Free speech is a familiar concept. It is an established ideal of liberalism and democratic politics, and the subject of political debate and conflict across diverse historical and cultural contexts. Free speech has not primarily been considered, however, as a set of lived, valued, and contested practices, mediated by various linguistic, ethical, and material forms. While anthropology has not traditionally occupied itself with free speech, it has extensive tools for bringing free speech into view beyond its quality as an abstract ideal or legal category. This entry borrows theoretical perspectives, as well as ethnographic examples produced by anthropologists, to shed light on free speech within a broader comparative frame. It begins by focusing on free speech as a dynamic value or virtue, asking: what is it about ‘free’ or ‘direct’ speech that people value when they value it? Secondly, the entry casts critical light on the idea of an individual as the universal ‘free speaker’, demonstrating how collective or disaggregated subjects can also practice free speech. Thirdly, it explores the material settings, contexts, or technologies through which free speech is curtailed or realised. Finally, the entry considers the idea of ‘voice’ as signalling modes of embodiment, and auditory phenomena such as noise, sound, and silence, which are not spoken language but can inform and expand our understanding of free speech.
Freedom of speech is a core tenet of liberal political philosophy, and a criterion frequently invoked to distinguish liberal democracies from their political others. In recent years, it has become a focus of extensive and embittered debates within the US and Europe. Some fear the rise of a ‘cancel culture’, and accuse proponents of ‘safe spaces’, ‘trigger warnings’, and ‘no-platforming’ of challenging freedom of speech. The latter in turn accuse their critics of invoking freedom of speech disingenuously in order to protect established interests. These debates invoke the notion of freedom of speech to apportion blame and responsibility for political injuries, but rarely involve a sustained analysis of the notion of freedom of speech itself. However they might disagree about the rights and wrongs of specific cases, the debating parties tend—with few exceptions—to subscribe to a familiar liberal vision in which freedom of speech, within certain limits, is broadly speaking good for individuals and polities, while silencing, except in certain carefully delimited cases, is broadly speaking bad. Despite appearances, these public debates are therefore still disagreements within, rather than about, a liberal consensus. Legal scholarship and classical political philosophy have given us more formal representations of this liberal space of disagreement over free speech and its limits.
Anthropologists can make a useful intervention by putting these familiar debates about freedom of speech into a broader comparative frame. This allows us to pick out, by contrast, some of the distinctive assumptions embedded in these familiar debates—assumptions about the nature of language, about speaking subjects and the polities they inhabit. These comparative explorations tend to challenge the idea that speech can ever be ‘free’ in any simple sense. Anthropologists have demonstrated extensive determinations—from grammar to sociolinguistics—that are entailed in any speech act; they have pointed to the pervasive and sometimes productive nature of silencing in social life; and they have shown the multiple ways in which authoritative speech is entangled in and produced by controls and limitations of other kinds of expression.
Nonetheless, it remains a persistent fact that many of the people anthropologists work with value, desire, or imagine something like freedom of speech as a particular goal, and mourn, fear, or protest its absence. Anthropologists have the resources to examine the varied ways in which free speech is imagined, valued, and practiced as a lived ideal in necessarily compromised and imperfect conditions.
Semiotic ideologies, religious and secular
The most sustained anthropological explorations of the question of freedom of speech have been in relation to recent debates around religious and secular representation. The case of the ‘Danish cartoon controversy’, in which satirical representations of the prophet Mohammed sparked outrage and violence, has been paradigmatic (Asad et al. 2013; Keane 2018, Favret-Saada 2015). This controversy was a natural entry-point into the subject of freedom of speech for anthropologists because of the wealth of material in the anthropology of religion focusing on comparable disputes about the morality and politics of speaking, silencing others or staying silent oneself, or of representing and stopping others from representing. Such ‘moral questions about semiotic form’ (Keane 2007: 6), arose, for instance, in the struggles of seventeenth century Quakers in England to separate out the word of God from everyday language as a ‘thing of the flesh’ (Baumann 1984). The Quakers’ project included a wholesale repudiation of accepted forms of politeness and honorific titles as insincere words that glorify the earthly person—a practice that exposed them to violence from offended interlocutors. The moral and political stakes of speech were similarly high in missionary encounters in non-Western contexts. For instance, Webb Keane details the struggles between Calvinist missionaries and followers of marapu (Sumbanese ancestral ritual) in the Dutch East Indies (modern Indonesia) about how to address spiritual entities. The Calvinists condemned the marapu followers’ uses of traditional ritual formulae as a violation of the ‘proper’ norm of speaking sincerely to God in one’s own words. Conversely marapu followers decried a form of hubris in Calvinist prayer aimed directly from the individual to the godhead without the mediation of ancestral formulae (Keane 2007: 176-96).
Considering liberal debates and concerns over freedom of speech alongside these cases points to the deep cultural assumptions about the nature and effects of language and representation that inform all of these moral struggles over semiotic form. Such assumptions about language and meaning have been described by anthropologists as ‘language ideologies’ (Woolard & Schieffelin 1994), or more broadly ‘semiotic ideologies’ (Keane 2007). By situating liberal concerns with freedom of speech within a particular (Western, modern, liberal, secular) semiotic ideology, anthropologists have thus opened up alternative angles on recent high-profile debates, such as the Danish cartoons controversy. Keane, for instance, argues that
[t]he classic [liberal] defence of freedom of expression draws, in part, on a semiotic ideology that takes words and pictures to be vehicles for the transmission of opinion or information among otherwise autonomous and unengaged parties and the information they bear to be itself so much inert content more or less independent of the activity of representation (2009: 58).
