1 Introduction

Since the first occurrence of the term in Sürek v. Turkey (1999), the European Court of Human Rights (ECtHR) has adopted a decidedly restrictive approach towards hate speech, thus becoming an international “centre of gravity” with regard to the implementation of hate speech bans [22: 181, 36]. While a fully-fledged definition is still missing, the Court provided its most extensive reflection on hate speech in Lilliendahl v. Iceland (2020):

The first category of the Court’s case-law on ‘hate speech’ is comprised of the gravest forms of ‘hate speech’, which the Court has considered to fall under Article 17 [: prohibition of abuse of rights] and thus excluded entirely from the protection of Article 10 [: right to freedom of expression] … The second category is comprised of ‘less grave’ forms of ‘hate speech’ which the Court has not considered to fall entirely outside the protection of Article 10, but which it has considered permissible for the Contracting States to restrict … Into this second category, the Court has not only put speech which explicitly calls for violence or other criminal acts, but has held that attacks on persons committed by insulting, holding up to ridicule or slandering specific groups of the population can be sufficient for allowing the authorities to favour combating prejudicial speech within the context of permitted restrictions on freedom of expression. [32 at para 34–36]

Both categories of hate speech outlined here typically refer to incitement to hatred or intolerance with respect to protected characteristics such as ethnicity, religion or sexual orientation, although the Court’s approach is not always consistent or predictable in this respect [27, 36]. Definitional difficulties aside, it is clear that certain forms of aggressive and disparaging humor (which the Lilliendahl judgment calls “holding up to ridicule”) can play an important role in what the Court recognizes as hate speech. Most recently, in ZB v. France (2021), the ECtHR explicitly acknowledged that hate speech is often “disguised as a humorous appearance which, precisely for that reason, may prove to be as dangerous as direct speech” [55 at para 20]—a finding that is largely confirmed by scholarly work on humor’s “cloaking” or normalizing effect on hate speech [20, 41, 48].

At the same time, this emphasis on the dangers of hateful jokes is potentially at odds with the vital importance of humor (i.e. expression that is meant as facetious or non-serious, and/or interpreted as such by at least some of its addressees) in democratic life, and with the special protection granted to it in Strasbourg jurisprudence. Focusing in particular on satirical humor, the landmark case Vereinigung Bildender Künstler v. Austria (2007) famously states that “satire is a form of artistic expression and social commentary and, by its inherent features of exaggeration and distortion of reality, naturally aims to provoke and agitate. Accordingly, any interference with an artist’s right to such expression must be examined with particular care” [51 at para 33]. The elusive and subjective nature of humor makes it particularly difficult to strike a balance between these two contrasting stances, and thereby draw a clear line between hate speech and protected expression. In the case of humorous material, it is therefore all the more important for courts to conduct a thorough, nuanced analysis of the impugned expression, based on a shared approach to the specific interpretive challenges posed by humor. Developing a systematic analytical framework is a necessary first step towards achieving greater consistency on a judicial level, irrespective of broader juridical debates on the effectiveness and desirable reach of hate speech restrictions. As shown by recent scholarship, an interdisciplinary perspective on humor-related jurisprudence can prove extremely helpful in this respect, by combining case-law analysis with insights from linguistics, semiotics and literary theory.Footnote 1

While a small number of studies have already addressed the treatment of humor and satire in ECtHR cases revolving around free speech and its limits [3, 17, 19], no specific attention has been paid so far to the complex relations between humor and hate speech. The present contribution will take a first step towards filling this gap by examining a corpus of 10 cases from a cross-disciplinary standpoint, aiming to show how humanities research can provide courts with useful conceptual tools for the interpretation of (potentially) disparaging humor. The following section will introduce the selected cases, as well as outlining the general structure of this article.

2 Corpus and General Outline

Through a series of combined keyword searches in HUDOC, we identified a corpus of ten ECtHR cases regarding hate speech, where at least one of the parties involved evoked the notions of humor, irony or satire to describe the impugned expression. The facts and outcomes of the selected cases are summarized below, in chronological order:

  1. 1.

    Seurot v. France (Application no. 57383/00, 18 May 2004). The applicant was a teacher in a private college. He wrote an article titled “Enough is enough,” which contained racist references to Muslim “hordes” and their “dirty, arrogant girls.” The article was published in the school newsletter and disseminated to pupils and parents, but Seurot was later dismissed for his racist and hateful writings. The applicant submitted that the article was “humorous” and was originally meant as a private letter, although he did not “take the precautions needed to prevent its publication.” Outcome of domestic proceedings: Both the applicant and the school director were charged with incitement to racial hatred. The applicant was sentenced to a fine of 5,000 francs. ECtHR finding: Seurot’s application was deemed inadmissible, since the interference was “necessary in a democratic society” (unanimous).Footnote 2

  2. 2.

    Leroy v. France (Application No 36109/03, 2 October 2008). The applicant is a cartoonist who illustrates for, inter alia, a Basque weekly magazine. On the day of the attack on the Twin Towers, he submitted a drawing of the attack with the caption “We had all dreamed of it… Hamas did it,” parodying a Sony advertising slogan. He stated his intention was to represent the destruction of the American empire. The cartoon was published on September 13th, 2001. Domestic proceedings: The applicant was charged with complicity in the glorification of terrorism, and the national court considered that damage had been done to public order in a region sensitive to terrorism. Both the applicant and the director of the magazine were sentenced to a fine of €1,500, with injunction to publish the judgment in the magazine. ECtHR finding: No violation of Article 10 (unanimous).

  3. 3.

    Féret v. Belgium (Application No 15615/07, 16 July 2009). The applicant, a member of the Belgian House of Representatives, was president of the political party National Front as well as responsible for the Party’s publications and website. The party distributed xenophobic content and jokes about non-European refugees and immigrants in Belgium, especially Muslims, including a leaflet blaming the 9/11 attacks on the “couscous clan.” The applicant was charged with “having incited discrimination, segregation, hatred or violence against a group, community or their members on the basis of the alleged race, colour, descent or national or ethnic origin.” Domestic proceedings: The applicant was charged with inciting discrimination and violence. He was sentenced to 250 hours of community service related to the integration of immigrants and a ten-month suspended prison sentence. He was also declared ineligible for ten years. ECtHR finding: No violation of Article 10 (4 votes to 3).

