Freedom of Expression

On 30 September 2005, the Danish newspaper Jyllands-Posten published 12 cartoons of the Prophet Muhammad (Rose, 2005). Placed under the heading ‘The face of Muhammad’ – and with a subtitle citing ‘freedom of expression’ – the cartoons were commissioned by the newspaper's editor, Fleming Rose, who wanted to challenge what he found to be an unreasonable fear of criticizing Muslims living in Denmark. Rose's efforts let to a diverse set of cartoons. The most notorious one was a cartoon of the Prophet with a detonated bomb in his turban. Another showed him as a next-door-neighbor circled by a half crescent suggesting both a halo and horns. Still another rejected the idea behind the Rose initiative and instead drew the Prophet as a seventh-grader at Valby Skole, a school outside Copenhagen known for its many Muslim immigrants. Representing a politics of resistance and subversion, the schoolboy took on the Jyllands-Posten journalists by pointing to a sentence written on the blackboard in Farsi: ‘The Jyllands-Posten journalists’, it said unapologetically, ‘are a bunch of reactionary provocateurs’.

Statements like this one anticipated the warlike conflicts that followed. Whereas Muslims in Denmark and around the world protested the cartoons, claiming they were harmed by them, others saw the cartoons as a legitimate exercise of the right to free speech. This article explores the framing of these positions in contemporary liberal theory. The objective is both to show the limits of liberal theory, in particular with regard to constituents who do not share liberalism's view of acceptable harm, and to discuss how these limits give us reason to supplement liberal theory with other resources from critical theory and phenomenology. So far the literature has framed reactions to the cartoons in terms of the legal limits of free speech versus sensitivity to harm. In an issue of International Migration devoted to the 12 cartoons, Tariq Modood argued that the cartoons should have been censured (but not censored) because they incited ‘racial hatred’ (Modood, 2006, p. 52). Opposing this argument, Randal Hansen and others did not see the cartoons as racial and insisted the cartoons were perfectly legal – if not the desirable outcome of ‘a liberal democratic framework’ (Hansen, 2006, p. 16; see also Post, 2007). Both positions expressed something essential about the liberal theory of free expression. One might even say that both Modood and Hansen were right: The cartoons showed we need more sensitivity to harm (which typically involves more regulation of free speech) and we need to be resilient and embrace disagreement (which typically involves less regulation of free speech). In this essay, I push past this binary to argue for a pluralization of the possible links between harm and law.

The motivation for this approach follows from the sense of intractability that characterized the first wave of interventions. Modood and Hansen saw their positions as mutually exclusive. And neither seemed interested in exploring each other's blind spots, setting the stage for a further bifurcation of the debate. On the side of the law, Kantians such Christian Rostbøll (2009) insisted that liberal theory can overcome its often unequal recognition of harm by basing freedom of expression on a Kantian conception of autonomy and not, as has been done historically, on a more stringent Millian character ideal that some argue democracies must promote if they want to secure free speech. On the side of harm, anthropologists such as Saba Mahmood expressed little faith in this solution, insisting that ‘the offense of the cartoons was not against a moral interdiction … but against a structure of affect, a habitus, that feels wounded’ (Mahmood, 2009, p. 78). The first theoretical responses to the cartoons thus reflected and may even have contributed further to the intractability that characterized the Danish cartoon war from the outset. Rather than explore the interplay between harm and law, theorists took sides and neglected the idea that to regulate harm through legal procedures is to use instruments that are themselves shaped by certain perceptions of harm.Footnote 1

Why this intractability and how might we negotiate it in ways that expand rather than shrink free speech? The answer, I suggest, lies in how liberal theory frames freedom of expression in terms of a choice between law and harm.Footnote 2 This framing has significant implications for how liberalism works in the context of deep pluralism. On the one hand, the frame enables liberalism to define a limited set of expressions as harmful and therefore to elide the diverse lived experiences of various groups. On the other hand, the frame veils liberalism's own contestability and cloaks the ‘creative instability’ that T.M. Scanlon attributes to liberal rights and that actually works to energize practices of free speech within and across social divides. Both outcomes encourage us to return to the plurality and energy intrinsic to the 12 Jyllands-Posten cartoons, and to explore what the liberal theory of free expression might learn from thinkers such as Merleau-Ponty and Deleuze who focus on the variable impacts and framings of free speech. The upshot, I argue, is a pluralization of harm and law that enables a self-critical theory of free expression, linking harmful speech to something like Scanlon's creative instability: To the affirmation of mutual contestation, social equality and respect for difference.