From this perspective, Muslims offended by cartoons of the Prophet are sometimes dismissed by liberal commentators as committing a category error, and one furthermore that designates them as insufficiently ‘modern’ in their continued attachment to the transcendent power of mere images (Asad et al. 2013: xiii). But as anthropologists such as Talal Asad have been at pains to point out, liberal freedom of speech also has well-defined limits, for instance in respect of patents, copyright, or pornography. These ‘liberal’ limits point to the extent to which liberal freedom of speech is premised on and limited by notions of property and ownership—ownership of one’s texts, ideas, or body (Asad et al. 2013). One might add that hate speech laws show that modern liberals do seem quite concerned with the capacity of words to do harm, at least in some contexts (Butler 2017; Heywood 2019). Or that contemporary laws of libel or insult in places like France or Germany have a genealogy that links them to honour codes, which many sociologists imagined to be extinct in ‘modernity’ (Candea 2019, Whitman 2000). While such comparisons may occasionally sound as if they are trying to score points by showing that liberals are not as liberal as they think, at its best this work provides a more subtle understanding, rather than a mere deconstruction, of aspirations to freedom of speech, liberal or otherwise. The point, as Asad puts it, is that ‘[t]he shape that free speech takes at different times and in different places [reflects] different structures of power and subjectivity’ (2013: 29).
Virtues: courage, truth, and risk
Another related way that anthropologists can contribute to our understanding of free speech is by examining its status as a value or a virtue. In a range of ethnographic contexts—perhaps most obviously but not exclusively those labelled as ‘liberal’—people understand ‘speaking freely’ to be a virtuous practice, and view the right to be able to do so to be an important value. Anthropology has an extensive conceptual apparatus with which to analyse and compare the ways in which people think about values and virtues in work on ethics (e.g. Faubion 2001; Laidlaw 2002, 2013; Robbins 2007, 2016; Lambek 2010; Keane 2015). In fact, one of the key conceptual sources for anthropological work on ethics, Michel Foucault, also had quite a lot to say about the genealogy of ‘free speech’ as a virtue.
Foucault’s late work on classical self-cultivation investigates how people work to make themselves into particular kinds of virtuous subjects. Despite its individualist overtones, self-cultivation does not occur in isolation. It is something done in a particular cultural and historical context, and in relation to others. In his final two lecture series at the Collège de France, Foucault sought to clarify this relationship between subject and context by turning to a very specific aspect of self-cultivation in the ancient world (2010; 2011). He believed that then—as now—there was a ‘necessary other person’ involved in work on the self. These are types of people whose role it is to help us decipher and establish the ‘truth’ of our selves (teachers, doctors, psychoanalysts, jurists, policemen). In the classical world, unlike ours, however, Foucault thought that this ‘necessary other’ was not an institutionally defined position. Rather, it was predicated on the possession of a particular virtue, namely parrhêsia, translated in the title of one of the lecture series as ‘the courage of truth’. To be the right sort of person to help others to work upon themselves, one had to possess the ability to speak freely and frankly, regardless of risk or consequence.
The history of this particular virtue in the ancient world is varied. For instance, there is what we might think of as ‘political’ parrhêsia, characteristic of pre-Socratic Athens. This is ‘free speech’ in which what is at stake are questions of the government of others. Later, and exemplified most obviously in Socrates, we find a virtuous ‘free speech’ that is much more concerned with ‘ethics’, and with the government of the self. Socrates eschews the political field to focus instead on the conduct of individuals, and to measure the gap between the way they think they ought to live and the way they actually do. Later still we find these modalities combined in the philosophy of the Cynics, who sought both to live their own lives as bare truth (naked and in the open) and to missionise this life to those around them, to make their lives speak as examples to others (Foucault 2011).
Like any concept, parrhêsia is situated in a particular context. Not all that is true about free speech in the ancient world applies to our own. While Foucault’s own account ends broadly speaking in the classical period, tracing the later history of parrhêsia gives us some insights into the origin of contemporary liberal notions of freedom of speech. Historian David Colclough argues, for instance, that classical parrhêsia served as one of the sources for imagining freedom of speech in seventeenth century England—the period which also gave us some of the classic sources of liberal defences of freedom of speech, such as Milton’s Areopagitica, or the works of John Locke. Somewhat ironically, however, Colclough notes that parrhêsia at that point was primarily a figure of rhetoric. Rhetorical manuals drew on examples from speeches by classical Greek and Roman orators, which consisted of prefacing one’s speech by warning that one’s position was controversial, daring, and likely to offend. For seventeenth century English commentators, ‘parrhêsia’ as a rhetorical figure therefore posed an inherent problem of sincerity. It could be a genuine warning and apology for speech that was necessary, but might offend. Equally, it could be merely a cynical way to flatter an audience by delivering, as if they were surprising or extreme, views which the speaker knew were perfectly conventional and likely to gain broad assent in any case.
Colclough notes that the debates around parrhêsia were only one amongst the cultural sources of seventeenth century English discussions of the value of free speech. Others included stories from the lives of Christian martyrs who had continued to speak the truth of their faith in the face of torture and death, or the legal prerogatives of unrestricted speech that applied (in principle at least) to parliamentary discussions. Colclough’s and Foucault's accounts point to the complex, diverse, and contested genealogy of liberal visions of freedom of speech.