  4. 4.

    M’Bala M’Bala v. France (Application no. 25239/13, 20 October 2015). The applicant—a comedian and political activist—put on a public performance in which he invited an academic, well known for his negationist views regarding the Shoah, to join him on stage. The applicant then staged a mise en scène involving a man dressed in stripped pyjamas (which resembled clothing worn by Nazi concentration-camp inmates) and wearing a yellow star with the word “Jew” written across it. This man gave a “prize” of a candlestick to the academic for his “unfrequentability and insolence.” Introducing the sketch, M’Bala M’Bala stated he wanted to “do better” than a previous performance which had been described as “the biggest antisemitic rally since the Second World War.” Domestic proceedings: The applicant was charged with public insults directed at a person or group of persons on account of their origin or of their belonging, or not belonging, to a given ethnic community, nation, race or religion. He was sentenced to a fine of €10,000. ECtHR finding: Inadmissible under Article 17, i.e. prohibition of abuse of rights (unanimous).

  5. 5.

    Sousa Goucha v. Portugal (Application no. 70434/12, 22 March 2016). The applicant was an openly gay TV host, who filed a complaint for defamation and insult after he was described in a television comedy show as “The best Portuguese female TV host.” He alleged his reputation and dignity had been damaged, the state had not protected his reputation and that his complaint had been dismissed in the national courts because of discrimination based on his homosexuality. For the purposes of our analysis, the main issue was whether the State had achieved a fair balance between the applicant’s right to protection of his reputation (Article 8) and the media’s right to freedom of expression (Article 10). Domestic proceedings: The applicant’s complaint was dismissed. ECtHR finding: No violation of Article 8 (unanimous).

  6. 6.

    Le Pen v. France (Application No 45416/16, 28 February 2017). The applicant is Jean-Marie Le Pen, the founder and honorary president of the National Front. He delivered a speech in which he used a pun equating Romani people with birds as they both “vol[ent] naturellement” [i.e. fly, but also steal, naturally]. Domestic proceedings: The applicant was charged with public insult to a group of persons on account of their belonging to an ethnic group. He was sentenced to a criminal fine of €5,000. He was also ordered to pay the plaintiff (a civil rights organization) the sum of €3,000 in damages and €3,000 pursuant to Article 475–1 of the Code of Criminal Procedure, as well as €1 in damages to each of the other complaining associations and €500 pursuant to the aforementioned article 475–1. Finally, the court ordered the publication of a press release setting out its decision, in a press organ chosen by the civil parties. ECtHR finding: Le Pen’s application was ruled inadmissible as the interference was necessary in a democratic society (unanimous).

  7. 7.

    Kaboğlu and Oran v. Turkey (Applications no. 1759/08, 50,766/10 and 50,782/10, 30 October 2018). The applicants, two university professors, had been appointed to a Human Rights body to provide the Government with “opinions, recommendations, proposals and reports relating to the promotion and protection of human rights.” Their report suggested that Turkey should revise the concept of citizenship in a multicultural, liberal and pluralist sense. The report created a backlash in ultranationalist newspapers, with the applicants receiving insults and threats in the press—some of which it was argued were phrased in a “satirical” style. The applicants complained about infringements to their rights to respect for a private life under Article 8, claiming that the State failed to protect their private lives against the attacks launched against them. Domestic proceedings: The applicants’ complaints were dismissed. ECtHR finding: Article 8 violation (unanimous). The Court decided that this was the main legal issue, and therefore did not examine the complaints under Article 10.

  8. 8.

    Z.B. v. France (Application No 46883/15, 2 September 2021). The case focuses on a joke referencing the 9/11 attacks printed on a T-shirt, which the applicant gave as a gift to his 3-year-old nephew in September 2012. The T-shirt bore the words “Jihad, né le 11 septembre” [Jihad, born on 9/11] and “je suis une bombe!” [I am a bomb!]. The child is actually called Jihad (which is a rather common name in the Arab world), and was born on September 11th, 2009. The term bomb can also mean ‘good looking’ in French. The T-shirt was worn only once at preschool, and was only seen by adults when the preschool’s director and one of the employees changed Jihad’s clothing in the bathroom. Domestic proceedings: The applicant and his sister (Jihad’s mother) were charged with the criminal offense of advocating crimes of willful attacks on life. The applicant received a two-month suspended prison sentence and a €4,000 fine, while Jihad’s mother received a one-month suspended sentence and a €2000 fine. ECtHR finding: No violation of Article 10 (unanimous).

  9. 9.

    Bonnet v. France (Application No 35364/19, 25 January 2022). The applicant, Alain Bonnet, is the president of “Égalité et Réconciliation,” a “political association” which owns a website and had published an article with an illustration which parodied a Charlie Hebdo front page. The Charlie Hebdo original depicted Belgian singer Stromae and alluded to his song “Papaoutai” [Dad, where are you?] as a darkly humorous commentary on the 2016 terrorist attacks in Brussels. The website illustration was entitled Chutzpah Hebdo (chutzpah being an Yiddish insult) and was accompanied by text reading “Attacks[:] the Zionists are in the square,” “Report[:] how the Mossad makes Molenbeek.” The illustration also included a reference to “disoriented historians” and a drawing of Charlie Chaplin (instead of Stromae) in front of a Star of David, under the title “Shoah où t’es?” [Shoah where are you?], surrounded by speech bubbles from drawings of soap, a lampshade, a shoe and a wig replying “here,” “there” and “and there too.” Domestic proceedings: The applicant was charged with public insult of a racial nature and contesting crimes against humanity. He was sentenced to pay a fine of €10,000. ECtHR finding: Bonnet’s application was deemed inadmissible, as the interference was necessary in a democratic society (unanimous).

  10. 10.