This, then, is the approach taken here: to rework the liberal theory of free expression from within, supplementing it with heretofore unappreciated resources of liberal theory and adding in elements of both critical theory and phenomenology. The first half of the article contextualizes the cartoon war before it develops a critique of liberalism as a frame that allocates the recognition of harm unequally. The second half focuses on the creative instability embedded in the liberal theory of free expression in order to propose another framing. I conclude with some remarks on liberalism and the future of free speech.

Harmful Cartoons?

No matter how liberalism construes the relationship between harm and law, we should be careful not to share Art Spiegelman's astonishment at ‘how banal and inoffensive the Jyllands-Posten cartoons’ should seem to anyone living in a liberal democracy (2006, p. 47). Such astonishment overlooks liberalism's power to frame visual experience, and may therefore blind us to the underlying processes that allowed the 12 Jyllands-Posten cartoons to become the object of a global ‘war’ causing the deaths of about 130 individuals (Cartoon Body Count, 2008). To avoid this mistake, we must broaden our perspective and analyze how and why the cartoons could seem harmful to some and not others.

One explanation might be Islam's prohibition of images. On both sides of the conflict, commentators have read this ban as accounting for the sensibilities of all Muslims who as a group were singularly wounded by the cartoons’ depictions, turning the debate into a question of whether liberalism could or should treat the ban as an acceptable limitation of free speech. But this reading overlooks how the ban is perceived variously by Muslims because it belongs to a more general hadith through which a chain of transmitters (called isnad) communicates texts (called matm) combining the words and deeds of the Prophet or some other religious authority. The hadith leaves the ban open to interpretation. Indeed, while the Qur’an includes the story of Abraham as an iconoclast – Abraham breaks the idols of a city because their depiction of god inspires false worship (Qur’an 21.51–70) – this has not ruled out the use of images in Muslim countries. First, images were used on coins in the second half of the seventh century, on official landmarks, and at places of worship such as the Dome of the Rock in Jerusalem and the Great Mosque of Damascus (Grabar, 2003, pp. 46–53). Second, in the exegeses of Ibn Kathīr and Sayyid Qutb, we learn that Abraham's iconoclasm represents not so much a ban on images as a warning against the way in which some images subvert the oneness of God through mockery and slander (Mirza, 2005, p. 425). Third, throughout the Middle East we find a tradition of political satire that uses cartoons as a means of critique and resistance (Göçek, 1998). Even though these uses only represent minority practices within the Muslim world, they nonetheless suggest an openness that draws Islam's ban on images into a more pluralistic world delimited by processes of politicization and retextualization.Footnote 3 Because liberals rarely attend to such experiential variety, their responses to the cartoons may have contributed to rather than alleviated the sense of intractability and bifurcation that followed.

But if the hadith alone does not explain why Muslims would see the cartoons as harmful, then what made them seem so? One contributing factor might have been the emergence of a populist discourse based on moral binaries, attachment to victimization and stigmatization of minorities – what we, in a Nietzschean vein, could call a culture of ressentiment. A carrier of this culture has been The Danish People's Party (DPP). Founded in 1995 by four dissenters from the far-right Progress Party, the DPP is currently Denmark's third largest party. Lending support to then Prime Minister Anders Fogh Rasmussen (recently replaced by Lars Løkke Rasmussen, another member of the same government), the party has been influential in mostly indirect ways, suggesting that failures in health care and social security stem from an unspoken alliance between socially privileged bureaucrats and Muslim immigrants. In the past 10 years, DPP has advocated stricter border control, limited membership of the EU and the protection of ‘Danish’ values. Feeding off an effective media strategy, this advocacy has empowered a stigmatizing discourse in which Muslims are perceived to be a threat to the future of Danish culture. Muslims living in Denmark note this perception and relate it to the publication of the 12 cartoons. According to one Imam living in Denmark, ‘It's not the cartoons, it is the way [the Prophet] is being presented … The pictures are saying that Muslims are terrorists, because he is a Muslim and he has a bomb in his head’ (quoted in Badkhen, 2006).