Anthropologists have used Foucault's discussion of parrhêsia to ask comparative questions about the ways in which freedom of speech is understood and valued in various contexts today. Pascal Boyer, for instance, has suggested that some contemporary political movements based on satire, such as Iceland’s iconoclastic ‘Best Party’—a joke political party that eventually achieved electoral success—may resemble aspects of ancient parrhêsia (2013). On the other hand, Harri Englund has pointed to the dangers of assuming that parrhêsia is portable beyond its own specific context (2018). In Finnish talk radio, he argues, what might look like ‘parrhesiastic’ speech on the part of individual callers is in fact a process carefully cultivated by the show’s hosts, an arrangement of multiple voices, rather than any individual ‘speaking truth to power’ (see below for a fuller discussion). As with many concepts, there is probably little to be gained by arguing over exactly how transposable the precise details of classical parrhêsia are or are not. The point is rather that one can ask of any context similar questions to those Foucault was asking about Ancient Greece, or Colclough about early Stuart England: what is it about ‘free’ or ‘direct’ speech that people value when they value it? To what ends is it directed? What role does it play in relation to the broader system of ethics in which it exists? How is speaking freely supposed to affect one’s relationship to oneself, and to others? These questions already move us in a much more anthropological direction than the classic juridical and political arguments over the extent of free speech rights, or the balance between freedom of speech and other legal protections.
Subjects: whose speech, and whose freedom?
Building on the above discussion, one might look more closely at who or what, in any given setting, counts as the free-speaking subject. If free speech is in some cultural contexts considered to be a virtue, we could ask: whose virtue is it? More generally, the ethnographic record compels us to move beyond a virtue-based understanding of ethics, and reconsider familiar assumptions about the individuality of speaking subjects, and the forms of freedom that characterise them.
Liberal freedom of speech could be understood as involving a specific ‘production format’ of speech (Goffman 1981), in which the speaker is simultaneously the utterer, the author, and the responsible agent of speech. Erving Goffman (1981) proposed the notion of production format to disentangle the complexity of conversation roles in communicative situations, arguing that the figure of the speaker should be differentiated into several analytical roles: the animator, i.e. the ‘sounding box’ physically pronouncing the words; the author, i.e. ‘someone who has selected the sentiments that are being expressed and the words in which they are encoded’; and the principal, ‘that is, someone whose position is established by the words that are spoken, someone whose beliefs have been told, someone who is committed to what the words say’ (Goffman 1981: 144). Separating the different conversational roles entailed in speaking and hearing, Goffman demonstrates that they might converge in the same social roles, and indeed the same person, as when we think of an autonomous, sincere speaker of liberal language ideologies; or equally, they might be distributed across several persons. One example of the latter might be the Wolof griots in Senegal—low-ranking poets hired to perform ritual insults on behalf of noblemen (e.g. Irvine 1989). Here, the roles of the animator and author converge on the individual speaker, while the principal is thought to be the collective whole to which the speaker belongs.
Building on Goffman’s work, anthropologists (e.g. Hill & Irvine 1992; Merlan & Rumsey 1991) have further explored the relation between complex, dynamic speaking roles, and the autonomy of speakers. For instance, in his ethnography of royal orators, or akeyame, in the Akan-speaking areas of Ghana in the 1980s and ‘90s, Kwesi Yankah describes them as ‘social mediators of speech’ and ‘specialists in the artistic reporting or representation of speech’ (1995: 8) as they act as ‘surrogate speakers’ for their chiefs. Yankah argues that the hierarchical subordination of the ‘surrogate speaker’—the orator—to their chief does not preclude autonomy in speech acts, for without the orator’s voice ‘a royal communicative act is incomplete’ (1995: 8). The duties of the orator ‘range from strict reporting to discretionary interpretation’, which means that relations of subordination that formally characterise surrogate speech might here entail ‘mutual reliance and dependency’ (Yankah 1995: 9). Akeyame are indispensable to royal speech, and, for instance in court judgments, ‘a greater part of akeyame’s contribution during prosecution is not structurally linked to a patron’s; it is independent’ (Yankah 1995: 163). Nevertheless, ‘in spite of its autonomy, the akeyame’s contribution is still made on behalf of the royal realm, to which they make occasional reference’. Yankah’s ethnography prompts us to question the autonomy of action inherent to the different speaking roles, and the way such autonomy is shaped by the social relations among persons performing these roles.
A similar reconfiguration of roles can be seen in the historical Soviet practice of self-criticism (samokritika), a form of speaking truth to power in which the author and addressee of speech are understood to be collective subjects, even when the speech act itself is performed by an individual person (Kharkhordin 1999; also Glaeser 2011). State socialist regimes that curtailed individual freedom of speech through explicit forms of official censorship were one of the key counterpoints against which liberal visions of freedom of speech were articulated throughout the twentieth century (cf. Boyer 2003, see below for a fuller discussion). Yet state socialism was not without its own imaginaries and practical repertories of free speech. Oleg Kharkhordin describes samokritika in Soviet Russia as a key element of socialist ethics and a means to achieving the revolutionary consciousness of the masses in the nascent Soviet state. Samokritika meant ‘an open statement by the working masses of their opinions on the weaknesses in Soviet … administrative apparatus and life’ (Viktorsky 1929: 266). Crucially, in samokritika within particular Communist Party cells or workers collectives, contemporary commentators saw not acts of individual confession or complaint but ‘the working class that upholds the proletarian dictatorship … criticiz[ing] and correct[ing] its own mistakes and failures by itself’ (Ingulov 1930: 97, in Kharkhordin 1999: 146). In theory, this notion of collective critical speech reflected the understanding of the Soviet state as an expression of class will; the ‘self’ of self-criticism referred to the working class as the sovereign of the ‘proletarian dictatorship’. In practice, however, bringing this collective subject into being through particular acts of speaking was no small feat.