    Ogurtsov v. Russia (Application No 61449/19, communicated on 9 March 2021). The applicant ran the CSKA football club’s fan account on social media. When the team lost to their main rivals Spartak he wrote a poem which he shared on social media and published in a poetry collection (300 copies printed), including lines such as “Spartak is shit, as are its fans, // I will wipe my ass with your crest!”. Domestic proceedings: The applicant was charged with inciting hatred towards the Spartak football club and fans “on account of their membership of a social group.” He was sentenced to a fine of 10,000 Russian roubles. ECtHR finding: n.a. (the case was only communicated). Nevertheless, we decided to include the case in the corpus, as it lends itself to some relevant considerations regarding the contextual factors involved in the interpretation of humor or satire (see below, Sect. 5).

Most of the cases listed above comply with the usual definition of hate speech as abusive expression targeting protected characteristics such as ethnicity, religion or sexual orientation. On the other hand, both Kaboğlu and Oran and Ogurtsov extend the notion of hate speech to what can be broadly defined as “social groups” (respectively progressive academics and supporters of a football club). Moreover, the two judgments regarding glorification of terrorism—Leroy and Z.B.—were included in the corpus as they both explicitly refer to “discours de haine” [hate speech], although the target cannot be easily identified in light of protected characteristics. Eight of our cases (Seurot, Leroy, Féret, M’Bala M’Bala, Le Pen, Z.B., Bonnet and Ogurtsov) were brought by the speakers/authors of the contentious expression, while two (Sousa Goucha and Kaboğlu and Oran) were brought by the targets/victims. Excluding Ogurtsov (which was only communicated), all seven cases brought by the speakers resulted in the applicant’s loss, through finding of a non-violation of Article 10 (3 cases) or inadmissibility under Article 17 (1 case) or due to the application being manifestly ill-founded (3 cases). Of the two cases brought by the target of the contentious expression, one overturned the national outcome (thus restricting disparaging speech), while the other upheld the previous finding of non-violation of Article 8. In short, 90% of the rulings resulted in the restriction of purportedly hateful speech, either by upholding the decision made by the national courts (8 cases) or by overturning a previous ruling protecting the expression at issue (1 case)—this percentage is strikingly higher than the one calculated by Jacob Mchangama and Natalie Alkiviadou in their analysis of ECtHR hate speech cases [36], which might suggest an even more pronounced tendency to restrict hate speech when conveyed in the form of humor. To conclude our preliminary remarks, it is worth noting that six out of ten cases are from France. This might be partly due to the particular attention paid to hate speech in the French Law on the Freedom of the Press, which makes it particularly easy for civil-right organizations to seek damages in case of incitement to hatred or discrimination with regard to protected characteristics.Footnote 3

The following sections will provide a cross-analysis of the cases, with particular attention to how linguistic, semiotic and literary research can assist courts in the interpretation of material presented as humorous within hate speech cases. Sections 3, 4 and 5 focus on the intrinsic and extrinsic factors that may be relevant to the Court’s interpretation—namely the humorous mechanisms at work in the contested text/utterance (3), the latter’s dialogue with previous verbal or non-verbal texts by means such as allusion, parody or commentary (4), and the role played by context in its various senses (5). Section 6 discusses the potential outcomes of the interpretive process, by proposing a typology of possible relations between humor and disparagement (where disparaging humor is only one of the potential interpretations). Section 7 reflects on the criteria that can be used by courts when the same given joke or cartoon can be understood differently by different audiences, which might complicate the task of assessing the speaker/author’s culpability. Lastly, Sect. 8 offers some concluding remarks, while highlighting future research avenues regarding the fine line between offense and harm in disparaging humor.

3 Textual Mechanisms

Over the past 2 decades, humor research has produced several theoretical models mapping the ways in which humorous texts work [4]—‘text’ being used here in its broad semiotic sense, covering both verbal and non-verbal media. As shown in previous studies [17, 19], Paul Simpson’s model of satirical discourse [45] can prove especially useful in the judicial context, due to its particular attention to the role of irony as well as to the multiform relations between humor and what is perceived as factual reality. While originally focusing on satire in particular, most of Simpson’s theoretical framework can be successfully transferred to the analysis of humorous communication in general. Most interestingly for our analysis, Simpson distinguishes between two fundamental categories of satirical (humorous) strategies, namely the metonymic and the metaphoric type. Metonymic strategies are processes of humorous distortion or reversal within the same given conceptual domain. Those include the following sub-types: (1) Saturation, ranging from verbal exaggeration (e.g. proclaiming that a corrupt politician is “the most corrupt ever”) to visual caricature (portraying the politician holding a huge bag full of money); (2) Attenuation—i.e. presenting something familiar from a naïve perspective, thereby highlighting the strangeness in what is usually taken for granted (describing the politician’s behavior as if an alien were witnessing corruption for the first time, thus implicitly denouncing the unacceptability of said behavior); (3) Negation, which reverses a given situation into the opposite polarity within the same domain (for example, representing our hypothetical politician as the least corrupt politician ever, or praising them for going through a whole week without taking any bribes). On the other hand, metaphoric strategies involve merging or juxtaposing ideas coming from different conceptual domains, usually by means of analogy (e.g. representing the corrupt politician as a vampire) [45: 111–151; the examples are ours].

Moving back to our corpus, M’Bala M’Bala v. France effectively illustrates the full range of what Simpson defines as metonymic strategies. On the one hand, the sketch as a whole is based on the mechanism of negation, as the public stigmatization of negationist historian Robert Faurisson is reversed into a “prize” awarded to him by a man dressed like a concentration-camp inmate. On the other, Faurisson’s speech during the sketch includes some elements of attenuation, as he seems to take a naïve, estranged perspective on the “special treatment” he receives as a Holocaust “revisionist,” claiming not to understand the reasons behind it and thereby denouncing its alleged strangeness: “I have been the object of special treatment ten times. Including one time when I came close to death and the one who saved me without knowing my name, when he found out my name the next day, told the police that he regretted saving my life” [34 at para 8]. Lastly, both the comedian and the historian also resort to exaggeration in their statements. For example, Faurisson calls his opposers affirmationists, and invites the audience to “spell the word as [they] please” [34 at para 8]—the implicit allusion to the alternate spelling affirma-Sionistes hyperbolically implies that all those who oppose Holocaust denial are automatically Zionists.

