Introduction: Freedom, Speech, and Anthropology’s Comparative Imaginary

Anthropologists have not, until relatively recently, paid much attention to freedom of speech as an ethnographic object. Where they have done so, it has been predominantly in light of arguments about the fundamentally limited and narrow nature of liberal understandings of language and signs. Webb Keane, for instance, wrote, “The classic defense 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). Liberal defenses of freedom of speech, on this view, are grounded in an unrealistic “vision of language purified of morally constitutive effects or independent of social ties” (ibid.: 70). They are therefore inherently blinkered toward more performative views of language and representation. This critique rejoins older arguments by cultural critics, such as Stanley Fish, who in effect charge defenders of free speech with a laughable “sticks and stones” approach to words.Footnote ¹

In deploying the notion of “semiotic ideology,” however, Keane is not merely critiquing liberal conceptions of freedom of speech—he is doing so comparatively. Keane’s piece was written in the context of the controversy which arose in 2005 following the publication of cartoons satirizing the prophet Muhammad by the Danish journal Jyllands-Posten. One of Keane’s main points is that a narrowly representationalist semiotic ideology explains the tone-deafness of Danish defenders of freedom of speech to Muslim expressions of offense and injury at the cartoons. Such reactions were cast as the mark of an essentially non-modern misunderstanding of the “true” nature of signs. In fact, Keane argues, the self-appointed defenders of Danish free speech were themselves “in the grip of a selective semiotic ideology … that leads them to misconstrue the nature of their own actions” (ibid.: 61).

Writing on the same topic, Talal Asad and Saba Mahmood deploy Keane’s work on semiotic ideology to make a slightly different point (Asad Reference Asad, Asad, Mahmood, Butler and Brown2013; Mahmood Reference Mahmood, Asad, Brown, Butler and Mahmood2013). Keane, wary of the danger of setting up civilizational dualisms, stopped short of saying anything substantive about the semiotic ideology which underpinned Muslim expressions of offense at the cartoons. Asad and Mahmood by contrast take precisely that step. Building on Keane’s characterization of liberal semiotic ideology, they seek to characterize an alternative semiotic ideology present, they argue, in some Islamic traditions. In this alternative ideology, signs are deeply entangled in social ties and rich with ethical force. The moral injury dealt by the cartoons stems, they assert, from a perceived attack on a relationship of intimacy and formative ethical dependency with the prophet.

As critical interventions, Asad and Mahmood’s essays are extremely powerful. And yet there is also something troubling about the way these arguments seem to replicate visions of a grand struggle between competing liberal and Islamic language ideologies—precisely the narrative that Keane, Asad, and Mahmood appear poised to defuse. As I have argued at greater length in another piece (Candea Reference Candea, Candea, Fedirko, Heywood and Wright2025), this is partly a matter of comparative strategy. Asad and Mahmood deploy a classic anthropological device: “frontal comparison,” in which a familiar “us” position is contrasted to an ethnographically substantiated “them” position. The move is designed to counteract stereotypes about Islamic reactions to the cartoons, but its form replicates the binary way in which popular debates about freedom of speech have themselves tended to take shape in Western public settings. I would not be the first to deplore the binarism of debates such as those surrounding the Danish cartoons (see, for instance, Tønder Reference Tønder2013: 8). What I wish to highlight is the peculiar scalar pattern of these debates, which seem forever to be opposing “them” and “us,” on different scales of abstraction. Whether they are cast as a contrast between “liberal democracies” and “totalitarian regimes,” between secularism and religious ontologies, between French republicanist intransigence and American multicultural tolerance, or, conversely between American First Amendment absolutism and European laws against hate-speech, between “the right” and “the left,” or between older and newer generations of progressive voices, all of these dualities recapitulate a vision of a struggle between those who would have more unfettered freedom and those who would have more social responsibility. These contrasts all point to radically different situations, cases, and instances of what Keane has termed “metapragmatic struggle” (2025). And yet the parallelism of form across scale enables free speech debates to gather a certain fractal or self-similar dynamic (not unlike the dynamic observed for public/private distinctions by Gal Reference Gal2002). On whatever scale it is considered, the opposition seems to emerge as forever the same, the stakes restated, over and again, in terms of a tension between unbounded individual freedom of expression and its proper, socially agreed limits.

In this context, a key contribution anthropologists might make to understandings of freedom of speech is to experiment with a different comparative imaginary—one that encourages us to count beyond two. To do so, this paper returns to the Danish cartoons case to elaborate on some of the complexities noted in passing by Keane, and which are lost in Asad and Mahmood’s arguments. Keane repeatedly acknowledges that the liberal semiotic ideology he is describing is merely one amongst a number of Western visions of signs. These points are made, however, as caveats to an argument focused on the link between free speech and representationalism. Thus, Keane notes that Western thinkers such as Wittgenstein and Austin argued against the separation of language from action, but he takes the very necessity of these arguments as proof of “the inertial power of the commonsense [representationalist] view” (2009: 58) of language. Similarly, Keane pointedly rejects the thought “that the sense of offense some Muslims expressed is fundamentally alien to ‘the West’ (as Spanish laws against lèse-majesté and American reactions to flag burning or the Piss Christ artwork make clear)” (ibid.). Yet these other Western understandings of signs are backgrounded in his argument, which focuses on establishing the dominance of one representationalist view as the explanatory framework for the Danish cartoons case. This is even clearer in Asad and Mahmood’s reduction of this debate to that of a clash between liberal representationalism and an alternative Islamic semiotic ideology. They, too, deploy caveats, but caveats easily fall out of view in the heady rush to frontal comparison.

By turning the comparative lens inwards, this article seeks to “stay with the trouble” of the multiplicity of Western semiotic ideologies indexed by Keane. The article takes as its ethnographic object a later iteration of the Danish cartoons controversy—the landmark ruling in which a French court judged that Charlie Hebdo’s republication of the cartoons did not constitute a hate speech offense (TGI Paris 2007). This ruling, as we shall see, both draws on and finds a place within what Scott has described as “the Whig-liberal narrative of blasphemous modernity, reciting a litany of martyrs to freethought who, by dying at the hands of inquisitorial religion, gave birth to our secular age—an enlightened epoch of audacious irreverence and salubrious profanation” (2023: 3).

In order to provincialize this type of narrative, anthropologists have in the main focused on multiplying and complicating accounts of liberalism’s putative “others.” While state legal orders may imagine themselves as sovereign, they are often shadowed by alternative ways of accounting blame, judging fault and causation, and seeking redress (see, for instance, Engel and Engel Reference Engel and Engel2010; Richland Reference Richland2008). In the same vein, an exploration of the everyday lives of Islamic legal and moral registers in contemporary Europe (e.g., Fernando Reference Fernando2010) provides valuable contexts for understanding what was truly at stake for those seeking to limit, through law, the publication of the Danish cartoons. Asad’s sophisticated genealogy of the Islamic legal concepts often mistranslated as “blasphemy” (2013; see also Scott Reference Scott2023: 7), and his earlier exploration of the Islamic duty of morally corrective criticism (Reference Asad1993; see also Bhojani and Clarke Reference Bhojani, Clarke, Candea, Heywood, Fedirko and Wright2025), partake of this tradition. As Bhojani and Clarke note in a forthcoming chapter (ibid.), reading the Danish cartoons controversy through the lens of Islamic notions of public morality and pedagogical speech would surely help to re-frame Muslim responses to the cartoons, whether it be those of “ordinary” French Muslims, or those of the judicial activists who brought court cases such as the one described here.