In the Communist Party discourse in the 1920s and early 30s, self-criticism ‘normally meant collective criticism by Party members of the weaknesses of the Party’ (Kharkhordin 1999: 146). Thus in the 1920s, ‘The Party continuously solicited self-criticism, which in practice meant urging rank-and-file members to criticize top leaders, in order to make the body of the Party homogenous’ (Kharkhordin 1999: 149). Party theorists who promoted samokritika as a form of accountability were aware that the imperative of collective speech gave rank-and-file workers an opportunity for political manoeuvring. When in 1928, self-criticism from below led to a wave of denunciations that evidently sought to settle personal scores, Party commentators had to remind Soviet workers to criticise collective, not individual, weaknesses. At a central Communist Party committee meeting in August 1928, for example, one high-ranking speaker proposed ‘a particular psychological technique’: ‘A worker was advised to imagine, before saying something critical of a manager, that the body he was kicking was not somebody else’s but his own, since in the Party view he was assaulting a corporate body of which he was a part’ (Kharkhordin 1999: 153).
Such critical truth-telling must be understood against the background of early Soviet techniques of the self and operations of power that aimed at creating socialist unity by orchestrating forms of action and speech that transcended individual subjects. Many Bolshevik revolutionaries wanted ‘to organize their experience and energy around an ideology that would help them lose their sense of self and acquire the sense of the collective’ (Williams 1980: 393). By submitting the self to the collective, revolutionaries aimed to achieve immortality through the lasting social effects of personal sacrifice. The notion of kollektiv—a collective of people united and transformed by the common experience of working on a particular task—is key to understanding samokritika. A kollektiv, typically a workplace collective, was imagined to act and think as one, and to exert group sovereignty that subsumed individual action under the imperative of a common goal. Regular, often ritualised acts of self-criticism revealed and analysed perceived flaws in the organisation of work, relations among workmates, or even between workers and their families, as seen in the light of communist ideals. But these acts of samokritika also objectified kollektiv before itself, helped it correct itself on its path to communism, and promoted its unity by strengthening horizontal surveillance of their members over one another. Unlike critical introspection or individual confession in Western Christianity, which Foucault sees as one of the historical forces underpinning modern individuation, samokritika was expected to be performed by workers and party members before—and on behalf of—their kollektiv. One was free to speak up as long as critique was directed at the self as part of the corporate whole of kollektiv, and in so far as it promoted the ‘fusion’ (spaika) of kollektiv into one. The subject and the object of samokritika was emphatically a ‘we’: a nested corporate subject, where a kollektiv of workers stood for and became aligned with both the proletarian class they represented, and the Communist Party leading that class. In the Party’s opinion at the time, ‘[T]hrough a certain person speaking up, the whole Party criticised itself’ (Kharkhordin 1999: 146).
These comparative cases remind us that the liberal framing of free speech as performed by individual persons is only one of many cultural possibilities. Yet comparisons of this kind shouldn’t lead us to assume that liberal visions of free speech are, by contrast, simply or uniformly individualist. Consider for instance the ‘speech’ of capitalist corporations. ‘Pronounced’ by corporate spokespeople, authored by PR and press offices, and attributed to the fictive legal person of the corporation, corporate speech rarely raises the question of freedom. Yet, in a recent landmark 2010 decision in Citizens United v. Federal Electoral Commission, which enshrined the status of corporations as legal persons enjoying the same rights as human persons, the US Supreme Court granted First Amendment protections to corporations’ and unions’ direct spending on political election campaigns. The court had designated election spending as a form of protected free speech. Susan Gal and Judith Irvine explain that the consequences of speech mattered in this instance, rather than the identity of the speakers: ‘making speech available as a source of information for the public’ so as to ensure the political ideal of a well-informed citizenry (2019: 9). The Court’s majority opinion that ‘prohibition on corporate … expenditures is a ban on speech’ rested on an equation of money, a resource necessary for corporate persons to orchestrate political speech, to speech itself (Gal & Irvine 2019: 9). The opposition to the ruling predominantly focused on dismantling this analogy, and demonstrating the false equality between natural (human) and fictitious (corporate) persons. In sum, the Citizens United decision revealed competing understandings of speech in the contemporary United States: on the one hand, a view that ‘takes speech to be a material thing, equivalent to money, and independent of speakers’, and on the other, one that ‘takes speech to be different from material objects, and freedom of speech to be embodied only in natural persons’ (Gal & Irvine 2019: 10).
These and other ethnographies help us understand that the model of a self-owning, rights-bearing individual subject of free speech is only one of multiple possible ways in which human societies have thought about and organised the relation between speech and freedom.
Platforms: censorship, materiality, and mediation
The question of who is the subject of free speech leads, in turn, to a focus on the material devices, spaces, and media they engage in their communicative practices. The question of who gets to speak doesn’t exhaust debates over freedom of speech—just as important is the question of who gets to be heard, and how. This issue has gained particular visibility in contemporary debates in the US and UK over ‘no-platforming’ on university campuses and beyond. No-platforming includes practices of boycotting or uninviting a speaker, blocking their access to a forum or debate, be it online or offline, because particular views they hold are deemed offensive or harmful. A ‘platform’ in this sense refers to a literal or metaphorical stage from which to address an audience. Critics of no-platforming cast it as a new form of censorship, part of a broader ‘cancel culture’ emanating from a progressive left which is increasingly unwilling to allow views it disagrees with to be publicly expressed. Proponents of no-platforming by contrast argue that they are not censoring anyone, but simply refusing to ‘amplify’ the voices of speakers deemed not only offensive or dangerous but also—crucially—privileged in their access to other high-profile platforms for being heard. Simultaneously, some proponents argue that the public media debates occasioned by no-platforming such high-profile speakers are themselves an occasion to give more ‘platform’ to marginalised voices. From this perspective, no-platforming can be cast as a form of epistemic justice, a righting of the scales in a world in which access to platforms for expression is unequal to begin with.