Although the cartoon war touched on more than sketched here, comments like these show how perceptions of harmful speech set the parameters for law and public culture within and across social divides. The parameters are never uniform in nature, something that was especially evident in the Danish case: Whereas a majority of Muslims saw the harm provoked by the Jyllands-Posten cartoons as a reason for regulating speech through law, the opposite could be said of constituents committed to values promoted by especially the DPP. Such contested, multiple perspectives challenge the liberal theory of free expression, which aims to be universally inclusive and above the fray. For these reasons, it may fail to appeal to constituents who do not share its framing of harm, law and free speech. Perhaps the question is not if liberalism frames harm unequally, but rather how and with what consequences for issues of contestation, intelligibility and recognition.

Harm in/of Liberalism

There is no way to assess this question without acknowledging liberalism as a rich and diverse tradition, which invokes a number of distinctions to ensure a robust commitment to free speech, focusing on the object (self-regarding versus other-regarding), the character (physical versus emotional) and the intensity (trivial versus non-trivial) of the harms in question.Footnote 4 Entrenched in the legal procedures and cultural mores of most Western democracies, these distinctions have become second-nature to many liberals. In recent years, however, liberals have shown an interest in developing a justification of free speech that doesn’t appeal to harm. Liberals point to an increased pluralism – what Rawls calls ‘the fact of pluralism’ – which appears to preclude a shared notion of what harm means. According to T.M. Scanlon, this means that we should modify Mill's harm principle, and instead use the principle of personal autonomy, which Scanlon defines as being ‘sovereign in deciding what to believe and in weighing competing reasons for action’ (Scanlon, 2003a, p. 15). Privileging this definition, Scanlon argues, allows liberalism to decide the limits of free speech by appealing to a general quality that all reasonable persons possess – that is, personal autonomy – avoiding the fluctuating and unstable nature of harm. Although some harms thus can remain impermissible because they evoke injury, panic, ridicule and defamation that weaken citizens’ ability to weigh competing evidence, or repress the ability to contest the judgment of others, a host of other harms remain permissible even though they arise from individuals having false beliefs, or from believing that some acts can be worth performing despite their harmful consequences. Indeed, ‘the harm of coming to have false beliefs’ is not one that an ‘autonomous man could allow the state to protect him against through restrictions on expression’ (Scanlon, 2003a, p. 17). This claim is more controversial than it first appears: Curled up inside it is a set of norms (autonomy, proceduralism and sovereignty) that citizens must internalize before they can determine the limits of harmful speech.

The issue here is one of what Butler (2009a, p. 6) calls ‘intelligibility’ and ‘recognizability’. Separating harm and law, why do liberals privilege some scheme of knowledge over others? And how might these schemes recognize some but not other modes of speech as being particularly problematic, thereby legitimizing certain practices of power and privilege? Such questions might lead us to assess the cultural differences between Western and non-Western conceptions of free speech. In addition, we should also focus on how the liberal frame disavows the interplay between, on the one hand, the legal norms defined by the principle of autonomy and, on the other hand, the processes of affect and perception that shape the citizens’ experience of autonomy and their judgments of harmful speech. Both responses enable a more nuanced and complex picture than the one currently envisioned by liberal theory (and its critics). In the Danish case, for example, we find that while most Muslims emphasized the harm caused by a defaming mockery of the Prophet, the majority of Danes focused on the harm caused by flag burnings, death threats and assaults on embassy buildings, all performed by or associated with the immigrant community. Moreover, to those who sided with Jyllands-Posten, the reaction against the cartoons seemed thin-skinned. Common retorts were thus ‘why can’t they [the Muslims] take a joke’ (Andersen, 2006), or ‘the cartoons are not as defamable as other cartoons in the history of political satire’ (Stjernfelt, 2006). Those on the other side did not see it that way. With the focus on the harm already done by the cartoons, decisions such as the one by the Danish government not to meet with ambassadors from the Middle East seemed only to deepen the wound. The Grand Mufti of Egypt expressed this view when he answered charges of violence by asking, ‘Why can’t you apologize now that you so evidently have harmed us?’ (in Thomsen and Hundsbæk, 2006).