While I will be attentive to these broader contexts throughout, this article takes a different route to unsettling and provincializing liberal certainties. I will focus inward, centering a self-consciously liberal and secular legal ruling in order to reveal the fractures and ambivalences at its very heart.Footnote ² Firstly, by foregrounding the specificities of a French civil law context, the article acts as a reminder that “French republicanism is not the same as anglophone Whig liberalism, and to pry them apart is to see ‘the West’ dissolving into a richly variegated transcolonial landscape” (Scott Reference Scott2023: 3). Secondly and more broadly, the article shows that, beneath the image of a tone-deaf liberal representationalism lie deeper passions and tensions at the heart of liberal visions of freedom of speech: enlightened reason, audacious martyrdom, and irreverent profanation are not always comfortable bedfellows.

The article’s first part examines the different semiotic ideologies indexed in the various twists and turns of the arguments defending the re-publication of the Danish cartoons. In a perceptive book written twenty years ago, communications scholar John Durham Peters (Reference Peters2005) highlighted the internal tensions and richly multiple genealogies of “liberal” perspectives on free speech. Building on Ernest Gellner’s (Reference Gellner1992) claims about a global clash between enlightenment doubt, cultural pluralism, and fundamentalism, Peters describes free speech debates in terms of three characteristic figures in a social drama: “Liberal tolerance, cultural transgression, and conservative offense” (2005: 14). Whatever one makes of Gellner and Peters’ specific terminology and characterizations, the broader heuristic form they deploy is noteworthy. Instead of a binary, we have three competing visions of the same problem, each of which challenges the other two “like rock, paper, scissors” (ibid.: 12). As a heuristic device for parsing controversies, this is a very productive way of defusing dualisms.Footnote ³

In that spirit, I will endeavor in the first part of this article to count to three. The ruling does appeal to, and in the end seems to plump for, a broadly Habermasian defense of the cartoons as instances of public debate and democratic deliberation—echoing the standard representationalist framework identified by Keane, Asad, and Mahmood. And yet I will show that this Habermasian vision jostles with two other semiotic ideologies at the heart of the ruling. One of these indexes the dangerous and unfettered provocativeness of antinomian laughter. The other tracks the potential of signs to deal relational damage through the medium of honor, reputation, and insult, but also recognizes the value of brave, risk-taking speech. These alternative semiotic ideologies are not presented as the inadmissible or pre-modern understandings of a religious “other,” but rather as familiar and valid considerations with a deep history in French and more broadly liberal secular public space.

In the second part of the paper, I show how the ruling tidies up this cacophony of semiotic ideologies into a neatly Habermasian conclusion. Following anthropologists of law such as Justin Richland (Reference Richland2013), I attend to the particular ways in which French law is founded and enacted through language. I argue that the form of this ruling dramatizes a tension within French legal language ideology itself, which has long been torn between the imperious brevity of legal literalism and a more discursive “democratic” style.

In sum, the article has two substantive aims in view. The first is to challenge the thought that liberal defenses of free speech are necessarily a mark of representationalist naïveté. Rather, they emerge from struggles and uneasy recombinations of very different understandings of representation—all simultaneously current in liberal public spaces. The second is to showcase a comparative technique for exploring the internal multiplicity of liberalism.

Part I. Three Structures of Plausibility for Freedom of Speech

Honor: Insult, Bravery, and Relational Damage

Anthropologists familiar with recent work on Islam and secularism are likely to see the outcome of what became known simply as “the caricatures trial” as entirely predictable. Commenting on the earlier Danish iteration of the cartoons affair, Asad and Mahmood argued that the juridical apparatus of the liberal state is unable to register the distinctive kind of injury caused by these images. In support of this claim, Mahmood quotes one young British Muslim interlocutor complaining of “the absolute lack of understanding on the part of my secular friends … at how upset people like myself felt on seeing the Prophet insulted in this way. It felt like it was a personal insult!” (2013: 69). Similarly, Mahmood quotes an elderly Egyptian man saying, “I would have felt less wounded if the object of ridicule were my own parents. And you know how hard it is to have bad things said about your parents, especially when they are deceased. But to have the Prophet scorned and abused this way, that was too much to bear!” (ibid.).

Mahmood advances the view that the juridical language of hate speech, to which some Muslim organizations turned in the wake of the publication of the cartoons, was unlikely to be successful in translating these concerns. Euro-American law favors “a language ideology in which the primary task of signs is the communication of referential meaning” (Mahmood Reference Mahmood, Asad, Brown, Butler and Mahmood2013: 77). This, for Mahmood and Asad, is part of a broader secularist apparatus of state power which seeks to “remake religion through the agency of the law” (ibid.: 87), and discipline the faithful into translating the embodied, relational, and affective demands of their faith into matters of individual choice and propositional “belief.” This setup is what renders illegible and illegitimate Muslims’ sense that the Danish cartoons attacked, not some abstract boundary of the Habermasian public sphere, but their personal relationship to the prophet.

The substance of the Paris tribunal’s 2007 “caricatures” decision as outlined above could be read merely as confirmation of these anthropological insights. Both the acquittal itself, and elements such as the definition of caricatures as a matter of free “communication of thoughts and opinions,” are straight out of the liberal language ideology identified by Mahmood, Asad, and others. And yet, a closer look at the substance of the decision shows that the ruling is in fact centrally concerned with the relationship between this vision of language and other rather different semiotic ideologies.

Of the many pictures in the issue, three specific images were retained by the prosecution. The first was an original drawing by Charlie Hebdo cartoonist Cabu, on the cover of the special issue. Under the caption “Muhammad outflanked by the fundamentalists,” the drawing depicts the prophet lamenting, “It’s tough to be loved by morons.” The other two drawings were reprinted from the original Jyllands-Posten publication. One depicted recently deceased suicide bombers lining up for heaven, met with the prophet shouting, “Stop, stop, we’ve run out of virgins.” The third was a picture of the prophet wearing a head covering which is also a bomb, on which is inscribed, in Arabic script, the shahada, a Muslim profession of faith.

The judgment takes its time on each of the three incriminated drawings, recalling the arguments of the parties and drawing on elements of context about the placement of the drawings and captions within the journal. In its evaluation the court does not dismiss claims that the images might be harmful or reduce them to mere communications of referential meaning. Rather, the decision considers at some length whether and to what extent the drawings can be taken as insults, and if so, to whom. The judgment relies on a view of images and words as performative, capable of acting upon and affecting relationships and persons, entangled in a web of relational flows of honor, insult, and reputation. Far from abstract considerations of “blasphemy” or “belief,” this section of the discussion speaks to French press law’s longstanding concerns with libel, honor, and reputation.

France’s current press law was enacted in 1881, during a decade which saw more than two hundred duels between journalists and readers who felt their honor had been slighted (Nye Reference Nye1998: 190). The attention of the legislators was fairly evenly split between, on the one hand, Habermasian concerns with the public sphere, democracy, and the power of the press to hold governments to account, and on the other a vision of language as a medium for honor, insult, and relational damage (Whitman Reference Whitman2000; Candea Reference Candea2019). It is striking how much this initial context still tells us about French free speech law today. The court which heard the caricatures case spends the majority of its time hearing cases of libel, defined in French as the “allegation or imputation of a fact which causes harm to the honor or esteem (à l’honneur ou à la consideration) of the person or body to whom the fact is imputed” (Loi du 29 juillet 1881 sur la liberté de la presse n.d.: 29). When something like a concern with “hate speech” was retrofitted to the 1881 press law in the 1970s, this was done simply by aggravating the penalty for libel and insult when directed at a person “on account of their belonging or not belonging to a particular ethnic group, nation, race, or religion.” (ibid., articles 32–33). Legally speaking, the offense which Charlie Hebdo was accused of had nothing to do with “hate” or “phobia” and everything to do with “insult”—“an affront, expression of contempt or invective” (ibid., article 29).