Whatever one makes of these arguments, these cases usefully focus attention on the important distinction between the formal right to speak and the substantive means for being heard by others. Both sides in arguments about no-platforming appeal in various ways to a difference between what one might call, following Isaiah Berlin (1969), a ‘negative’ freedom of speech (the freedom from, for instance, legal impediments to speech) and a ‘positive’ freedom of speech (freedom to speak, which includes the means of accessing a platform from which to do so).
It is precisely because such a distinction between ‘freedom from’ and ‘freedom to’ is so difficult to make in practice that debates and concerns over freedom of speech are so often also arguments over material settings, devices, and media, in the broadest sense: objects, spaces, and techniques that mediate communication. Thus, while freedom of speech is often imagined as a single abstract principle relating to intangible contents and messages (political opinions, artistic expression, scientific knowledge), the history of changing understandings of freedom of speech is inseparable from the rise and transformation of a host of technologies of mediation: mass-circulation newspapers (Keane 2008), radio stations (Englund 2018), the cinema industry (Mazzarella 2013), television, or the internet (Coleman 2009, Gershon 2014). These material devices, spaces, and media may seem like mere background when talk is of principles. And yet they profoundly shape what ‘freedom of speech’ can concretely mean in any given situation, in ways that are historically and culturally variable. Matters of principle take multiple forms through very concrete questions of access and presence: who can speak where and who can hear them? How long can people speak for and must they take turns? What kinds of expression, beyond the spoken or written word, can be made available and under what modalities? What does it cost? How far does it reach?
Thus, on the one hand, new media have frequently been linked with new possibilities for freed and challenging expression—the heavily internet-mediated uprisings in Arabic-speaking countries in 2011 being a classic case in point. On the other hand, the mediation of expression is often the most obvious means through which it can be impeded, filtered, and censored—from the explicit work of film censorship boards, for instance, to the subtle pre-publication pressures of in-house legal advisors in publishing houses (cf. Candea 2019). Mediation in this sense is not merely a matter of technology but of the particular social relations, forms of intervention, and expertise that different technologies enable and require. For instance, in his above-mentioned work on a vox populi phone-in radio show in Finland, Englund (2018) notes how radio hosts in practice manage conversations with callers whose anti-immigration views they find unpalatable. Rather than cut them off, or even directly challenge them, the hosts steer the conversation in subtle ways in order to ‘strive for harmony’, while making space for their callers’ ‘need to be heard’ (Englund 2018: 108). It is interesting to put this example alongside Dominic Boyer’s archival exploration of the practices of state censors in East Germany (2002). Boyer shows that the classic vision of censors as mere administrative agents of deletion—erasing offending passages or cancelling entire texts—underplays censors’ view of themselves as involved in an intellectual, even productive, enterprise akin to the work of editing. Censors intervened not merely in ideological matters, but also concerned themselves with questions of style and quality; they often worked in a back-and-forth (albeit unequal) dialogue with the authors of the work. Boyer argues provocatively that state censorship was thus not always that different from the practices of editorial intervention, review, and selection practiced by academic journals. Whatever one makes of the latter comparison (see Candea 2019) these two cases are useful to think of side by side because they highlight the extent to which concerns with freedom of expression in any particular case are inseparable from the particularity of the medium through which that expression occurs. Live airtime and peer-review, for instance, both bear on the shaping, allowing, and curtailing of expression, but they do so in profoundly different ways.
While material mediation poses the question of access, it is therefore not sufficient to think of freedom of expression merely as a singular good of which one can have more or less. Changes in media also involve changes in the nature of what is expressed. In his ethnography of film censorship in colonial and postcolonial India, William Mazzarella (2013) explores the distinctive affective power of cinema as a form of mass mediation. The moving image, in its sociocultural setting of production and reception, does things to people in embodied ways, things that cannot be reduced to or deduced from an analysis of its contents, meanings, or the ideas it ‘encodes’. This in part explains the permanence and broad acceptability of film censorship even in settings in which other forms of censorship—such as official censorship of the press—have been abandoned. But more broadly, as Mazzarella notes, a history of censorship shows the extent to which the attention of censors—and, one might add, the experimentation of producers of 'content'—recurrently focuses on new media and their new ways of generating affects, just as it moves away from media which have grown familiar and old: newspapers, the radio, film, television, the internet. Anthropologists studying censorship in practice have thus contributed to the chorus of challenges brought by social theorists (Bourdieu 1991; Fish 1994; Butler 1997; cf. Bunn 2005 for an overview) against arguments for freedom of expression in which expression is set apart as a special form of conduct which is essentially about conveying contents. Attending to the materiality of media reminds us not only of the material constraints on expression, but also of its material effects.
But the materiality of media also reveals how imaginaries of freedom of expression are transformed together with the appearance of new forms and visions of the public. Thus Ilana Gershon (2014) argues that the rise of social media has contributed to the emergence of a new conception of the public, in terms of access, reachability, and responsiveness. By contrast to the classic liberal visions of a public defined as a collection of anonymous strangers (Warner 2005), publics defined by accessibility—epitomised on platforms like Facebook—are experienced by their participants as collectives structured by links extending from close friends to distant acquaintances. In these kinds of publics named relations entail accountability, a responsible and graduated use of the information that is exchanged.
