absence could be blamed on the silent: a numerical majority that does not vocalize itself is likely to set off a spiraling of silence whereby active minority voices become amplified (Noelle-Neumann Reference Noelle-Neumann1993). But constituencies are formed and mobilized by framing a conflict, hence the silent majority is evoked as “silenced” (by elites, privileged and/or noisy minorities) rather than simply or even necessarily silent.

This is just a first layer of meaning, however. The white canvass is a metaphor commonly used for silence. This is because it is empty, blank, but also because it is a receptive surface, responsive to shifts in color, light, and structure. Silence is not unlike it in that it is “available for various projections of meaning because no meaning or definition can “match” the notion of silence” (Loevlie Reference Loevlie2003, 9). It thus lends itself to be used as an empty signifier, indeed, one could argue, as the empty signifier par excellence (Laclau Reference Laclau2005). This gives it a protean quality. It can open up to protect and accommodate difference. It can absorb different particular grievances and demands. It can establish an equivalence between these demands negatively through the projection of a shared antagonistic position. It can also neutralize differences between demands in the process.

To appreciate this, we need to turn from silence as an empty container, waiting to be filled, to silence as full presence. For silence is not just capable of receiving meaning. It is also capable of giving meaning, without pinning it down to particular signifieds. The enduring appeal of silence as a beyond lies precisely in its alleged capacity to provide a quasi-mythical meaning to the inexpressible: “all those experiences, insights, institutions, feelings, states ‘whereof one cannot speak’” (Loevlie Reference Loevlie2003, 11). Silence brings these into a relation of equivalence, allowing for a potential “universal identification in difference to take place” (Saldaña-Portillo Reference Saldaña-Portillo2003, 196). As this happens, silence becomes an “incommensurable universal signifier” (Laclau Reference Laclau2005, 95) forming a collective subject, which becomes hegemonic, but at the risk of any difference within silence becoming “dangerously close to losing all specificity” (Saldaña-Portillo Reference Saldaña-Portillo2003, 196).

The effect is reinforced by how we are primed to hear silence. Where speech is seen as differentiating, silence is seen as internally undifferentiated and “beyond all the distinction speech can make” (Muers Reference Muers2004, 11). As such, voice stands for distinctive individual political judgments, which, once transformed into votes, are aggregated and counted in. Silence, by contrast, is perceived as undifferentiated, or, at best, binarily discursive. This reflects itself in the tendency to resolve silence into reductive binary meanings, “yes” or “no,” assent or dissent, etc., and in the call for referenda to enable majorities purportedly silenced by normal politics to express themselves. Forced to operate in this narrow way “silence” acquires an acclamatory character, which returns a “clumped” effect (Schwartzberg Reference Schwartzberg2010, 453).

This analysis of how our culturally given understandings of silence work themselves into claims to represent an alleged “silent majority” shows how such claims may attract a diverse following by preserving their ambiguity while leaving the represented at the mercy of “elite minorities with privileged access to technologies and institutions of claim-making” (Saward Reference Saward2006, 304) and displacing involvement by “silenc[ing] the constituencies or audiences that they may in part constitute by evoking” (Saward Reference Saward2010, 55). Let me explain.

A key tenant of constructivist accounts of political representation is that representation is relational: it must presuppose and support agency and judgment on both parts, of representatives (as makers of claims) and the represented (as those challenged to take them up) alike (Pitkin Reference Pitkin1967, 155). Those claiming to represent a silent constituency can enhance their agency at the expense of the represented by playing with the equivocal meaning of silence to mobilize resentment while willfully reading a constituency out of the evoked silence in ways that foreclose agonistic engagement and only empower preconceited views. This precludes responsiveness, including responsiveness to the potentially agonistic nature of silence itself. But if domination is a concern, so is the risk of impairing the represented’s agency. To take the consequences of constructivism seriously one needs to see the represented’s capacity for action and judgment as (partly) constituted in terms of the claim made (Enroth Reference Enroth2017). Claims about a “silent majority” can be challenging in this respect because they ask constituents to judge, decide, and act under a particular self-image, rendering them as people who must be spoken for.

Another way that claims may disallow talk-back is by immunizing themselves against contestation. Claims to represent must put the representatives at some risk (Enroth Reference Enroth2017). Yet where unaccompanied by a specification of interests, policies and reforms claims about purported silent majorities leave the represented without a focus of accountability. Also, where casting the silent as dutiful, judicious citizens, expressing themselves only through proper electoral channels, they de-authorize verbal objection and seek to induce a spiral of silence whereby opponents (cast as unreasonable and as minoritarian) may be silenced or become less willing to express dissent. While attempts at foreclosing objection are part and parcel of claim-making and can only be assuaged at system level through confrontation with competing claims, the silent majority claim, especially where the reference group is large, disallows contestation in its own terms. It trades on the ambiguous and unverifiable nature of the evoked silence to resist falsification. As Tocqueville noted, it is not uncommon for those with a precarious or uncertain electoral win to trump up their legitimacy by claiming the support of a majority still “out there.”