The other case in the corpus involving antisemitism (Bonnet v. France) is, instead, an example of predominantly metaphoric disparaging humor. The main metaphoric link is the one between the “disoriented historians” looking for traces of the Holocaust (“Shoah where are you?”) on the one hand, and the singer Stromae—as represented in the Charlie Hebdo cover—looking for his missing father in the song Papaoutai (“Dad where are you?”). The primary goal of this metaphor is, of course, to cast doubt on the historical truth of the Holocaust; parallel to that, the cartoon also builds on the original Charlie Hebdo reference to the 2016 terrorist attacks in Brussels, which the applicant blames on Israel (“Report[:] how the Mossad makes Molenbeek”). Notably, the impugned cartoon replaces the face of Stromae with that of Charlie Chaplin—which, one could assume, is meant to represent the disoriented historians, as well as being a symbol of Jewishness (due to the actor’s purported Jewish origins).Footnote 4 On a rhetorical/semiotic level, this can partly be seen as a metonymy or synecdoche, Chaplin being presented as a pars pro toto within the conceptual domain of Jewishness. However, a metaphoric component is also present, based on the apparent analogy established between the Shoah and the sphere of cinematic fiction, evoked here through the Charlot reference. This ‘cinematic’ trope is particularly frequent in antisemitic discourse, as confirmed for example by a 2018 Italian case regarding a far-right politician who wore a negationist T-shirt bearing an “Auschwitzland” logo (with an obvious visual hint at Disneyland) [9]. In that case too, the Disney allusion was used to insinuate that the Holocaust was rather a matter of (Disney-like) fiction than historical reality; this was taken into account by the first instance court, which sentenced the politician to a €9050 fine (the appeal is still pending) [8]. As far as Bonnet is concerned, the metaphoric implications underlying the reference to Charlie Chaplin are not mentioned anywhere in the ECtHR decision or in the national proceedings, although they do seem to constitute an important aspect of the cartoon’s overall message.

While M’Bala M’Bala and Bonnet offer particularly clear examples of metonymic and metaphoric techniques respectively, the other cases in the corpus can also be analyzed in light of Simpson’s model. The table below provides a complete overview of the textual mechanisms at work in the impugned humorous expressions within each case (Table 1).

Table 1 Metonymic and metaphoric strategies in the selected cases

As suggested by this overview, disparaging or controversial humor can employ a wide variety of rhetorical tools, spanning over the whole spectrum of Simpson’s metonymic and metaphoric strategies. Looked at from this perspective, the ECtHR’s tendency to define humor and satire only in terms of “exaggeration and distortion of reality” [51 at para 33] might be limiting, as this phrasing does not really capture the nature of figurative devices such as negation or metaphor. As shown by recent work, a more widespread awareness of the diversity of humor’s textual mechanisms could lead the Court to more nuanced and accurate analyses, which may sometimes have a significant impact on the outcome of a given case (see [19], with particular regard to Palomo Sánchez and Others v. Spain). As far as our corpus is concerned, however, the Court paid adequate attention to the rhetorical and semiotic strategies at work in the texts under discussion—which often resorted to humor as a way to muddy the waters between hate speech and “just a joke” (see for instance the role of puns in Féret v. Belgium and Le Pen v. France).

That being said, the ECtHR’s reasoning is not entirely convincing in at least one case, namely Z.B. v. France—where, despite lamenting the lack of a “more developed motivation” [55 at para 66], the final decision essentially confirmed the cursory interpretation put forward on a national level by the Nîmes Court of Appeal:

Certain attributes of the child (his first name, day and month of birth) and the use of the term “bomb,” which cannot reasonably be claimed to refer to the beauty of the boy, are magnified through the turn of phrase, […] and in reality serve as a pretext to valorize, unequivocally, and through the deliberate association of terms referring to mass violence, willful attacks on life. [55 at para 11; our emphasis]

To begin with, the idea that the T-shirt “unequivocally” glorifies the 9/11 attacks is far from indisputable, as proved by the fact that the first instance court of Avignon had reached the opposite conclusion [55 at para 8]. As will be further discussed in Sect. 6, the applicant’s joke—as distasteful as it might be—could actually be an attempt to ironize on the widespread Islamophobic prejudice whereby a child named Jihad, who was furthermore born on 9/11, can only be a potential terrorist. However, this aspect will be explored in more detail later. At this stage of our analysis, it is more relevant to reflect on the Court of Appeal’s statement that the French term bombe “cannot reasonably be claimed to refer to the beauty of the boy,” as this directly relates to the textual mechanisms underlying the joke. From a rhetorical perspective, the Court of Appeal’s reasoning is clearly flawed—“I am a bomb” is a rather conventional metaphor, where the source domain (bomb) is mapped onto the target domain (child) based on a perceived analogy which is well established in the French language.Footnote 5 Therefore, technically, the word bombe does refer to the beauty of the child; and more generally, the T-shirt’s allusions to the 9/11 attacks are part of a rhetorical construction where terrorism is only evoked as a source domain, ultimately referring to the child via metaphor and negation/reversal. This marks a notable difference, for example, from the Leroy v. France cartoon, which was indeed about the 9/11 attacks—the latter being the target domain of the fundamental metaphor underpinning the cartoon, with the Sony product as the source.

In this sense, the Z.B. decision highlights an issue that also emerged in previous cases such as Nix v. Germany (2018), where the ECtHR confirmed the conviction of a German blogger for using Nazi symbols. In Nix, comparably to Z.B., the controversial symbols (a picture of Heinrich Himmler wearing a swastika armband) had only been evoked as the source domain—the metaphoric target being the alleged racism shown by an employment officer towards the applicant’s German-Nepalese daughter. As previously pointed out [36: 1027], the Nix ruling could have benefitted from greater attention to “the practical link between the expression and the actual ideology”; something similar applies to Z.B., where a joke using the 9/11 attacks as its source domain is automatically equated with an actual glorification of terrorism. As we aimed to suggest in this section, a closer engagement with the semiotic and rhetorical devices underlying the impugned jokes could help the ECtHR (and courts more generally) to achieve a more fine-grained analysis in this respect.