Gershon analyses the tensions between this ‘new’ vision of the public as a network of knowable persons enmeshed in relations with one another and the older vision of the public as a collective of strangers, from the perspective of young social media users whose comfort zone is broadly situated in the former. These younger informants, Gershon argues, ‘often believe that members of a public will experience certain obligations in managing information, and as a result will act responsibly. At the same time, they imagine that they can anticipate who might read their material’ (Gershon 2014:80). Yet these new online publics are also the home of internet ‘trolls’—anonymous users who post inflammatory comments or target and harry other users with pranks and attacks which seem designed to puncture this feeling of online safety. Gershon follows Gabriella Coleman (2011) in characterising trolls as self-appointed crusaders for a return to an older vision of the public as a collective of strangers who do not take things personally. It is thus unsurprising, perhaps, that Gershon’s informants feel that the public sphere beyond their own familiar and accountable networks is a space of risk, and ‘anonymity a cover for antagonism’ (Gershon 2014: 84).
In sum, Gershon’s argument shows how these new online public/private borderlands are the scene of struggles and accommodations between radically different ethics and politics of communication. These contested spaces increasingly overspill the porous boundaries between online and offline communication. Shifting struggles are illustrated in the rise of a bevy of neologisms—‘echo chambers’, ‘safe spaces’, ‘snowflakes’, ‘haters’, ‘trigger warnings’—which purport to diagnose communicational pathologies or, on the contrary, hoped-for solutions to the risks of expression through shifting and ambiguous media. Returning to the opening problematic of this section, one might say that attending to the materiality of media suggests that being heard is not simply a right, but can also be a vector of risk.
Voice: embodiment, affect and sound
Apart from being shaped by the materiality of their settings, practices of free speech are also constituted by what they look, feel, and sound like. Anthropologists have studied linguistic and vocal practices that do not involve the kinds of reasoned, articulate forms of speech ideally associated with democratic participation, but rather emphasise the embodied and affective nature of communication. Theories of free speech and political engagement have typically been premised on the idea of citizens having a ‘voice’ within the polis, with that voice understood as a transparent representation of the conscious, self-interested, individual self (Kunreuther 2014, 2018; Weidman 2014). In recent explorations of how voice manifests as part of the production and transformation of publics and political movements in various cultural contexts, though, where voice is often still used as a metaphor for political participation, actual practices of voicing involve bodies, sounds, and collectives of people in ways that do not map neatly on to traditional liberal notions of political and free speech. The notion of voice has been helpful as a way to consider political speech, as it can shift our attention away from the linguistic and semiotic content of the speech at hand, and focus instead on the actual sounds being produced and circulated, which in turn brings to light the various bodies and materialities at play in the making of free speech.
In her study of the sounds of protest events in Kathmandu, Laura Kunreuther (2018) shows how artists’ and demonstrators’ use of various kinds of noise—produced by cars moving through the city, protestors banging on pots and pans, and the radio broadcast of recordings of human crying—transform what is generally thought of as unruly, unproductive sound into political engagement. Kunreuther describes a 107-day demonstration in front of the Prime Minister’s residence protesting violence against women, in which everyday noises were repurposed to indicate popular support for the movement and a challenge to civic life as usual that, yet, was expressed through its very own auditory forms. The use of domestic items such as pots, pans, and plates, for example, served to bring the home and domestic sphere into the public and political realm, and in particular evoked the status of women as those who generally perform household labour and whose experiences of being subject to violence often remain hidden. Beyond these immediate resonances, the noise of the banging acted to reveal popular anger and discontent, as Kunreuther suggests, ‘signifying through noise a breakdown in communication between ruler and ruled’ (2018: 23). In this way, noise becomes a form of political 'speech' and a way in which protestors can shape the forms of their expression without necessarily having to use words at all.
Similarly, Kunreuther shows how sounds produced by humans, but that are not made up of words, can speak volumes as part of the non-linguistic, affective realm of politics. In a performance piece by a Nepali artist, staged during the Maoist insurgency and in the context of regular state violence against protesters, recordings of mothers and babies crying were compiled and broadcast both at the site of the street performance and on all national FM stations (2018: 14-15). The sound of the wailing was effective in calling forth a national, secular public given the anonymity of the voices heard, who, although clearly women and children, were not identifiable through accent, social class or caste, ethnicity or religion. Combined with the imagery evoked of the women, heard as mothers of the nation, and of a genre of sound mostly heard in funerary and wedding rituals, the broadcast had the effect of sidestepping the government/Maoist divide, with both sides claiming the piece was condemning the other. There was a sense, then, that a purer, more human voice was made possible through the use of the immediacy of the cry, devoid of language but able to express meanings otherwise hampered in the context of civil conflict. This interpretation stands in contrast to those theorists of liberalism who have framed the bodily and collective energy of the crowd as a threat to the measured, reasonable publics of deliberative democracy (see Cody 2011, 2015; Mazzarella 2015), and draws instead on theories of popular assembly that reframe how the gathering of publics and collectivities can be a central and transformative part of democratic and other political processes (Butler 2015, Butler & Athanasiou 2013).
Finally, even silence can be thought of as a form of free speech. The absence of words, sound, or noise is a tool that protesters in diverse contexts have employed to communicate opposition to government practices of censorship, war, and oppression. As an easily translatable technology of protest, silence symbolises popular dispossession or a government’s lack of listening to what is being said out loud in the public sphere. Through the intensity of the silence of a large crowd gathered in a normally noisy public space, it has a solemn emotional character while also emphasising popular cohesion in support of a political position. As Kunreuther (2018) explains, in its use both by the performance artist who employed silence in parallel with the broadcast crying described above, and by journalists and media personnel at other moments in Nepali history to highlight government censorship, silence recalls the modern liberal subject. It implies silent concentration and rational, reflective engagement with the political, but does so without concealing the bodily and collective instantiation of these democratic subjects, given the centrality of embodied presence to the protest. As Athena Athanasiou also observes about the use of silent vigils by activists in post-conflict Serbia, silence can be a powerful, subversive force precisely because it can express forms of mourning and of protesting injustice that, when people attempt to voice them through language, become tied up in the limits and politically exclusionary nature of speech and representation (Athanasiou 2005, 2017).