Statements like these underscore a deep ambiguity in contemporary liberalism: On the one hand, it seems that liberalism can be co-opted by multiple constituents, each of which seeks to formulate their claim for recognition in ways that seem intelligible to the other side, refusing a neat separation of law and harm in an effort to call on liberalism to be creative and more responsive to context. On the other hand, however, liberals also insist that there has to be a way of limiting this creativity by grounding it in a universal principle that appeals (and therefore seems legitimate) to everyone involved. It is in this regard that liberalism's reliance on personal autonomy proves politically inadequate. Indeed, each side of the Danish cartoon war can adopt the principle of autonomy for its own purposes, engendering an undecidability that unsettles most liberals. Whereas Muslims can see the defamation of the Prophet as a threat to their autonomy, which suffers because of the discrimination implied by equating Islam with terrorism, Danes can see the assaults on embassies as a politics of fear that short-circuits an independent consideration of the 12 cartoons. To the extent that both of these interpretations are valid, liberalism remains less decisive than it seeks to be, limited instead to a frame in competition with others. On this account, we might say, liberalism is always part of the agon, and does not exist ‘above the fray’. It turns out that the creative instability endorsed by Scanlon as an aspect of liberal rights goes all the way down.

To be sure, in a case like the Danish cartoon war, the possibilities for contestation are never distributed evenly. Constituents for whom liberalism's conceptions of intelligibility and recognizability seem unproblematic are favored. Scanlon highlights this problem when he introduces a consequentialist consideration to adjudicate cases where unlimited free speech has unacceptable consequences (Scanlon, 2003b, p. 152). The move to consequences is a good one, but it requires that we attend to actual political contestation, since the attempt to judge harm independently of its significance for the parties involved tends to create a backlash in which government officials add insult to injury by privileging one set of impermissible harms over another set. The Danish cartoon war helps us once again to capture this problem, which we might call the ‘harm of justification’ that occurs when claims of harm are turned aside by reasons developed outside the context to which they are applied. This problem arose when the Danish government claimed it was neutral in its indifference to the cartoons. In so doing, the government favored the Danish majority's conception of harm, augmenting the harm caused by the defamation of the Prophet.

I highlight these points to show how the liberal frame's separation of harm and law implicates it in the intractability it seeks to attenuate. The frame does so because it abstracts from the perceptions and affects that animate free speech, and because it distributes the recognition of harm unequally in order to limit a practice of contestation that liberalism at its very best seeks to promote. It is Scanlon, rather than Rawls (1999) and Habermas (1998), who enables us to undo this impasse. When Scanlon explains that rights harbor within themselves a ‘creative instability’ (2003b, p. 154), he distinguishes among three elements: (1) the ends, (2) the means and (3) the linking empirical beliefs. Whereas the first two elements touch on the normative reasons we give for rights such as the one to free expression, the last – what Scanlon calls empirical beliefs ‘about the motivation of the relevant actors, about the opportunities to act that are available to them, and about the collective results of the decisions they are likely to make’ (2003b, p. 152) – stresses the motivations for executing this right, as well as the consequences that follow from this execution. Scanlon notes that empirical beliefs are subject to change because they rely on context-dependent encounters with the consequences that follow from implementing the right to free expression in this or that way. Key here is Scanlon's claim that empirical beliefs can also draw our attention to the fact that this right does not form a ‘coherent whole’ but instead entails ‘a dynamic quality’ and ‘creativity instability’ that ‘can lead to an almost constant process of revision’ (2003b, p. 154).

The turn to creative instability holds great promise. Striking an almost poetic chord, the term indicates that there are other resources in the liberal canon that we can use to frame harm differently, and thus avoid the bifurcation of harm and law that currently characterizes the debate. In fact, we may read Scanlon's felicitous term as suggesting that a more self-critical theory of free expression requires that liberals (and political theorists more generally) attend to the ‘dynamic quality’ that organizes the interplay between harm and law, taking into consideration not only legal norms but also visual experience, affective bonds, critical responsiveness and creative intervention. How does the incompleteness of perception motivate citizens to act? What role does the encounter with harm play in the link between perception, affect and action? How do these registers shape the citizenry's appreciation of harmful speech? To address these questions, I now turn to Merleau-Ponty, who may help supplement some of these underappreciated possibilities in contemporary liberal theory.