4 Intertextuality

In the present section, our focus shifts from the intrinsic features of the humorous text to its dialogue with other texts. The title uses the word intertextuality in this rather generic sense; however, some terminological clarifications may be useful before discussing specific cases. In his seminal work Palimpsests: Literature in the Second Degree (first published in 1982), literary theorist Gérard Genette distinguished between five types of what he called transtextuality, that is to say “all that sets the text in a relationship, whether obvious or concealed, with other texts” [16: 1]: (1) Intertextuality in the strict sense, namely echoing an excerpt from a previous text on a microscopic (e.g. sentence) level by means of quotation, plagiarism or allusion; (2) Paratextuality, i.e. a text’s relation with other texts physically surrounding it (such as titles, prefaces, back-cover descriptions, etc.); (3) Hypertextuality, describing a more structural relation between two texts, as in the case of parody or pastiche (akin to intertextuality in the narrow sense, but on a more macroscopic level); (4) Metatextuality, i.e. when a later text B can be seen (partly or entirely) as a commentary on a previous text A; (5) Architextuality, designing a text’s relation with the perceived conventions of the genre or genres it is ascribed to. Types 2 and 5 (para- and architextuality) will be discussed in the next section focusing on context at large, as they are usually subsumed under the umbrella text “context” in both legal cases and humor scholarship. Types 1, 3 and 4, instead, will be discussed together in this section, as they all refer to something more specific—namely the ways in which a given text (whether humorous or not) alludes to or comments upon a given, identifiable previous text.

To begin with, metatextuality is indeed an important dimension in some of our corpus cases. For example, in Kaboğlu and Oran v. Turkey, both the national courts and the ECtHR consider the impugned articles insulting the applicants as a response or commentary to the report published by the two academics. Nevertheless, the Government and the Strasbourg court assess this metatextual relation in very different terms. According to the national courts, the abusive language characterizing the articles was justified by “the attack which the applicants had allegedly launched in their report against their ideological adversaries by presenting them as paranoid” [25 at para 64]. In contrast to that, the ECtHR pointed out that the report simply expressed the authors’ opinion “without … using derogatory or insulting language in connection with those holding different views” [25 at para 83]. Somewhat comparably, M’Bala M’Bala v. France also involves a reflection on the metatextual dialogue between the controversial sketch and a previous critical review by French philosopher Bernard-Henry Lévy. According to the applicant, the idea of staging a sketch with Robert Faurisson “was a response to a provocation, [as] he was responding to Bernard Henry Lévy’s criticism—which in his view was exaggerated—of his earlier show” [34 at para 29]. However, both the French courts and the ECtHR concur that this does not constitute a sufficient justification for the comedian’s disparaging humor.

Other cases, instead, are more closely related to inter- and/or hypertextuality—the separation between the two is not always clearcut, as the micro and macro levels can hardly be distinguished when it comes to extremely compact forms of communication such as jokes or cartoons. When analyzing the interaction between a humorous text and a previous text, however, Genette’s typology of hypertextual relations might be especially useful. More precisely, Genette identifies four basic hypertextual operations: (1) Parody: Adapting the letter of a given text to a different situation, usually with a comic effect (whether at the expense of the original text, or of the situation being described); (2) Travesty: Transposing the subject/plot of a previous text into a different style; (3) Caricature: Imitating the manner or style of a previous text for satirical purposes; (4) Pastiche: Imitating the manner/style of a previous text, without a clear critical or satirical intent.

In particular, the notion of parody is relevant to cases like Leroy v. France and Bonnet v. France. Building on Genette’s work, Linda Hutcheon defined parody as a form of “repetition with critical distance” [23: 6], which is an apt description of how the impugned texts respectively deal with the Sony slogan (Leroy) and Charlie Hebdo’s Stromae front page (Bonnet). In both cases, the critical distance from the object of parody is actually integral to the overall message. The Leroy cartoon, for example, transposes and adapts the Sony commercial to an incongruous context, with an implicit satire of the consumerist rhetoric underlying the slogan; this intertextual twist ultimately reinforces the cartoon’s attempt to “criticize capitalism and American imperialism” [30 at para 34]. Similarly, in Bonnet, the parody includes an implicit criticism of Charlie Hebdo and its notorious stance against antisemitism—significantly, in an official statement on the case published on the applicant’s website, the magazine is accused of being “an organ of Zionist and Atlantist propaganda” [52].

In both Leroy and Bonnet, a stronger engagement with literary-theoretical work on parody would not have led to a different outcome; however, it would have certainly contributed to a more thorough analysis of the impugned expressions. This also applies to other cases that are not included in our corpus, but still concern humor and abusive language, such as Eon v. France (2013) for example.Footnote 6 Last but not least, Genette’s and Hutcheon’s definitions could also set the basis for more terminological consistency in ECtHR case law with regard to intertextuality (in the broad sense). Terms like “pastiche” [30 at para 6, 13, 34, 42; 5 at para 19] and “parody” [5 at para 1 and 5] are currently used interchangeably in Strasbourg jurisprudence, regardless of the specific nature of the inter- or hypertextual relation at issue; even more strikingly, the notion of parody is repeatedly evoked in Sousa Goucha v. Portugal [47 at para 17, 37, 50], despite the lack of any noticeable parodic element in the disputed joke.Footnote 7

5 Context

The notion of “context” is often referred to in free speech jurisprudence, and our corpus cases are no exception. After all, the subjectivity and slipperiness of humor make it particularly important to examine the contentious expression “in the light of the circumstances and the whole context” [34 at para 37]. This stance is perfectly in line with recent developments in humor studies, which place increasing emphasis on the pragmatic (i.e., contextual) dimension of humor production and reception. However, the term context can refer to several different factors involved in humorous communication, which might prevent courts from maintaining a consistent and exhaustive approach in this respect. Building on the theoretical overview provided by Villy Tsakona’s Recontextualizing Humor (2020), this section will explore the notion of context in its various dimensions, illustrating these dimensions through examples from the selected cases. Despite not being originally meant for the analysis of legal decisions, five of the contextual levels identified in Tsakona’s overview [49: 14] strike us as particularly relevant from a juridical standpoint:

  1. 1.