Free speech, therefore, may take the form of non-linguistic noise and sound, bodily presence, and symbolic resonance, as much as it can involve verbal forms of expression. By focusing ethnographically on the material, embodied, and affective forms through which political voice actually takes shape, we see that free speech is in practice a much wider and more diverse phenomenon than its abstraction as a category of liberal thought implies.
While anthropologists have not as yet written much on the subject of freedom of speech, this entry points to anthropological studies of language, voice, ethics, subjectivity, and media that can help to complement, critique, and contextualise political scientific, legal, and philosophical accounts of the subject. One upshot of these studies is to put canonical liberal visions of freedom of speech in comparative and historical perspective, as one amongst a range of ways of imagining the proper relationship between subjects, speech, and freedom. Another effect of these studies is to highlight the ways in which visions of free speech—whatever cultural form they espouse—take shape within and against specific material and embodied possibilities and constraints. In these ways, anthropology can enrich our understandings of free speech as a multiple, contested, and frequently unattainable horizon of desire and action.
The writing of this article was supported by the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement 683033).
Athanasiou, A. 2005. Reflections on the politics of mourning: feminist ethics and aolitics in the age of empire. Historein 5, 40-57 (available on-line: https://doi.org/10.12681/historein.72).
— — — 2017. Agonistic mourning: political dissidence and the women in black. Edinburgh: University Press.
Asad, T., W. Brown, J. Butler & S. Mahmood 2013. Is critique secular? Blasphemy, injury, and free speech (2nd revised ed.). New York: Fordham University Press.
Bauman, R. 1984. Let your words be few: symbolism of speaking and silence among seventeenth-century Quakers. Tucson: Wheatmark.
Berlin, I. 1969. Four essays on liberty. Oxford: University Press.
Bishara, A. 2013. Back stories: U.S. news production and Palestinian politics. Stanford: University Press.
Bourdieu, P. 1991. Language and symbolic power. Cambridge: Polity Press.
Boyer, D. 2003. Censorship as a vocation: the institutions, practices, and cultural logic of media control in the German democratic republic. Comparative Studies in Society and History 45, 511-45.
Boyer, P. 2013. Simply the best: parody and political sincerity in Iceland. American Ethnologist 40, 276-87.
Bunn, M. 2015. Reimagining repression: new censorship theory and after. History and Theory 54, 25-44.
Butler, J. 1997. Excitable speech: a politics of the performative. London: Routledge.
— — — 2015. Notes toward the performative theory of assembly. Cambridge, Mass.: Harvard University Press.
Butler, J. & A. Athanasiou 2013. Dispossession: the performative in the political. Cambridge: Polity Press.
Candea, M. 2019. Silencing oneself, silencing others: rethinking censorship comparatively [introduction]. Terrain (available on-line: http://journals.openedition.org/terrain/18773). Accessed 3 February 2020.
— — — 2019. The duelling ethic and the spirit of libel law: matters and materials of honour in France. Law Text Culture 23, 29.
Cody, F. 2011. Publics and politics. Annual Review of Anthropology 40, 37-52.
— — — 2015. Populist publics: print capitalism and crowd violence beyond liberal frameworks. Comparative Studies of South Asia, Africa, and the Middle East 35(1), 50-65.
Coleman, G. 2009. Code is speech: legal tinkering, expertise, and protest among free and open source software developers. Cultural Anthropology 24, 420-54.
Coleman, E.G. 2011. Anonymous: from the lulz to collective action. In Politics in the age of secrecy and transparency (ed.) E.G. Coleman (available on-line: http://mediacommons.org/tne/pieces/anonymous-traveling-pure-lulz-land-political-territories). Accessed 3 April 2020.
Englund, H. 2018. The front line of free speech: beyond parrhêsia in Finland’s migrant debate. American Ethnologist 45, 100-11.
Faubion, J.D. 2001. Toward an anthropology of ethics: Foucault and the pedagogies of autopoiesis. Representations 74, 83-104.
Favret-Saada, J. 2015. Comment produire une crise mondiale avec douze petits dessins. Paris: Fayard.
Fish, S.E. 1994. There’s no such thing as free speech, and it’s a good thing too. Oxford: University Press.
Foucault, M. 2010. The government of the self and others. London: Palgrave Macmillan.
— — — 2011. The courage of truth. London: Palgrave Macmillan.
Gal, S. & J.T. Irvine 2019. Signs of difference: language and ideology in social life. Cambridge: University Press.
Gershon, I. 2014. Publish and be damned: new media publics and neoliberal risk. Ethnography 15, 70-87.
Glaeser, A. 2011. Political epistemics: the secret police, the opposition, and the end of East German socialism. Chicago: University Press
Heywood, P. 2019. Fascism, uncensored: legalism and neo-fascist pilgrimage in Predappio, Italy. Terrain 72, 86-103.
Hill, J.H. & J.T. Irvine (eds) 1992. Responsibility and evidence in oral discourse. Cambridge: University Press.
Irvine, J.T. 1989. When talk isn’t cheap: language and political economy. American Ethnologist 16(2), 248-67.
Keane, W. 2007. Christian moderns: freedom and fetish in the mission encounter (1st ed.). Berkeley: University of California Press.