Harm from the Perspective of Perception

Unlike Scanlon who begins with the assumption of personal autonomy, Merleau-Ponty begins his inquiry by examining the heteronomy intrinsic to perception.Footnote 5 Imagine a person looking at a painting. The first thing this person aims at is to turn the painting into ‘an unbroken text’ with no gaps between ‘the previous instant and that of the following’, fulfilling what Merleau-Ponty calls the ‘tacit thesis’ of picture-perfect images (Merleau-Ponty, 1962, p. 54). In order to develop the painting's perspective, however, the person must also develop a sense of depth that is ‘hidden’ to and ‘simultaneous’ with the pursuit of picture-perfect images (Merleau-Ponty, 1968, p. 219). This paradox, Merleau-Ponty argues, creates an ongoing oscillation between depth and surface and, by extension, between richness and anxiety. For even as the depth of perception hints at how rich a person's experience of a painting can be, it also undermines the thesis of picture-perfect images defined by clear lines of demarcation between the perceiver and the perceived, engendering an anxiety that may encourage the perceived to reduce or even eliminate the very richness of perception.Footnote 6 What we have, then, are tree elements that jointly define perception: a thesis of picture-perfect images that expects the perceived world to appear as an unbroken text; a depth that enables the perceiver and the perceived to slip in and out of each other; and an enriching as well as anxiety-inducing difference, one subject to affirmation or disavowal.

What makes this analysis relevant to liberal theory is the fact that it, too, encourages us to see freedom of expression in the context of creative instability. As Scanlon did with the right to free expression, so Merleau-Ponty shows how perception is not a ‘coherent whole’, but instead possesses a ‘dynamic quality’ that pervades every image of self, other and world. To this quality, Merleau-Ponty adds a scheme of intelligibility and recognizability that paves the way for a more nuanced understanding of the interplay between harm and law. On the one hand, we can now speak of harm as arising from perception's inability to fulfill its own thesis of picture-perfect images. (This is what the anxiety mentioned above is about.) On the other hand, we can also speak of harm as emerging along a continuum of outcomes, each of which points to what Merleau-Ponty (2000, p. 109) would call harm's ‘sense or … future’. At one end of the continuum, harm is thus the sign (if not the instigator) of an instability that must be replaced by a legally codified world of perceptual stability and fixed identity. At the other end, harm represents the richness that lies in perception's incompleteness and that opens up to a world of contestation and difference. Fraught with its own experience of harm, this world may nonetheless be affirmable because it connects with the value of visual depth and sensorial abundance.

Merleau-Ponty would be the first to admit that whatever follows from this continuum is not normative in the sense of being valid at all times and in all places. To Merleau-Ponty, however, this admission is not a weakness, but underscores the heteronomy that haunts the organization of both law and harm. To see why this admission can be an important element in the framing of free speech, consider the cartoon depicting the Prophet Muhammad with a bomb in his turban (the ‘bomb cartoon’). Cartoons like this one seem especially prone to force a bifurcation of law and harm. Merleau-Ponty's analysis changes this by inviting us to see the cartoon as a question of perception in which the cartoon is judged neither by harm or law, each defined separately, but rather by its call to see things in a certain limited way, not open to perception's own instability and richness. Using black and white colors, quoting the Islamic creed on an ignited bomb serving as the Prophet's turban, the bomb cartoon thus combines visuals in a way that collapses Islam, Muslim identity and terrorism, and as such blocks an exploration of these terms in conjunction with their various effects on Danish culture. Indeed, the cartoon implies that we don’t need such exploration, since ‘we’ (that is, ethnic Danes) already know that Muslims see no value in reciprocating any interest ‘we’ might show in their faith. This assumption makes the bomb cartoon emblematic of how some types of harm can foreclose an expansion of free expression, replacing mutual contestation (which, of course, has its own harms) with the opposite – moral binaries, cultural stereotypes, and clear lines of causality for pain and suffering.