    Sociocultural assumptions on humor use, namely shared beliefs regarding appropriate and inappropriate uses of humor within a given culture or society. Despite not being particularly prominent in our cases, this aspect is often evoked in ECtHR humor jurisprudence—especially as a basis for acknowledging a wide margin of appreciation to national courts, which are considered best placed to assess the sociocultural specificities of humor within a given region.Footnote 8

  2. 2.

    Genre, i.e. the discursive form(s) to which the text belongs or is perceived to belong, based on its more or less creative dialogue with a given set of conventions. This dimension clearly relates to what Genette defined as ‘architextuality’ (see above, Sect. 4). Issues of genre are integral to the humorous framing of the impugned text/utterance in many of the corpus cases. In Sousa Goucha v. Portugal, for example, the joke is favorably framed by the courts in light of “the playful and irreverent style of the television comedy show and its usual humor” [47 at para 53]. In contrast to that, the sketch at the center of M’Bala M’Bala v. France is not ascribed to any specific stand-up comedy genre, but rather to the discursive mode of the (political) rally: “in the course of the offending sketch, the show took on the nature of a rally and was no longer a form of entertainment” [34 at para 39]. Conversely to humorist-turned-politician M’Bala M’Mbala, but with a similar outcome, politician-turned-humorist Jean Marie Le Pen does not enjoy the benefit of (comedic) genre, as his racist joke is clearly embedded within the serious discursive mode of political propaganda: “Jean-Marie Le Pen is not a professional comedian, which he does not object to; the impugned phrase was uttered in the context of a political speech” [31 at para 7, quoted from the Paris Court of Appeal]. Notably, in hate speech cases, politically motivated humor seems to be less protected than jokes which are considered as merely entertainment-oriented. While the reasoning in both M’Bala M’Bala and Le Pen is convincing, this criterion seems generally problematic, as it appears to establish an objective divide between dangerous political humor on the one hand, and gratuitous/aimless humor on the other—something that has long been contested in humor scholarship [33, 41]. The same applies to the mirror opposite of this criterion, namely the idea that humor engaging with topics of public interest is more worthy of protection than jokes that are perceived as “gratuitously offensive”—an argument that is applied by the ECtHR in Z.B. v. France among others.Footnote 9 Rather than being inherently considered more or less deserving of protection, genres (from stand-up comedy or infotainment to political speeches) can be effectively referred to by courts as contextually grounded sets of conventions, against which the specific features of a given text can be better assessed. Ogurtsov v. Russia also raises some interesting issues in this sense, since the contentious expression in this case belongs to the genre of football chants, which are notoriously characterized by hyperbolic, aggressive language against the opponent. Looked at from this perspective, the insulting tone of the impugned lines is somewhat tempered by their formulaic nature. However, these considerations can only remain speculative at this stage, since the case was only communicated by the Court to the Russian authorities.Footnote 10

  3. 3.

    Specific communication setting: the spatial, chronological and socio-political circumstances in which a given humorous text is produced and (predictably) circulated. This dimension is particularly broad, and its relevance is quite evident in most corpus cases. In Leroy v. France, the ECtHR’s finding of non-violation is supported by the fact that “the cartoon was published on September 13th [2001], when the whole world was still shocked by the news … in a politically sensitive region [: the French Basque Country]” [30 at para 45]. Comparably, Z.B. v. France highlights the “particular resonance” assumed by the T-shirt within its spatial and chronological context, with reference to the terrorist attacks carried out by Mohammed Merah in March 2012, which killed seven people including three children and a teacher at a Jewish school [55 at para 63]. Based on these considerations, the Court upheld the French courts’ finding that “the applicant cannot rely on the long delay between the attacks of September 11 and the wearing of the disputed inscriptions. The context of a terrorist threat would, on the contrary, be such as to increase its responsibility” [55 at para 42]. Moreover, a thorough examination of the communication setting should also extend to positionality issues, such as the social status of both the target and the author/speaker of the contentious expression. With regard to the target, for instance, Sousa Goucha v. Portugal stresses the need for public figures such as the applicant to be more tolerant towards public scrutiny and criticism; this does not apply, instead, in Kaboğlu and Oran v. Turkey, since the applicants’ status as academics “could not be compared” to that of politicians in this sense [25 at para 74]. As to the status of the speaker, Féret v. Belgium and Le Pen v. France seem particularly relevant, as they both aim to strike a fair balance between the freedom that should be afforded to political debate on the one hand, and the duty for elected politicians not to spread intolerance on the other [13 at para 75, 31 at para 37].

  4. 4.

    Co-textual cues, meaning the verbal or non-verbal hints surrounding the text and guiding its interpretation. This aspect is generally given adequate attention in our corpus, as exemplified by M’Bala M’Bala’s reliance on various cues to confirm the antisemitic nature of the impugned sketch:

[The applicant announced] by way of introduction that he intended to “do better” than in a previous show which had allegedly been described as the “biggest anti-Semitic rally since the Second World War”. The judges took the view that the sketch, presented by the applicant as a “quenelle”, an expression which, according to the Court of Appeal, evoked sodomy, had been addressed to persons of Jewish origin or faith as a community. That finding by the domestic courts was based on an assessment of the facts with which the Court can agree. In particular, it has no doubt that the offending sketch in the applicant’s show had a strong anti-Semitic content. [34 at para 34–35].

As acknowledged by the Court, the comedian’s introductory remarks—as well as his definition of the sketch as a “quenelle”—fulfill a role akin to Genette’s “paratext” (see above, Sect. 4), by serving as a threshold to the actual text and driving the interpretation.

  1. 5.