In its consideration of the drawings, the ruling begins by noting that they are unambiguously insulting, before turning to the question of whom, specifically, is being insulted. The judges analyze the depiction of the prophet in the first two drawings and claim that these two images actually paint a positive, rather than a defamatory, picture of the prophet, as, in the first case, “driven to despair by the betrayal of the spirit of his message” (TGI Paris 2007), and in the second case, as explicitly opposed to suicide bombing. On this basis the court agrees with the repeated claim of the defense, that the first two drawings are “clearly” targeting their insults at “fundamentalists” and not at “all Muslims,” and therefore cannot be considered as insults to persons on the basis of their belonging to a religion (ibid.).Footnote ⁴ Whatever one makes of the strength or weakness of these arguments, the point is that they are cast in the same idiom of insult as the claims of Mahmood’s interlocutors, quoted above.

What is French law doing when it seeks to protect persons’ honor and extract reparations for insult? Here we can turn to a rich anthropological, philosophical, and sociological literature which has dissected the notion of honor in a number of different ways. Authors have grappled in particular with the paradoxical relation, within the notion of honor, of two seemingly different ideas. “Honor” relates both to an internal quality (something like “virtue” or “integrity”) and to the external evaluation of this quality (something like “reputation”) (Abu-Lughod Reference Abu-Lughod1986; Appiah Reference Appiah2011; Pitt-Rivers Reference Pitt-Rivers1977; Stewart Reference Stewart1994). The eighteenth-century philosophes—who produced the first abundant literature on the subject—sought to distinguish the false attractions of mere reputation from the true virtue of inward honor (LaVaque-Manty Reference LaVaque-Manty2006; Taylor Reference Taylor and Gutmann1994). Late twentieth-century accounts influenced by a grand narrative of modernization saw this polysemy instead as the mark of a historical shift in European understandings of the person: a progressive internalization of honor, which leads to its eventual obsolescence and replacement by “dignity.” Thus, in Berger’s influential account, honor is pre-modern, feudal, inegalitarian and role-bound; dignity is modern, bourgeois, egalitarian (Berger Reference Berger1970). “In a world of honor,” the individual’s true identity is essentially located in his social roles (“the individual is the social symbols emblazoned on his escutcheon”), whereas “in a world of dignity,” the social roles are masks, disguises which hide the true self, “the solitary self … intrinsic humanity divested of all socially imposed roles or norms” (ibid.: 342). In modern society, according to Berger, “honor” is obsolete. Tellingly, he grounds his point in an argument about law, which deserves quoting in full:

There is a striking analogy between Berger’s arguments here and Mahmood’s claim that “liberal law” cannot understand the relational damage dealt by signs. This is no coincidence—the sociological accounts of Western liberal modernity on which authors such as Asad and Mahmood draw are themselves profoundly influenced by the characterizations of an earlier modernization theory, recast now in a critical vein.Footnote ⁵

These arguments are prima facie incorrect when leveled at French law: insult and libel are certainly actionable and recognized as real injury in French law, even though there is much debate among jurists as to why and how this should be so (Beignier and Foyer Reference Beignier and Foyer2014). And yet French law, for all its differences from Anglo-American common law, remains unambiguously a species of “liberal law” in the key ways highlighted by authors such as Asad and Mahmood, including its focus on individual responsibility. The vision of liberal law as irredeemably representationalist glosses over its internal tensions.

The same is true of the adjective “modern.” Before we rush to conclude, with Berger, that its concern with honor makes French law “archaic,” consider the elegant way in which the vexed duality of honor (both internal truth and relational interpersonal effect) resolves postmodern conundrums about the specific nature of verbal injury. Judith Butler, in one of the landmark contributions to this question, observes, “There is no language specific to the problem of verbal injury which is, as it were, forced to draw its vocabulary from physical injury” (Reference Butler1997: 4). Butler has in mind here U.S. debates over hate speech in which anti-racist and feminist activists had frequently reached for analogies with physical harm in order to characterize the harms produced by “words that wound” (Matsuda Reference Matsuda1993). Butler points out that while these moves can be effective up to a point, they mistake real differences between speech harms and bodily harms at their peril. In particular, the metaphor of “words that wound” leads to an over-emphasis on the direct power of words—their illocutionary force. It decontextualizes each speech act from the wider networks of signification and historical sedimentation which gives it its power to harm and to heal, to make and unmake persons. Pace Butler, however, this language specific to the problem of verbal injury exists: it is the language of honor—or if that sounds too archaic, then insult and reputation. This is a language which remains very much alive, constantly subverting attempts to distinguish between internal qualities and external relations. Honor has long provided a way to think about the ways in which signs, and particularly public ones, can shape, fashion, and unmake persons to their core.

Far from being deaf, therefore, to the injury made to the faithful on seeing, as Mahmood’s interlocutors put it, the prophet insulted, scorned, or abused, this is precisely the language in which the decision considers the three drawings. This becomes even clearer in the decision’s account of the third drawing (the “bomb” drawing). There, the court rejects the defense’s claim that this can be interpreted in the same way—as merely an insult to fundamentalists or terrorists. The court here recalls the claim by an academic witness at the trial, Abdelwahhab Meddeb, according to whom the very physiognomy of the prophet as portrayed in this picture is part of a “long Islamophobic tradition depicting the prophet as bellicose and lustful” (TGI Paris 2007). The decision at this point notes that “this drawing appears, in itself and taken in isolation, of such a nature as to insult all of the adepts of this faith, and to besmirch their reputation [les atteindre dans leur consideration]” (ibid.). In these sections of the caricatures ruling, the court unselfconsciously embraces a rather un-secular sounding semiotic ideology in which insults to the person of the prophet could be a medium of relational damage to the faithful.

In sum, as demonstrated by the hundreds of libel cases prosecuted by the French Chamber of the Press and of Public Liberties every year, liberal law is perfectly able to see beyond a naively representationalist vision in which words and pictures are distinguished from any relational power or effect. Indeed, the performative semiotic ideology in which personal integrity is demonstrated, damaged, and upheld through the relational use of signs, pervades the law well beyond the question of insult and libel. Legal language in France is suffused with appeals to honor, from the basic mechanics of sworn statements (attestation sur l’honneur), to the numinous power and responsibilities of judges themselves.Footnote ⁶ Beyond the archaic-sounding wordings of French law, this semiotic ideology in which persons are made and unmade by relational signs is widespread and commonplace in a range of Western liberal understandings of verbal injury (see, for instance, Post Reference Post1986 on libel; and Herzog Reference Herzog2017 on defaming the dead).

Indeed, the reference to sworn statements points to another way in which the semiotic ideology of honor pervades the caricatures case. Both in their appeals to the court and in their broader dissemination of their cases, Charlie Hebdo’s lawyers and journalists insistently framed the affair as involving bravery, risk, and the dangers of speaking one’s mind. The decision takes up these themes intertextually as part of its outline of the context of the case (soberly titled “a reminder of the facts”). There, the caricatures decision recalls that the initial publication by Jyllands-Posten had been motivated as a reaction to “self-censorship” by those who feared to write about Islam in the wake of the murder of Theo Van Gogh. The decision quotes a passage from the special issue, written by the journalist Caroline Fourest, in which the republication of the drawings is justified in the following terms: “The newspapers which ‘dared’ publish the caricatures of Muhammad are now threatened with reprisals, as are the states and their citizens who are considered as accomplices of the blasphemy. Faced with this tidal wave of violence, Charlie tries to analyze this polemic and its consequences. A way of showing that freedom of expression must be stronger than intimidation” (ibid.). This framing evokes a familiar vision of free speech, and one which rejoins Foucault’s description of parrhesia, “the courage of truth”: “Verbal activity in which a speaker expresses his personal relationship to truth, and risks his life because he recognizes truth-telling as a duty … the speaker uses his freedom and chooses frankness instead of persuasion, truth instead of falsehood or silence, the risk of death instead of life and security, criticism instead of flattery, and moral duty instead of self-interest and moral apathy” (Reference Foucault, Davidson and Burchell2011: 12).