These are risks, not inevitabilities. Claims are not good or bad. They must be assessed not in the abstract, but in their particular performance. Take for instance the meaning and effects of the descriptions under which the represented are asked to act. The claim may be the same, but these meanings and effects may change radically depending on who makes the claim and how the representative relationship is posited by the claim and enacted. Consider, for instance, a leader of Chicago Black Power seeking a seat on city government by representing blacks as the real silenced majority in the city while seeking to reinforce their agency and political efficacy by community organizing, engaging political education programs, and building multi-racial and ethnic coalitions grounded in class solidarity. The leader’s social positioning and efforts to promote relationships with—and between—the represented, which may enable the causes of silence to be identified and spoken of, and discrepancies in accounts of causes and solutions to be interpellated, averts fears of domination and displacement. By contrast, consider Donald Trump’s claim to represent the silent majority of Americans in the 2016 U.S. Presidential Election. Combined with claims such as “I am your voice. I alone can fix it. I will restore law and order,” the claim displaces the judgment of the represented by positing the speaker as the sole knowledgeable, authoritative and empowered subject. It also uses evasiveness (“fix it”) to deprive the spoken for from a focus of accountability, while positioning them as passive followers to be “championed for from afar” (Alcoff Reference Alcoff1991, 24).

In emphasizing the constitutive dimension of representation, constructivists foreground its importance in mobilizing and challenging different forms of exclusion that result in effectively silenced constituencies. Unauthorized bids for support challenging marginalized constituencies to accept them will always be necessary to bootstrap a “productive form of antagonism” out of antagonisms that are currently muted or even “largely silent, pre-rational and unarticulated.” (McNay Reference McNay, Gonzalez-Arnal, Jagger and Lennon2012, 240). However, with their almost exclusive focus on uptake and systemic reflexivity, constructivists can lack part of the normative vocabulary which is necessary to engage such claims critically.Footnote ⁵

Even under conditions of pluralism, publicity, and reflexivity, to infer a presence or passive authorization from silence (non-objection) (Runciman Reference Runciman2007; Saward Reference Saward2010, 152) is deeply problematic. It downplays the fact that the “agonistic capacity to engage contestatory activity speaks of being in a position of relative power,” “not in a position of relative powerlessness where such process of contestation may be experienced as profoundly alienating” (McNay Reference McNay, Gonzalez-Arnal, Jagger and Lennon2012). It also deemphasizes the fact that agents are empowered to judge and object not only by systemic conditions of reflexivity through competitive claim-making (Disch Reference Disch2011), but also by how they are constituted by claims to represent. Yet, where silence is taken as univocal rather than potentially agonistic, where it is taken for an absence, not a potential presence, and where the silenced are addressed as non-agents rather than as actual or potential speakers who are currently unheard (Muers Reference Muers2004, 44), their agency is undercut.

This brings me back to my extended notion of responsiveness. In arguing that representatives should show an openness to being interpellated by the silenced’s address, I am not assuming that this is because the silenced’s speech is simply found, necessarily emancipatory, or reflective of their “true” interests. Nor am I suggesting that silent constituencies should be left to speak for themselves and be listened to rather than represented. This would be to make the mistake of assuming their voice—and their capacity for voice—as given rather than a constructed position. The point is rather that one’s ability to speak is contingent on the experience of being listened to. Hence, only representation involving generative listening—a listening that pays heed to while also actively drawing out—and which speaks with and to rather than simply about and for, can be constitutive of subjects that may come to voice, challenge, and subvert (Alcoff Reference Alcoff1991, 23; Spivak Reference Spivak, Nelson and Grossberg1988).

While only representative claims-making performed in this way may activate and empower effectively silenced constituencies, one must be cautious. The representative system is responsible for extensive silencing effects, since it is constitutive of and constituted by many of the exclusions underwriting them (Devenney Reference Devenney2019). While these exclusions may have become naturalized by treating silence as assenting, prophylactic, negligible—or indeed by offloading responsibility for silence onto citizens—they have a “presence, history, and form” (Rich Reference Rich2013). This is productive of systemic institutional and structural processes (e.g., the underrepresentation of women, campaign financing privileging big money donors, etc.) whereby certain claims and claimants are allowed and taken up, others are shut down, and other claims, while taking shape, struggle to gain traction. Unless constructivists engage the notion of structural injustice (Young Reference Young2011, 52), they will remain unable to scrutinize and potentially challenge the background conditions—including constitutive representations of race, gender, and class—preventing the inclusion and empowerment of many through representation.