    Reactions and comments, i.e. the ways in which the audience’s response sheds light on the contextual functioning of a given joke. Once again, this dimension is well illustrated by M’Bala M’Bala, where the public’s reaction to the sketch is used to support the Court’s interpretation: “The reactions of members of the audience showed that the anti-Semitic and revisionist significance of the sketch was perceived by them (or at least some of them), as it then was by the domestic courts, the remark ‘Faurisson is right’ in particular having been shouted out” [34 at para 37]. Similarly, in Le Pen v. France, the audience’s “laughter and applause” are taken into consideration when assessing the potential harm brought about by the politician’s joke [31 at para 6, quoted from the Paris Criminal Court]. Considering this contextual dimension is easier when the humorous expression is performed in front of a live audience (as with stand-up comedy, jokes within political speeches, etc.), while it becomes more difficult in mediated forms of humor—from satirical novels to memes. In this latter scenario, as will be discussed in Sect. 7, courts are usually left speculating on how the joke could be potentially received by a hypothetical reasonable public.

As suggested by the overview outlined above, context is a multi-layered, composite notion, and a thorough examination of its various dimensions is vital in order to ensure a fair analysis of the humorous text. To this end, it might be worth mentioning one last contextual aspect, which literary theorist Liesbeth Korthals Altes (borrowing on Aristotle and Cicero) defined as the prior ethos of the author/speaker—i.e., “the image an audience already has of the speaker on the basis of his reputation, previous deeds, or generally known character traits” [26: 5]. The relevance of prior ethos to humor interpretation is fully acknowledged by the ECtHR in cases like M’Bala M’Bala, which stresses how the applicant had previously “displayed a strong political commitment by standing in a number of elections,” and “at the material time he had already been convicted for proffering a racial insult” [34 at para 37]. Similarly, in Bonnet, the applicant’s recidivism plays an important role in the final assessment of the proportionality of his sentence (“he had … already been definitively sentenced six times between June 11, 2008 and February 11, 2016, including twice for incitement to national, racial or religious discrimination, and once for defamation of an individual on the grounds of sexual orientation or gender identity” [5 at para 58]). On the other hand, a closer engagement with prior ethos seems desirable in Z.B. v. France, where the Court states, “the fact that the applicant has no ties with any terrorist movement whatsoever, or has not subscribed to a terrorist ideology, cannot attenuate the scope of the disputed message” [55 at para 60]. On the contrary, the speaker’s ideological ethos seems particularly relevant when it comes to criminal charges like glorification of terrorism—even more so when the impugned text is a rather ambiguous joke (more on which in the following paragraphs).

6 Meaning Reconstruction

The last three sections mapped the intrinsic and extrinsic factors that might play a role in the interpretive process—from the joke’s formal textual features (Sect. 3) and its dialogue with previous texts (4) to its interaction with context in its various senses (5). Based on this complex set of factors, courts are usually called to construe the meaning of the disputed humorous expression, with a view to assessing whether it amounts to disparaging and, ultimately, hate-inciting speech. The relationship between humor and disparagement is a multi-faceted one, as the former can not only be a disguise for the latter, but also a way to attenuate, deflect or even undermine the potentially derogatory component of a given utterance. While this topic is still vastly unexplored in humor scholarship, recent work has identified three basic outcomes for the interpretation of humor dealing with potentially disparaging tropes [18]: (1) Disparaging humor, aiming to dehumanize and vilify the direct (first-degree) target of the joke; (2) Sarcastic disparagement, e.g. using racist or sexist tropes in the second degree, as a way to denounce someone else’s racism or sexism; (3) Taboo-breaking humor, where the offensive component is supposedly not meant to disparage its first-degree target, but rather to question or disrupt the perceived taboo status of a given topic, as is often the case with disaster jokes [28]. This might be for the thrill of boundary-pushing, as a coping strategy, or in order to make a broader point about freedom of expression.Footnote 11

Most of our cases revolve around whether a certain expression should be considered as disparaging humor amounting to hate speech, or rather as taboo-breaking humor. In M’Bala M’Bala, the applicant argued that “his sketch was not to be interpreted in the first degree; he had sought to show that in France any allusion to the Holocaust which ran counter to the requisite respect for the latter was regarded as an aggression, whilst the questioning of other genocides was tolerated” [34 at para 29]. Likewise, in Bonnet, the applicant insists that the ultimate target of his humor was not the Jewish people, but Holocaust historians and their alleged dogmas [5 at para 46–47]. Another comparable case is Leroy v. France, as the cartoonist stated that he did not mean to disparage the victims of the 9/11 attacks, but rather to highlight inequalities and double standards in media representation: “What makes their [: US victims and their families] pain so much more media-worthy than that of the Iraqis bombed every month by American and British aircraft? … The tens of thousands of children who die each year from the direct causes of the global embargo imposed by the US on Iraq do not weigh very heavily in terms of current events” [30 at para 10]. As described above, in all three cases the ECtHR countered the applicants’ claims by relying on textual and contextual factors, thus making a convincing case for interpreting the jokes as outright disparaging as opposed to taboo-breaking.

In other cases, instead, the dispute does not focus on who is the ultimate target of the joke, but rather on the degree of aggressiveness of the contentious expression. In Le Pen v. France, for example, the applicant attempted—unsuccessfully—to downplay the aggressive nature of his statement, by relying on the polysemy of French verb voler, which can mean ‘stealing’ but also ‘flying’: “The domestic courts could not accept the classification as an insult, since the statements complained of created doubt as to the meaning which the prosecution had given to them, which could simply mean that it is in the nature of the Roma of Eastern Europe to move” [31 at para 9]. However, Le Pen’s statement clearly followed the standard structure of a canned joke, where the reference to the other meaning of voler is essential for the punchline to land—which is confirmed by the audience’s reaction (see above, Sect. 5). While not engaging with these formal considerations, both the national courts and the ECtHR understandably dismissed the politician’s claim. In Sousa Goucha, instead, the ECtHR concurred with the domestic courts, arguing that the joke was only mildly disparaging; its aim “had not been to attack the applicant’s sexual orientation,” and therefore did not meet the “‘certain level of seriousness’ requirement established in the Court’s case law” [47 at para 31, 20]. The Court here seems to rely on an implicit distinction between ‘casual’ forms of (homophobic, sexist or racist) disparagement on the one hand, and fully-fledged hate speech on the other—while the former can and should be contrasted by cultural means, only the latter can result in legally cognizable harm. Disparaging humor often falls into the grey area between these two terms; however, as will be further discussed in the final section, empirical humor research could provide some useful coordinates in this respect.