Such rhetorical articulations of “the courage of truth” to personal integrity and bravery are commonplace in contemporary debates over freedom of speech and have a deep history (Colclough Reference Colclough2009). What is rarely noted, however, is that this characterization of free speech as parrhesia harks to a vision of persons, language, and truth which is at odds with the standard view of liberal language ideology as a matter of civil Habermasian public debate between detached rational actors. Far from being mere “opinions” or “information,” words and images in this view are a token of a person’s integrity, an integrity which must be tested and demonstrated in risky engagements with others. This vision of bravery in the face of threat and of speaking the truth at all costs is a far more natural conceptual bedfellow to the universe of dueling, insult, and honor than to the peaceful bourgeois imaginary of a conciliatory “marketplace of ideas.” These two modalities of signs—the incalculable excess of honor and the reasonable claims of the marketplace—have been in a complex and tense relationship since the nineteenth century in French visions of the press (Berenson Reference Berenson1992; Nye Reference Nye1998).

How, then, does the ruling succeed in subordinating this vision of language as performative medium for courage and integrity, insult and relational damage, to the more Habermasian semiotic ideology indexed by the outcome of the decision? Before answering this question, let us pause to explore the contours of a third semiotic ideology interwoven into this ruling.

Carnival: Antinomian Excess and the Democratic Function of Satire

As noted, one of the most often quoted passages of the Charlie Hebdo decision is its definition of the genre of caricature as essentially a contribution to democratic debate—a straightforward piece of Habermasian representationalism. And yet, the very need to explicitly define caricatures in this way indexes the fact that this vision of caricature, while potentially persuasive, does not “go without saying” in contemporary France. Indeed, the performative vision of language as a medium for honor and insult—outlined in the previous section—was not the only strand which challenged the dominance of Habermasian representationalism in this decision. A third strand peeks through in the alternative vision of caricature and laughter which the ruling is so keen to defuse. This is a vision of laughter, and free expression more generally, as nihilistic antinomian excess.Footnote ⁷

In his perceptive historical study of understandings of censorship in France since the eighteenth century, literary theorist Nicholas Harrison distinguishes two strands in French visions of freedom of speech (Reference Harrison1995). The first is the familiar mode I have described here as Habermasian: a reasoned defense of freedom of speech within legal limits as a democratic good. The French constitution’s protection of “free communication of thoughts and opinions … while remaining answerable for any abuse of this freedom in the cases determined by law” is a good example of this former genre. The other is an attitude Harrison characterizes as “anti-censorship”—a generalized refusal of any form of limit, bound to an aesthetic of excess. As an exemplar of this other strand, Harrison takes the Marquis de Sade, pornographer and radical républicain for whom, famously, “philosophy must say everything” (“la philosophie doit tout dire”) (cited in ibid.: 205). While Sade’s pornographic, violent, and misogynistic opus may be seen as an outlier, Harrison argues persuasively that “anti-censorship,” with its twinned impulses towards totality and excess, can be traced throughout the history of French literature and theory. It surfaces in the productions of the surrealists, influenced by the Freudian vision of the liberating power of anti-censorship. One finds the echo of Harrison’s “anti-censorship” renewed in the antinomianism of May 68, with its famous slogan “it is forbidden to forbid” (Bourg Reference Bourg2017). The “spirit of May 68” is also an important part of the symbology of Charlie Hebdo, a journal which appeared in its current form in 1970, when its predecessor was banned after mocking other newspapers’ obsequious coverage of the death of Charles de Gaulle—although the censorship was officially justified on the grounds of the “pornographic” character of some of its drawings.

Harrison’s contrast between a civic-minded defense of free speech and an antinomian “tout dire” is a useful guide for tracing some of the tensions inhabiting the position of Charlie Hebdo and its supporters. At times, as in Caroline Fourest’s quote above, the publication of the cartoons was cast in terms of a deadly serious defense of “republican values” and the French public sphere. And yet this seriousness clashed with the self-professed irreverence, antinomianism, and generalized disrespect for all forms of authority of the publication whose tagline to this day remains “irresponsible newspaper” (“journal irresponsable”). Tensions between public-spirited seriousness and antinomian “fun” are highly evident in the documentary that covered the trial (Leconte Reference Leconte2008), essentially from the perspective of Charlie’s supporters. Early on, Charlie Hebdo journalist Riss is interviewed explaining the serious democratic purpose behind the cartoons. Acknowledging that laughter can be seen as a form of denigration or insult, Riss explains, deeply seriously, that this was never his intent as a cartoonist. Why was he laughing? “Well, this may seem surprising, but I was laughing to tell Muslims: you’re a part of French democracy.”

In line with this somewhat strained vision of laughter as a device for republican integration (or forceful incorporation), throughout the caricatures trial the defendants and commentators sympathetic to them were mostly careful to cast the court as an ally in what they portrayed as the journal’s democratic struggle for freedom of speech and against the breakup of the public sphere. Yet there were also moments when the authority of the court, the law, and the judges seemed to be the target of Charlie Hebdo’s antinomian bent. Editor Philippe Val and other commentators occasionally expressed sadness and shock at the fact that such a matter could come to court in the first place. The tension between the authority of the court and the distinctively antinomian “spirit” of Charlie Hebdo came across strikingly in a moment of the documentary where Val gleefully recounts that, when the judge tried to limit the number of witnesses which his side could cite in support, Val reported with a chuckle that he instructed his lawyer to tell the judge “to go fuck himself.” As Bourg notes, “Antinomianism does not mean simple transgression; it also suggests the rejection of norms seen as restrictive in favor of a redemption that will come “outside the law” (2017: 7).

These oscillations between antinomian laughter and republican seriousness speak of a deeper struggle within liberal semiotic ideologies of humor. Harrison’s account of “anti-censorship” bears a striking parallel to Bakhtin’s influential characterization of the carnivalesque as an inherently anti-hegemonic force, forever upending the pretensions of power, mixing the crudely sexual with a life-affirming force. Bakhtin’s vision of the inherent powers of laughter taps into elements of liberal commonsense which dovetail with aspects of Charlie Hebdo’s defense. In their depiction of the stakes of the case as a confrontation between the irrepressible right to laugh at everything and the serious sternness of religious censorship, Charlie Hebdo’s supporters might as well have quoted Bakhtin’s claim that “festive folk laughter presents an element of victory not only over supernatural awe, over the sacred, over death; it also means the defeat of power, of earthly kings, of the earthly upper classes, of all that oppresses and restricts. Medieval laughter, ‘when it triumphed over the fear inspired by the mystery of the world and by power, boldly unveiled the truth about both. It resisted praise, flattery, hypocrisy’” (Reference Bakhtin1984: 116).

Yet as Charlie’s critics then and since have been quick to point out (Trudeau Reference Trudeau2015), and as countless academic critics of Bakhtin over the past four decades have ceaselessly repeated, “For every laugh in the face of autocracy, there [is] another laugh by the powerful at the expense of the weak” (Beard Reference Beard2015: 6). This stands as an important reminder that, as a rejection of laws and norms, antinomianism is not necessarily the preserve of “the weak.” It can just as easily be wielded by those in a position of structural power who feel, rightly or wrongly, that laws and norms constrain them.