In reply to this reading, liberals may counter that even if we ought to criticize the bomb cartoon, we risk a slippery slope if we censure the interests that citizens have in calling something to the attention of a wider audience (cf. Scanlon, 2003c, p. 102). Without a protection of these interests, enforced by free speech as a constitutional right, Merleau-Ponty's alternative may in fact be used to justify a content-regulation that undercuts public scrutiny and mutual contestation. Merleau-Ponty, I think, would disagree. His willingness to speak against preconceived dogmas, as well as his interest in a speech situation oriented more toward contestation than regulation, suggest that for him, too, we should not try to restrict content in any predetermined way. But in response to Scanlon and other liberals, Merleau-Ponty might add that distinguishing between different kinds of interests (as Scanlon does), and then using these interests to categorize different domains of speech, is itself a framing of free speech, one that not only makes the separation of interests possible, but also names certain domains as ‘interests’. From the perspective of Merleau-Ponty, such moves demand critical scrutiny. For him, the issue is less how to limit harmful speech legally, and more how our hidden assumptions frame the interplay between harm and law. Merleau-Ponty therefore does not see a contradiction between the two demands posed by the twelve Jyllands-Posten cartoons – that is, greater sensitivity to harm and a reluctance to regulate the right to free speech through law.

Still, we need to press further ahead if we want to grasp the conditions that enable a framing of free speech based on mutual contestation, social equality and respect for difference. One way to do so is to follow Merleau-Ponty's correlation of perception with the more general category of expression, which he defines as a synthesizing mode of empowerment in which the expressed does not exist before expression, but instead fuses seemingly disparate elements in ‘a’ world. The fusion occurs at the level of the styles that format the background of speech, at the level of the tones and pauses that direct the rhythm of speech, and at the level of the perspectives that outline the standpoint of speech (Merleau-Ponty, 1964, pp. 183, 187). Merleau-Ponty stresses that these levels do not form a coherent whole, but always add something new to expression. Like perception, expression is thus part of a revisionary process similar to the one Scanlon identifies: ‘To express oneself is’, Merleau-Ponty says, ‘a paradoxical enterprise, since it presupposes … a fund of kindred expressions, […] and … from this fund the form used should detach itself and remain new enough to arouse attention’ (Merleau-Ponty, 1973, p. 35).

The idea that expression hinges on the ability to ‘arouse attention’ is helpful as it allows us to further develop the interplay between harm and law in situations where there is no pregiven definition of harmful speech. In these situations, harm stands out – or, as Merleau-Ponty puts it, ‘arouses attention’ – because it offends those subject to it. But as cases like the Danish cartoon war show, the way citizens react to this offense depends primarily on the political context that structures their initial reaction. We should therefore also speak of a second way in which the arousal of attention matters – as the gaze that folds the initial experience into the web of meaning and power that sustains free speech (and the legal norms it implies). The upshot is a circular movement in which both the attention created by and the attention paid to harm frame free speech, placing it in the context of different worlds and different futures. Merleau-Ponty foregrounds this plurality by emphasizing the creative instability of not only rights but also visual experience, orienting us toward the perceptual richness postulated by creative instability and highlighted by the contestability that arises from expressions being subject to revision.

Interestingly, a good example of this reframing can be found in Jyllands-Posten's original series of images, which also featured cartoons attentive to the politics of framing. One, the ‘mirror cartoon’, stands out because it refuses the underlying premise of Jyllands-Posten's project: The cartoon draws a scene from a police station in which a middle-aged man looks at a police line-up through a one-way mirror. Although all in the line-up wear a turban, none resembles the Prophet. In fact, they all seem to caricature someone else (number ‘2’, for example, is the chairwoman of the DPP). Even so, the man looking through the mirror states that he ‘does not recognize him’. This way of framing the ‘frame’ and ‘the framer’ is highly suggestive. Calling out the ignorance with which most Danes discuss Islam, as well as highlighting the power of the frame that insulates this ignorance from contestation, the mirror cartoon is both a comment on how a majority culture allocates the recognition of harm unequally and an attempt to bring out the creative instability embedded in this situation by mirroring it: Is the viewer the middle aged man in the cartoon? Or not? Free from claims to superiority or impartiality, the mirror cartoon solicits the viewer into its frame, and may thus be a model for a self-critical framing of free speech based on mutual contestation, social equality and respect for difference.Footnote 7

Affective Becomings

Theorists interested in the Danish cartoon war rarely notice the potential of the mirror cartoon to inform or expand standard views of free speech. They either accept the liberal frame as it is, or emphasize the harm that many Muslims invoke, one that seems to follow from ‘distinctively different conceptions of the subject, religiosity, harm, and semiosis’ (Mahmood, 2009, p. 88). To further unsettle the idea that we must choose between these approaches, I turn now to Gilles Deleuze who, as we shall see, makes a valuable contribution to the discussion of free speech: His work on affect in particular enables us to explain why the bomb cartoon – and not the mirror cartoon – became the center of attention, and it allows us to develop the ways in which different harms empower different legal norms, deepening the creative instability that keeps troubling contemporary liberalism.