— — — 2008. Freedom and blasphemy: on Indonesian press bans and Danish cartoons. Public Culture 21(1), 47-76.
— — — 2015. Ethical life: its natural and social histories. Princeton: University Press.
Kharkhordin, O. 1999. The collective and the individual in Russia: a study of practice. Berkeley: University of California Press.
Kunreuther, L. 2014. Voicing subjects: public intimacy and mediation in Kathmandu. Berkeley: University of California Press.
— — — 2018. Sounds of democracy: performance, protest, and political subjectivity. Cultural Anthropology 33(1), 1-31.
Laidlaw, J. 2002. For an anthropology of freedom and ethics. Journal of the Royal Anthropological Institute 8, 311-32.
— — — 2013. The subject of virtue: an anthropology of ethics and freedom. Cambridge: University Press.
Lambek, M. (ed.) 2010. Ordinary ethics: anthropology, language and action. New York: Fordham University Press.
Larson, J.L. 2013. Critical thinking in Slovakia after socialism. Rochester: University Press.
Levinson, S. 1983. Pragmatics. Cambridge: University Press.
Mazzarella, W. 2013. Censorium: cinema and the open edge of mass publicity. Durham, N.C.: Duke University Press.
— — — 2015. Totalitarian tears: does the crowd really mean it? Cultural Anthropology 30(1), 91-112.
Robbins, J. 2007. Between reproduction and freedom: morality, values, and radical cultural change. Ethnos 72, 293-314.
— — — 2016. What is the matter with transcendence? On the place of religion in the new anthropology of ethics. Journal of the Royal Anthropological Institute 22, 767-808.
Schäfers, M. 2017. Voice. In The Cambridge Encyclopedia of Anthropology (eds) F. Stein, S. Lazar, M. Candea, H. Diemberger, J. Robbins, A. Sanchez & R. Stasch (available on-line: http://doi.org/10.29164/17voice).
Viktorsky, G. 1929. Samokritika v sisteme sovetskoi demokratii (Self-criticism in the system of Soviet democracy). In Sovetskaia demokratiia (ed.) I.M. Steklov. Moscow: Sovetskoe Stroitel’stvo.
Warner, M. 2005. Publics and counterpublics (new ed. edition). Cambridge, Mass.: MIT Press.
Weidman, A. 2014. Voice and anthropology. Annual Review of Anthropology 43, 37-51.
Whitman, J.Q. 2000. Enforcing civility and respect: three societies. The Yale Law Journal 109, 1279.
Williams, R. 1980. Collective immortality: the syndicalist origins of proletarian culture, 1905-1910. Slavic Review 39(3), 389-402.
Woolard, K.A. & B.B. Schieffelin 1994. Language ideology. Annual Review of Anthropology 23, 55-82.
Yankah, K. 1995. Speaking for the chief: ȯkyeame and the politics of Akan royal oratory. Bloomington: Indiana University Press.
Note on contributors
Matei Candea is a Reader in Social Anthropology at the University of Cambridge and former editor of the Journal of the Royal Anthropological Institute (2013-2016). He is the author of Corsican fragments: difference, knowledge and fieldwork (2010, Indiana), and editor of The social after Gabriel Tarde (2010, Routledge) and Detachment: essays on the limits of relational thinking (2015, Manchester University Press) with Jo Cook, Catherine Trundle and Tom Yarrow. He has published a number of articles on politics, identity, hospitality, human-animal relations, behavioural science, and anthropological comparison.
Matei Candea, Department of Social Anthropology, Free School Lane, Cambridge CB2 3RF, firstname.lastname@example.org
Paolo Heywood is Assistant Professor of Social Anthropology at Durham University. He is the author of After difference: queer activism in anthropological theory (2018, Berghahn).
Paolo Heywood, Department of Anthropology, Durham University, Dawson Building, South Road, Durham, DH1 3LE, email@example.com
Fiona Wright is a Research Fellow at the University of Edinburgh. She works on care, activism, dissent, and ethics, and how they are linked to sovereignty and violence. She is the author of The Israeli radical left: an ethics of complicity (2018, University of Pennsylvania Press).
Fiona Wright, Advanced Care Research Centre, Usher Institute, University of Edinburgh, 9 Little France Road, Edinburgh BioQuarter, Edinburgh, EH16 4UX, firstname.lastname@example.org
Taras Fedirko is a British Academy Research Fellow in social anthropology at the University of St Andrews. He is the editor of Grammars of liberalism, a special collection in Social Anthropology with Farhan Samanani and Hugh Williamson. His research focuses on middle-class professionals involved in promoting political liberalism at the core (Britain) and semi-periphery (Ukraine) of global capitalism.
Taras Fedirko, Department of Social Anthropology, 71 North St, St Andrews KY16 9AJ, email@example.com
 It is also worth noting that stark contrasts between ‘Western/liberal’ and ‘Muslim’ language ideologies or perspectives on the Danish cartoon controversy overwrite the diversity of understandings within each of these ensembles, which are hardly mutually exclusive—as these anthropologists themselves acknowledge (Keane 2008: 57; Asad et al. 2013: viii). For a different anthropological reading of the case, which puts the emphasis on how specific actors worked to produce a global sense of a singular ‘Muslim reaction to the cartoons’, see Favret-Saada (2015).
 An instrument of socialist reflexivity and resistance, the notion of samokritika became a tool of punitive power towards the end of the 1930s, when it shaped the stakes and form of (forced) confessions of defendants during the infamous Stalinist show trials.
 This and other anthropological work on voice is explored by Marlene Schäfers (this volume).