The key point is that, irrespective of which side one is on in these liberal debates, visions of laughter as anti-authoritarian, and visions of laughter as oppressive “punching down,” share a sense that laughter is fundamentally about power—a type of force, an untamed energy, jouissance, or vital flow. This vision does not straightforwardly resolve into the image of the cartoons as a matter of communication of opinions and ideas. While it shares with the semiotic ideology of Honor a concern for the performative effects of expression, the high seriousness of the language of insult, bravery, and reputation does not sit easily with the crude antinomianism of the “irresponsible journal.”

Part II. Varieties of French Jurisdiction

Having teased out the multiplicity of visions of language indexed in the French re-run of the caricatures controversy, I will focus in this part of the article on the techniques through which the ruling manages to reduce or distill these multiplicities down to one single, broadly Habermasian conclusion. This section builds on Justin Richland’s call to attend to what he terms “juris-diction” (2013), or the ways in which legal language grounds the force of the law. In line with the broader theme of this article, moreover, a key point is that French law is itself internally riven between different visions of the nature of legal language, different modes of juris-diction. While the substance of the caricatures decision reflected on and proposed to adjudicate a question pertaining to religious authority and individual freedom, the form of the decision implicitly reflected a particular attitude to the question of legal authority. It spoke to an enduring set of concerns in French legal history over the source and nature of the power of judges, and the relationship between speech and authority in legal decisions.

Democratic Context and Imperious Brevity

In tidying up the various visions of language outlined above, the ruling makes ample use of that most powerful move: the appeal to context. Since humor, as we have seen, can veer dangerously close to a sheer antinomian force, the ruling finds it necessary to contextualize the drawings with an explicit definition of the genre of caricature. The often-quoted definition of caricatures, in full, reads as follows: “Considering that any caricature can be analyzed as a portrait which breaks with the rules of good taste in order to fulfill a parodic function; that exaggeration functions in the same way as a witticism, which allows one to evade censorship, to use irony as an instrument of social and political critique, by eliciting judgment and debate; … Considering that the literary genre of caricature, albeit intentionally provocative, is therefore a part of freedom of expression and communication of thoughts and opinions” (TGI Paris 2007).

In cases such as these, there seems to be, as Peters puts it, “an under-the-table transactional ethic in the free speech story, a curious coupling of straitlaced defenders of liberty and wacky or wicked pushers of limits” (2005: 16). Reason’s acknowledgment of the value of Carnival, however, is also a functional subordination. Note the careful framing through which the ruling subordinates humor to a functionalist explanation (“a parodic function,” “functions in the same way as,” “an instrument”), in order to deduce that caricature as a genre is, after all, not merely antinomian but rather an aid to (if not straightforwardly an instance of) democratic debate and does as such fall under the legal protections of freedom of speech.

Similarly, the claims of Honor and insult are carefully considered, as we saw above, but they are ultimately defused through an appeal to “context.” Having granted that the third (“bomb”) drawing is “of a nature to insult all of the adepts of [the Muslim] faith,” the ruling nevertheless states that this drawing “cannot, from the perspective of penal law, be appreciated independently of the context of its publication.” This is where the decision cashes in on the three and a half pages of “context” introduced earlier under the bland heading, “A reminder of the facts.” In that section, the publication of the drawings in Charlie Hebdo is emplotted in a longer narrative beginning with a resistance by Jyllands-Posten to the perceived threat of self-censorship among journalists, brought on by the murder of Theo Van Gogh. The passage reports the violent protests sparked internationally by the initial publication. “It is in these circumstances,” the decision continues, that Charlie Hebdo published its special issue about the “Muhammad caricatures” on the 8 February 2006. This reminder of “the facts” then zooms in on the internal contextualization of the pictures within the special issue, quoting at length from the article by Caroline Fourest which explains the re-publication of the drawings as part of an effort to “analyze the polemic and its consequences” as well as a performative, demonstrative defense of freedom of speech. Based on this “context” and on a detailed consideration of its size and placement in the journal and other texts printed alongside it, the decision therefore asserts that even the third picture, however reprehensible in itself, “can only be seen as participating in a reflection on a debate of ideas” and “in a public debate of general interest.” It cannot therefore be characterized as an insult which justifies a limitation of the right of free expression. As one of the lawyers for the plaintiffs noted with some satisfaction, “What I retain is that the court has clearly stated that this type of caricature could be condemned, and that in another context, it would be” (quoted in Kiejman and Malka Reference Kiejman and Malka2019: 22, my emphasis).

Critics will be quick to note all that is occluded in this seemingly plain and clear presentation of “the facts.” The way in which the decision contextualizes the Danish cartoons affair follows almost step by step the way in which this affair was characterized by the defendants throughout the trial. From the perspective of the plaintiffs, one might ask why this story begins where it does—with the murder of a film maker—rather than with, for instance, the prevalence of anti-Muslim sentiment in early twenty-first-century European societies, socioeconomic discrimination against Muslim citizens in France, or the history of French colonialism (see Tønder Reference Tønder2013: 113; Scott Reference Scott2023: 4). The fact that these are not considered elements of relevant context is a pre-judgment of the issue. The power to set what counts as relevant context and how an event is emplotted in a broader narrative is a classic device of the seemingly abstract, or indeed “pedagogic,” deployment of legal speech.

The most striking fact about the Charlie Hebdo ruling, however, is likely to pass unnoticed for readers accustomed to British or American common-law rulings. This is the extent to which the ruling engages in an exposition of context and argumentation at all. Vincent Crapanzano expresses a widespread ideal type when he writes that, unlike U.S. rulings, “French decisions are terse, rarely more than a sentence, expressed in syllogistic form; the arguments that lie behind them have been debated in camera, the rapports of the reporting judge and the conclusion of the advocate general, are not public. They support a transcendent, universalizing picture of the law as a purely rational enterprise from which the Law garners its authority and independence” (Reference Crapanzano2001: 242).

Strikingly at odds with this model, the caricatures decision stretches to twenty-three pages in folio, including three and half pages setting out the context of the “Danish cartoons affair” and a further eight pages on motivation. This may not seem much by the standards of, say, a U.S. high court ruling, but in the French context this is significant and distinctive. This was not lost on French commentators, many of whom were captivated not simply by the outcome of the ruling but also by its form. Legal scholar Camille Viennot, writing some years later, praises “the pedagogy shown by the 17th Chamber … in outlining both the social and the juridical stakes of the so called ‘Muhammad caricatures’ affair. It is therefore regrettable that only the dispositif of this decision [the final lines which state the acquittal] has really been publicized beyond juridical circles, without further analysis of the motives which led to the acquittal” (2015: 281–82). Viennot was slightly overstating the case. In an article published on the day of the decision, which reprised key quotes from the decision, le Monde’s legal reporter Pascale Robert-Diard gave a blow-by-blow account of the main steps of the decision’s reasoning. Robert-Diard hailed the decision as “beautifully crafted both juridically and politically,” enthusiastically concluding: “Isn’t law beautiful?” (Robert-Diard Reference Robert-Diard2007).

To understand what was so “beautiful” about the caricatures decision for these commentators, we need to take a step back and consider the changing form and expectations of French legal decisions more broadly. Crapanzano’s quote above provides an accurate portrayal of a certain classic view of French judgments, which are expected to trace a plain and simple logical deduction, not a sophisticated, narrated, and argued piece of legal reasoning. This is most clearly visible in the terse judgments of the cour de cassation,Footnote ⁸ which, as legal scholar Ruth Sefton-Green notes, “take the form of a juridical syllogism which expresses an irrefutable logic: The solution is derived from a major proposition (the legal rule), followed by a minor proposition (the facts of the case) and an often abrupt conclusion” (2011: 92).