Even though many interpret Merleau-Ponty and Deleuze as belonging to two opposing strains in French philosophy, they share a set of basic assumptions regarding the study of experience.Footnote 8 Similar to Merleau-Ponty's view of perception, Deleuze sees affect as a mode of world-making that keeps subverting its own being. Deleuze substantiates this impression by showing how affects always appear as heterologous combinations of joy and sadness. Hope, for example, combines the joy of looking towards a desired goal with the sadness of not knowing whether the goal will become real. This uncertainty is in turn related to fear, which mixes the sadness of a perilous situation with the joy of hoping that the danger is short-lived. In both cases, no one controls what follows from these affects. In fact, neither individuals nor collectives ‘have’ affects, if by the verb ‘to have’ we mean the possession of entities available for exchange or trade. Instead, affects suspend one configuration of the body and allow another to emerge. That is, they defy confinement to one setting; they entail their own registers of power; and on these registers, they modify individual as well as collective bodies. Deleuze and his collaborator, Félix Guattari, summarize these features as modes of becoming: ‘To the relations composing, decomposing, or modifying an individual there correspond intensities that affect it, augmenting or diminishing its power to act … Affects are becomings’ (Deleuze and Guattari, 1987, p. 256).

Deleuze and Guattari challenge the possibility of personal autonomy, but contemporary liberals may nonetheless profit from their view of affect, which gives a more complex account of what it means to be subject to harm. Indeed, according to Deleuze, we should never assume that the encounter with harm always will produce the same reaction – one of desecration and woundedness, justifying legal protection, if not political retribution. As an experience of not only perception but also affect, harm is a plurality circumscribed by different degrees of power (understood in terms of what both Spinoza and Nietzsche discuss as potentia). We might even speak of two types of harm that build on the insights Merleau-Ponty develops from the perspective of perception: Whereas ‘sad harm’ would be the feeling of trying but ultimately failing to achieve picture-perfect images, ‘joyful harm’ would be the feeling of disorientation that arises from the richness of perceptual experience and that opens up alternative resources heretofore unseen. Deleuze encourages us to make this move, but warns at the same time against assuming that one always perpetuates empowerment, with the other being nothing but a feeling of powerlessness (cf. Deleuze, 1983, p. 62). This caution is well taken with regard to the Danish cartoon war where, as we have seen, joyful affects some times arose from the pleasure of somebody else's pain, and sadness some times lead to a shared sense of community, especially when it linked up with concerns for caring and empathy. To recognize both possibilities, we must develop a more complex continuum, which moves between an active capacity that expands connections across self, other and world, and a reactive capacity that contracts connections across self, other and world. Each end of the continuum expresses a different ‘sense or future’ of harm: As harm either expands or contracts connections among constituents, it engenders political interventions ranging from mutual contestation and juridical flexibility to social antagonisms and legal discrimination.

The idea that harm can generate a variety of responses, futures or politics points to the dangers of a bifurcation of harm and law, and invites us to consider a third option, one I shall say more about in the Conclusion. But first it might be helpful to clarify why the bomb cartoon – and not the mirror cartoon – became the center of attention in the Danish cartoon war. Although skeptics have wondered why Muslims couldn’t accept the bomb cartoon as a joke, Deleuze's account of affect as becoming goes beyond naïve wonder: Deleuze's account calls on us to note how the bomb cartoon was privileged by an affective context in which Muslims living in Denmark struggled with stigmatization, and where some groups responded by exploiting uncertainty about how to interpret Islam's prohibition of images. The exploitation funneled a perception of all Muslims as premodern fundamentalists, and fed into the reactive powers associated with sad harm, nourishing an affective connection between the bomb cartoon, the world of hadiths and the so-called war on terror. In this vicious circle, the mirror cartoon was left as a minoritarian outlier incapable of challenging majority assumptions regarding the nature of secular politics and democratic pluralism. The affective context, in other words, directed the public eye to the bomb cartoon rather than the mirror cartoon.