Sefton-Green argues persuasively that the traditional brevity of French legal judgments encodes a different relationship between language and authority to that often described for common-law systems. Common-law judgments seek to persuade by showing justice to have been done (Chainais Reference Chainais2014: 254; Sefton-Green Reference Sefton-Green2011: 93)—which may include spelling out dissensions between judges, hesitations, and counterarguments for all to see. By contrast, Chainais, quoting Arendt, characterizes the French tradition of brevity in terms of an opposition between authority and justification—a truly authoritative source does not need to persuade or explain itself (2014: 247). From that perspective, as Crapanzano puts it, “To those used to codified law, such as French law, American common law seems a strange bird. The judicial opinions that serve as precedent seem wordy, digressive, often contradictory, self-justificatory, revealing, at times to the point of confession (always an American obsession), of the debates, the anguish, the indecision that lie behind them, and thus the vulnerability. It is as though they undermine the law, or rather the Law, at the same time as they produce it” (2001: 242).

As the caricatures ruling itself demonstrates, such ideal-typical distinctions between common law and civil law systems are partial and simplistic. And yet they reflect distinctions drawn by actors within these systems. French judges, lawyers, and commentators deploy these distinctions between archetypal French brevity and common-law loquacity self-consciously and critically to situate themselves. Much as lawyers themselves like to project them outward, onto a contrast between French and common-law systems, the two visions of the relationship between language and authority are also internal to French Law.

The tension harks back to the complex history of French legal decisions (Texier Reference Texier2000). In the pre-revolutionary period French judges were perceived to have great latitude in interpreting laws and, in effect, legislating from the bench. Pre-revolutionary judges also considered it a formal privilege of their status that they were not technically required to justify their decisions. Like many such privileges, it was the object of mounting resentment in the pre-revolutionary period and was self-consciously abolished in the revolutionary restructuring of the French justice system. This involved an attempt to cut judges down to size, redefining them as relatively modest and interchangeable civil servants, merely applying codified law to particular cases. An express obligation to justify legal decisions was formally introduced in 1790, as part and parcel of a self-conscious effort to strip judges of their erstwhile numinous authority and make them accountable to the public. Chainais (Reference Chainais2014) notes that, to this day, the extent to which French courts justify and explain their decisions tends to be in inverse proportion to their sense of their own authority. At one end of the scale, the Conseil d’état, which is both an administrative high court and the descendant of the Sovereign’s “privy council,” has been notorious for the “imperious brevity” of its judgments—sometimes literally one-liners. Lower-level courts, by contrast, have historically been more prolix, taking seriously their duty to make their judgments accountable to the public they serve.

The historical background I have sketched above underpins a broad sense of the contemporary among French legal commentators. This sense is that a traditional civil law vision of judges as mere civil servants applying a fully codified law, is now giving way to a system in which jurisprudence and precedent play a greater role, at least informally, on the model of common-law systems—notably under the unifying pressure of international instances such as the European Court of Human Rights. Thus, legal scholars Chainais and Sefton-Green encourage and welcome what they see as a growing impact of common-law understandings of the nature and authority of judgments on French jurisdiction. The narrative, dialogical conception of judicial authority which they perceive in common-law judgments strikes them as more “transparent” and “democratic,” as well as more “pedagogical.” The old-style French imperious brevity emerges from their writing less as authoritative, and more as arbitrarily authoritarian (Sefton-Green Reference Sefton-Green2011). For such commentators, longer, more self-consciously justified decisions speak of a shift towards greater democratic accountability. Their contrast recalls the equally normative contrast drawn by Crapanzano between, on the one hand, the “literalism” of conservative U.S. high court judges and Christian fundamentalists, alike serving the Word with “a strict commitment to what is taken to be “literal” or “true” meaning,” and on the other, the more “open interpretive styles” characteristic of liberals (2001: xvi).

Within this context, the caricatures decision, with its extensive justifications, pages of context, and careful consideration of points and counterpoints, stands as the epitome of the type of shift Chainais and Sefton-Green are calling for. The ruling’s appeals to context are thus not an unmarked device—they are themselves recognized by other French commentators as a sign of a distinctively democratic discursive approach to judgment, one which breaks with the older French style of imperious brevity. Contextualizing, in other words, is itself held up as a valuable democratic practice.

Once we see this, we can see also a powerful recursive dynamic in the ruling. For in contextualizing Charlie Hebdo’s republication of the cartoons, the ruling is casting the latter as, in essence, a contextualization. The “bomb” cartoon in particular, identified by the ruling as “in another context” truly reprehensible, is saved by the fact that it is carefully contextualized by the editors of Charlie Hebdo: it is reproduced “in a very small format,” amidst other caricatures and on the same page as an editorial which clearly explains the purpose of the journal in republishing these cartoons. Most of all, “the drawing in question, which is only the reproduction of a caricature published by a Danish newspaper, is included in a special issue whose cover “editorializes” the totality of its content and serves as a general presentation of the position of Charlie Hebdo” (TGI Paris 2007).

Here we see again the features of a representationalist language ideology identified by Keane, Asad, and Mahmood: the cartoons’ capacity to harm is subsumed within questions about interpretation, context, and the intention of the authors. But the key point is that in this extensive contextualization of the cartoons, the ruling is not merely struggling against powerful alternative semiotic ideologies at the heart of contemporary French liberal secular space—it is also performatively setting itself against an alternative vision of French legal language, in which context and argument do not matter, and the judge’s job is merely to apply the numinous letter of the law. When Pascale Robert-Diard in le Monde or legal scholar Camille Viennot praise the judgment’s “pedagogical” exploration of social and juridical context, or the “subtlety of the argumentation” in relation to the third drawing, they are echoing a broader sense of what makes this judgment feel exceptional. Within the broader setting of French decisions, the form of the caricatures judgment itself indexes a self-conscious stance towards law’s authority: that it needs to be carefully and extensively justified through argument. As I outlined above, this stance is not anodyne in France but rather aligns with at least some people’s broader sense of the “direction” in which French law-as-Justice (Constable Reference Constable2014) should be going: away from arbitrary authoritarianism and towards democratic debate.

Writing about another notorious area of French legal dispute over Islam—veiling in public—anthropologist Mayanthi Fernando argues that the seeming aporia faced by French Muslim women who claim that veiling is both a choice and a religious obligation is actually an aporia internal to the secular liberal project itself, which the state in effect displaces onto Muslim subjects. Fernando writes that “criticism of the religious authority and normativity of Islam … entails a deferral of tensions inherent in a secular-republican project that ostensibly emancipates individuals from various forms of authority (the church, custom, etc.) by bringing the normative disciplinary authority of the school to bear on its subjects” (2010: 30).

Fernando’s point is perceptive, and its relevance to our case striking. The caricatures trial was also, in important ways, an extended reflection on the proper relationship between authority and autonomy in the Republic. Firstly and most obviously, the defendants and the large majority of French public commentators on the trial framed the entire case as one which opposed Islamic religious authority to the freedom of speech of individuals in a democracy (Boe Reference Boe, Titley, Freedman, Khiabany and Mondon2017). But the trial can also be read as a reflection on the tension between authority and autonomy at the heart of French law itself. The substance of the caricatures decision—cast as a vindication of the importance of public debate in the face of religious authoritarianism—is echoed also in the form of the judgment itself, as an extended, carefully contextualized, precise argument.

Implicit Jurisprudence and Sociological Generalization

Yet, for all its commitment to explication, context, and discussion, there was one crucial zone of silence in the caricatures decision, which might strike an observer unfamiliar with French legal decisions as particularly odd: the decision cited no jurisprudence. This is particularly surprising since it silences one of the most obvious legal “contexts” of this ruling: in France as in Britain and elsewhere in Europe, the 1990s and 2000s saw the sustained rise of religious legal activism, particularly by Christian and Muslim organizations, which contested the secular state’s attempts to “remake religion through the agency of the law” (Asad Reference Asad, Asad, Mahmood, Butler and Brown2013: 87), by using the secular state’s own legal instruments (Favret-Saada Reference Favret-Saada2017; see also Scott Reference Scott2023: 10–12; McIvor Reference McIvor2020). A rich and complex European jurisprudence has developed as a result around the disputed difference between the problematic of blasphemy (insults to God), and the problematic of hate-speech (insults to believers on account of their faith). As Scott shows in his study of India’s article 295A, the two problematics are difficult to tease apart legally, philosophically, and sociologically: the self-consciously secular Indian law against religious insult remains “blasphemy-adjacent” (Scott Reference Scott2023: 7), while in the UK the introduction of provisions against religious insult in the 2006 Racial and Religious hatred act made blasphemy [law] redundant (ibid.: 12). In French law, blasphemy has not been an offense since 1881,Footnote ⁹ but first Christian and then Muslim legal activists have for the past forty years campaigned for the recognition of wounded religious sentiments in ways which their critics at least denounce as blasphemy-adjacent (Favret-Saada Reference Favret-Saada2017). The result has been a spate of cases, many of them heard by this same court, as well as appeals and high court decisions, all parsing and probing the problematic distinction between blasphemy and religious insult. As we shall see, the caricatures ruling took up a very self-conscious position within this jurisprudence, but it did so “silently.”

The caricatures decision, in the brief consideration of the legal motives for the decision (one and a half pages out of the total twenty-three), does not explicitly mention any jurisprudential precedents, but only points to two codified texts: the French Press Law of 1881, and article 10 of the European Convention on Human Rights. Additionally and strikingly, this section focusing on legal motives includes what looks at first glance like a bit of sociological generalization: “Considering that, in France, which is a secular and pluralist society, the respect of all beliefs goes together with the freedom to criticize any and all religions […] that blasphemy, which insults a divinity or religion is not punished, by contrast to insult, when this entails a personal and direct attack, directed against a person or group of people by reason of their religious belonging” (TGI Paris 2007, my emphasis).

Legal scholars, however, note that the decision is in fact steeped in a careful consideration of jurisprudential tendencies concerning religious offense. As Camille Viennot points out, in some cases the caricatures decision even reproduces the language of previous decisions without citing them. For instance, the formulation I quoted above borrows directly and without citation from a decision published the previous year by the Cour de cassation. The Cour had judged that a parody of Da Vinci’s Last Supper, incriminated by a Catholic organization, “did not entail an insult, a personal and direct attack directed against a person or group of people by reason of their religious belonging” (quoted in Viennot Reference Viennot2015). Similarly, the caricatures decision replicates, without citing it, the language of the famous 1976 Handyside ruling of the ECHR. The latter had affirmed that freedom of speech “is applicable not only to ‘information’ or ‘ideas’ that are favorably received or regarded as inoffensive or as a matter of indifference, but also to those that offend, shock or disturb the State or any sector of the population. Such are the demands of that pluralism, tolerance and broadmindedness without which there is no ‘democratic society’” (ibid.).

The caricatures decision notes—again without any explicit mention of the earlier precedent—that Freedom of speech “is applicable not only to ‘information’ or ‘ideas’ that are favorably received or regarded as inoffensive or as a matter of indifference, in a given society but also to those that offend, shock or disturb, as required by the principles of pluralism and tolerance which are particularly important in a period characterized by the coexistence of many beliefs and faiths within the nation” (TGI Paris 2007).

What to a common-law scholar might look like reprehensible acts of plagiarism, actually highlights a distinctive form of juris-diction (Richland Reference Richland2013) central to French decisions. In refusing to acknowledge jurisprudence, the decision conformed to the legal fiction according to which French judicial decisions derive directly from written law. The fact that this is a fiction is eminently clear to anyone familiar with French legal worlds. As John Bell (Reference Bell2001) points out, the sometimes-impenetrable brevity of French court judgments operates symbiotically with a slew of doctrinal discussion, interpretation, and explanation. While French judgments self-consciously avoid interpreting and commenting on precedents, let alone establishing them, this work is complemented by doctrinal writings of legal scholars who pore over, excavate, and contextualize jurisprudential tendencies, as Viennot does here. The effect of this division of labor between judgment and doctrine preserves the fiction that judges are not making the law but rather applying it while, at the same time, and in full view of the French legal community, judges build and reflect on existing jurisprudence in often sophisticated ways.

One of the obvious effects of this practice is that juridical decisions are, as it were, two-faced. On the one hand, to the legal community, they index subtle and often fascinating shifts and continuities in an ongoing jurisprudential conversation. In the case of free speech law, this conversation is extremely technical and intricate, and also the preserve of a small, tightly knit community of experts among whom the judges and lawyers of the 17th chamber play a central role. Thus, in the caricatures case lawyers for both the plaintiffs and for the defense, in their closing statements, painstakingly traced the relevant jurisprudence with more than a little technical delight.

To the broader public, however, this conversation is, broadly speaking, invisible. Each decision, by not explicitly situating itself in terms of jurisprudence, stands as an embodiment of the vision of the application of written law, and beyond that, of philosophical, cultural, or sociological principles presented as self-evident. The distinction between an insult to people by reason of their faith and an attack on people’s religious sensibility, for instance, appears as if it were a self-evident matter “in France.” Yet the Cour de cassation whose distinction the caricatures decision is plagiarizing here was itself reversing a decision by the court of appeal of Toulouse, which had incriminated the parody of the Last Supper, judging that it could have “been experienced as an offense by the Catholic community by reason of its beliefs and practices” (quoted in Viennot Reference Viennot2015). A live legal debate over the nature of offense, belief, and the experience of religion is thus occluded, and rendered as a single, self-evident pronouncement.

Conclusion: Counting to Three

The 2007 Charlie Hebdo ruling came back to the forefront of French public consciousness in the wake of the 2015 attack on the journal’s offices, in which two members of Al Qaeda murdered ten cartoonists and two policemen in explicit retaliation for “insulting the prophet.” Predictably, this shocking event relaunched discourses about a “clash of civilizations” between the “liberal West” and “archaic Islam”. In response, anthropologists might be tempted to dust off the critical counter-generalization of a clash between a liberal semiotic ideology in which signs are mere opinions exchanged by independent transactors and an alternative, Islamic semiotic ideology in which signs can deal relational damage to persons. And yet an alternative comparative move might have identified in this event a grotesque echo of some home-grown French struggles over violence and limits of language, freedom, and the press. Current French press law was designed to regulate the respective claims of semiotic injury and public democratic expression. The 1881 law, the bulk of which is still in place, was established at a time when newspapers had fencing rooms installed in their offices in order to prepare journalists to defend their words at the point of a sword (Nye Reference Nye1998; Whitman Reference Whitman2000; Candea Reference Candea2019). Three decades later, while Europe was on the verge of a world war, France’s newspapers thrilled the nation for weeks with coverage of the trial of Mme Caillaux, who had murdered the editor of the leading newspaper Le Figaro in his office in retaliation for a smear campaign against her husband (Berenson Reference Berenson1992). The contexts that we remember matter.