And yet, since the encounter with harm is never unidirectional, leading to one reaction, one feeling, we should be careful not to expect constituents in general to connect harmful speech with what Wendy Brown calls ‘wounded attachments’ – that is, attachments that collapse the distance between the deed of harm and the one subject to this deed, victimizing citizens who become ‘invested in [their] own subjection’ (Brown, 1995, p. 70). Although the Danish cartoon war unfolded in a way that seems to support such an analysis, taking harm seriously need not mean we yield to woundedness. Such an expectation may in fact add to the current impasse because it treats harm univocally and thereby occludes the creative instability embedded in the interplay between harm and law. If Deleuze and Merleau-Ponty are right in their analyses of the processes that sustain free speech, then there is no harm outside its framing. One lesson we learn from the Danish cartoon war is thus that framings not only privilege one kind of harm over others, but also privilege some but not other expressions as harm. This lesson encourages us to politicize all attachments – whether harmful or not, wounded or not – foregrounding the question of worldliness: that is, the question of whether specific attachments either empower or isolate, expand or contract connections across self, other and world.

Conclusion

The question of worldliness brings us to the importance of what Butler (2009a, p. 10) characterizes as the ‘perpetual breakage’ of every frame. According to Butler, this breakage occurs because a frame's ‘conditions of reproducibility’ must adjust to contingent contexts, which constantly change, creating a process in which ‘[w]hat is taken for granted in one instance becomes thematized critically or even incredulously in another’ (ibid.). This is a more politicized version of Scanlon's point about the contingency of the right to free expression. As we saw earlier, liberal theory frames harm in a way that privileges some modes of speech over others in order to diminish pluralism, and yet given it is fraught with ‘creative instability’, on Scanlon's account, liberalism may also be open to more pluralization. This openness, I have argued, should encourage non-liberals to adopt liberalism's affirmation of free speech – even when it entails satire, mockery, ridicule and scorn – while at the same time seeking to destabilize the principle of personal autonomy that currently sustains liberal theory.

To elaborate on this insight, I would like to conclude by way of three principles that are neither politically impartial nor philosophically incontestable, but instead represent what Deleuze and Guattari call an ‘intensity map’ that demarcates the ‘flows of intensities […] their continuums and conjunctions of affects, [and] microperceptions’ (Deleuze and Guattari, 1987, pp. 164, 162). Understood in this way, the principles both recognize how different sensibilities inflect the same legal norm in different ways, and designate a new framing of harm based on mutual contestation, social equality and respect for difference:

  1. 1

    The principle of creative instability according to which registers of affection, perception and discourse frame harm through pluralizing processes of becoming.

  2. 2

    The principle of affirmative power according to which states of joy rather than sadness are turned to nourish the affective as well as normative affirmation of creative instability.

  3. 3

    The principle of pluralizing worlds according to which the right to free expression privileges mutual contestation, social equality and concern for difference out of respect for the modern condition of deep pluralism.

To speculate on what the outcome would be if we allowed these principles to inflect our commitment to free expression would be just that – speculation. And yet, if we acknowledge that part of political theory's mission is to speculate about possible futures, then perhaps we could say that a society committed to creative instability, joyful affects and political pluralization would be one in which free speech is more open to its own contestation than contemporary liberalism allows. This imaginary society would not allow us to miss the fact that the 12 Jyllands-Posten cartoons – variously differentiated in their style, context and purpose – commit different kinds of harm. Moreover, the society would be one that directs the public eye more toward cartoons such as the mirror cartoon and less toward cartoons such as the bomb cartoon. Indeed, the bomb cartoon would be too flat and uninteresting to hold the gaze of a society of persons pitched on creative instability, joyful affirmation and political pluralization. For these persons, the main challenge would be one of creating the conditions necessary for turning a cartoon war into an agonistic exchange, one in which encounters with harm are not so much obstacles to freedom as signs of difference and contestation. Might this not be the best we could hope for in a society characterized by deep pluralism and social complexity? Or alternatively put: This is indeed the best we can hope and work for in a society characterized by deep pluralism and social complexity.