I have argued in this article that liberal visions of freedom of speech are not reducible to a mere misunderstanding of the real power and effects of representation. The first part explored three interwoven modalities in which freedom of speech is envisioned in debates such as those surrounding the “Muhammad cartoons.” In the second part of the article I moved from these models of free speech to an examination of the pragmatics of legal language as it seeks to adjudicate such instances of “metapragmatic struggle” (Keane Reference Keane, Candea, Heywood, Fedirko and Wright2025). In their own internal struggles over what can be said and what must be left out of a written judgment, about the limits of “context” and “explanation,” judges are themselves playing out different visions of the relationship between the performative force of law and public Habermasian debate. In this respect, for all its peculiarities, French law remains a species of liberal secular law, riven, as Fernando (Reference Fernando2010) shows, by internal contradictions over the relationship between authority and freedom.

Scaling up from this particular case, the broader point is that anthropologists are not alone in critiquing a “vision of language purified of morally constitutive effects or independent of social ties” (Keane Reference Keane2009: 70)—critiques of and alternatives to this vision are rife in the very fabric of liberal public and legal debates over freedom of speech, in France and elsewhere (see, for instance, Kramer Reference Kramer2011, Peters Reference Peters2005; Candea Reference Candea, Candea, Fedirko, Heywood and Wright2025).

As I suggested in the first part of this article, alongside the avowedly representationalist intuitions of a Habermasian defense of freedom of speech within limits, two other structures of plausibility for free speech can be teased out of this case, which may have wider comparative purchase.

One of these, which I have indexed to Carnival, stakes no particular claim to democracy, rationality, or progress. Its main wellspring is oppositional, antinomian—this is first and foremost anti-censorship, a rejection of settled boundaries, authorities, orders, and structure. Silencing is censorship by definition, irrespective of its source or its legal or expert credentials. Publicity, privacy, and context are not central themes of concern. The paradigm for “speech” here is not the communication of contents, but rather flow, vital force and jouissance. Speech is “expression”—it is an activity, a doing, often most visible when it comes as an upending, a reversal, a desacralizing. Truthfulness, meaning, or sincerity are not necessary—sarcasm, irony, and cynicism are fine. The authority of reason is no more acceptable here than that of good taste or civility. This is the speech of a lawyer telling a judge to “go fuck himself.” While on the face of it defined by the denial of any kind of limit, this mode of freedom of speech does enable, and even structurally requires, certain exclusions. In order to be recognized under this mode, expression must be new, challenging, and vital. This mode provides no language in which to defend the conventional or the half-hearted. Indeed, this mode of freedom of speech often comes with a hint that expression is a zero-sum game—to speak is to silence someone else. Silencing the voices of convention and conformity is thus precisely the point.

So far we have only counted, again, to two. But the struggles within this ruling, between authority, insult, and explanation, point the way to a third structure of plausibility. In this one—let us call it Honor—speech is the word one stands by: speech is promise, testimony, insult, critique, and exposure. This is grounded neither in a hope for the rational benefits of public communication nor in a burning desire to resist censorship and convention, but in a special relationship between truth-telling and personhood. This commitment can manifest as a disregard for merely human social conventions, public-private distinctions, and the calculations of propriety, but it has no truck with rule-breaking for its own sake, sarcasm, or nonsense. This is freedom of speech envisaged, as in Foucault’s account of parrhesia, not merely as a right but also as a duty. This is the free speech of prophets and martyrs, both religious and secular, but it is also, on a less dramatic scale, the free speech of activists, of witnesses, and more broadly of persons “of good faith”—“honest” and “honorable” persons, persons who are characterized by a special commitment to truthfulness and draw an entitlement from it. In the caricatures decision, this structure of plausibility encodes the role of the witnesses at the trial, and those of Charlie Hebdo caricaturists insofar as they can be assumed to be sincere. But it is also the language of insult which attends to words’ ability to cause relational damage, and the necessity for responsibility and reparation. It is the language in which the claims of the injured party can be heard, at least transiently and faintly.

In defense of his ambitious project of teasing out fifteen modes of existence, Bruno Latour perceptively noted,

With Latour’s blessing, one might attempt to systematize the above argument “by way of a chart” (see table 1). The point is not to exhaust the possibilities—as if three were somehow the magic number! The point is simply to resist the inexorable pull of dualism.

Table 1.

Three modes of free speech.

Sometimes these structures of plausibility can be made to fall into alignment. Reason says, “Free public communication normally leads to the greater good.” Carnival says, “I will not be censored.” Honor says, “I will speak the truth whatever the cost.” The three statements may seem entirely in agreement. As Boe notes, in some of the discourse surrounding the caricatures case, “humor and satire are linked to integration and to belonging, thus becoming essential markers of a new “we,” characterized by “humor and democracy,” in opposition to a “them” characterized by “dictatorship and violence” (Reference Boe, Titley, Freedman, Khiabany and Mondon2017: 175).

Other scholars have observed the paradoxical yet deeply effective recombination of the language of honor and dueling with liberal conceptions of the public sphere—from Kant’s strange foray into dueling as a marker of public morality (LaVaque-Manty Reference LaVaque-Manty2006), to the deep interweaving of dueling with French bourgeois consciousness in the late nineteenth century. Antinomianism can also claim the mantle of sincere truth-telling. Bakhtin’s romanticized vision of Carnival as simultaneously antinomian, democratic, and truth-telling is an example of an attempt to fuse these three structures of plausibility together. Charlie Hebdo’s once undisputed appeal in France as an icon of freedom of speech can partly be explained in a similar vein by the ways in which it drew on and wove together these three structures of plausibility.

Yet such alchemies are unstable, as critiques of Bakhtin—which “run into thousands” (Beard Reference Beard2015: 234)—have endlessly pointed out. Tensions between these structures of plausibility surface inexorably. Reason will see in Carnival’s outpourings mere gibberish which is not deserving of support and the posturings of Honor will seem like a dangerous extremism for which there is no place in the public sphere. Antinomian Carnival clearly has little truck with the pieties of the other two modes, and Honor can be as dismissive of the bloodless proceduralism of reason as of the antinomian rantings of Carnival. Reflecting on the caricatures trial in the immediate wake of the 2015 murders, Charlie Hebdo cartoonist Luz highlights the risk of performative contradiction when Charlie is “taken seriously”: “The media made a mountain out of our cartoons, when on a worldwide scale, we are merely a damn teenage fanzine.… Since the cartoons of Muhammad, the irresponsible nature of cartoons has gradually disappeared.… Charb believed we could continue to overcome taboos and symbols. But today, we are the symbol. How can you destroy a symbol when it is yourself?” Conversely, in identifying these internal tensions, we can begin to see the shape of accommodations and partial connections with visions of free speech which are otherwise all too easily cast as non-liberal or non-secular. Many of those who criticized Charlie Hebdo’s editorial decisions or pointed to the structural racism underpinning the caricatures debate understood themselves to be speaking their truth bravely, at personal risk, against the grain.

Plaintiffs and their lawyers in the caricatures case drew together appeals to wounded religious subjectivities with broadly Habermasian claims to keeping open a space of public dissension beyond republicanist pieties. Conversely, liberal concerns with honor and reputation speak a language in which some at least of the claims of wounded religious subjectivities can be heard. Moving in the opposite direction, Islamic pedagogies of public criticism (Bhojani and Clarke Reference Bhojani, Clarke, Candea, Heywood, Fedirko and Wright2025) might open up pathways for another type of accommodation with Habermasian ones, if the latter are decoupled from the temptations of antinomian iconoclasm. These are just tentative suggestions, but they all involve working against the grain of binarisms.

As I have argued throughout this paper, comparisons and contextualization have distinctive purposes and effects, whether they are those of lawyers, defendants, or anthropologists. Part of the work of judgments such as the caricatures ruling is to tidy up the multiplicity at the heart of liberal understandings of freedom of speech into one neat narrative, which takes us from the general vision of France as a national entity through to an obvious conclusion. We do not as anthropologists need to do the same: once we learn to count beyond two, richer comparative worlds, horizons, and hopes come into view.