Lastly, Z.B. v. France deserves a separate discussion in light of the tripartite typology outlined above. As discussed in Sect. 3, the ECtHR echoed the Nîmes Court of Appeal’s interpretation of the T-shirt as an unmistakable example of (severely) disparaging humor against the victims of jihadist terrorism: “Certain attributes of the child (his first name, day and month of birth) and the use of the term ‘bomb’ … serve as a pretext to valorize, unequivocally, and through the deliberate association of terms referring to mass violence, willful attacks on life” [55 at para 11]. Despite the court’s assertiveness, though, the phrases on the T-shirt could just as easily be construed as taboo-breaking humor, where the alleged fun originates precisely from the perceived inappropriateness of the joke. However distasteful, this type of humor can hardly be equated to an apology for terrorism, just like the thousands of disaster jokes that started circulating soon after 9/11 do not amount to a glorification of the attacks [28]. Even more importantly, the T-shirt can also be interpreted as an instance of sarcastic disparagement, appropriating and implicitly debunking the common Islamophobic trope whereby all Muslims are considered terrorists until proven otherwise [54]—let alone a family where a child is called Jihad (as previously mentioned, the name Jihad is actually quite common in the Arab world, and does not have any inherent fundamentalist connotations).Footnote 12 This type of ironic (self-)disparagement is indeed quite widespread in the work of comedians from a Muslim background [2]. Shortly after 9/11, for example, British stand-up comedian Shazia Mirza famously opened her set with the line “My name is Shazia Mirza, or at least that’s what it says on my pilot’s license” [37]—which far from being a glorification of terrorism, was meant as a sarcastic critique of increasingly widespread prejudices against Muslims after the attacks. However, it is not our goal here to identify one correct interpretation of Z.B.’s T-shirt, as the joke inevitably retains a degree of subjectivity. However, regardless of what one might think of the final outcome of the case, the Court of Appeal’s hasty classification of the joke as an apology for terrorism (which was readily accepted by the Court of Cassation and the ECtHR) exemplifies a concerning tendency to automatically interpret humor on highly sensitive topics as disparaging humor—thus leaving little room for legitimate (however shocking or distasteful) forms of ironic disparagement or taboo-breaking humor. Looked at from this perspective, the Z.B. decision could have indeed paid greater attention to the specific textual features of the joke and the ethos of the speaker (as suggested in Sects. 3 and 5 respectively), as well as to the audience presumably addressed by the applicant (to which we now turn).

7 Actual and Presumed Audiences

As shown in the previous paragraphs, humor often lends itself to contrasting interpretations. As a consequence, courts often have to shift their focus from the production of a given joke to its reception—i.e., borrowing Stuart Hall’s terminology, from the “encoding” to the “decoding” of the message [21]. This faces courts with the task of distinguishing between reasonable interpretations of the joke (for which the speaker might be held accountable) and misreadings for which the speaker clearly bears no responsibility. In order to do so, judges can occasionally rely on the empirical evidence provided by an audience’s reaction to the contentious humorous utterance—this is the case with M’Bala M’Bala and Le Pen v. France, where the public’s response to the comedian’s sketch and the politician’s pun sheds light on their plausible reception (see above, Sect. 5). However, in most instances of mediated humor without a live audience, this empirical test is unavailable. To make up for the absence of an actual public, courts usually resort to the legal fiction of the “reasonable reader”—namely the idea that the speaker should be held responsible for how her/his utterance could be interpreted by a “reasonable” (or, depending on the preferred phrasing, “ordinary,” “right-thinking,” etc.) member of the public. In our corpus, this criterion is used—more or less explicitly—in Sousa Goucha, Kaboğlu and Oran and Z.B.:

The Court further observes that in the judgment of Nikowitz and Verlagsgruppe News (cited above) it introduced the criterion of the reasonable reader when approaching issues relating to satirical material. … [The domestic courts] considered that for a reasonable person, the joke would not be perceived as defamation because it referred to the applicant’s characteristics, his behavior and way of expressing himself. [47 at para 50, 53]

Certain passages of the articles are ambiguous in that they would seem to be stereotypical phrases with a nationalist ideological wording, but could also be read as condoning violence, at least by some readers with insufficient knowledge of the jargon in question who are liable to take the words in question literally. [25 at para 84]

Certain attributes of the child (his first name, day and month of birth) and the use of the term “bomb,” which cannot reasonably be claimed to refer to the beauty of the boy, are magnified through the turn of phrase, […] and in reality serve as a pretext to valorize, unequivocally, … willful attacks on life. [55 at para 11; our emphasis]

To be sure, the “reasonable reader” test is a well-established criterion in legal doctrine, and can prove useful when tackling the ambiguity of a given utterance (whether humorous or not). At the same time, as Z.B. in particular illustrates, this notion lends itself to being used rather subjectively, and to inadequately replace a thorough analysis of the different interpretations potentially allowed for by the same text. More generally, from a pragmatic and literary-theoretical perspective, the notion of the reasonable reader is “vague and unsatisfactory, not least because it begs a series of large questions about who the ‘reasonable’ or ‘right-thinking members of society’ might actually be” [35: 186].

The vagueness of the reasonable reader standard, and the need for a more context-sensitive approach, are particularly evident in Féret v. Belgium—where the majority found it necessary to consider not only the allegedly reasonable interpretation of the applicant’s jokes and slogans, but also the “irrational” response they were likely to trigger among people more prone to xenophobia: “Such a discourse is inevitably bound to create in the public, especially among the less informed [moins averti], sentiments of contempt, rejection, and hate towards foreigners” [13 at para 73]. This deviation from the usual test was harshly criticized by the three dissenting judges: “The judgment considers some human beings and a whole sector of society as ‘simpletons’ incapable of responding to arguments and counterarguments due to the irresistible pulsion of their irrational emotions” [13, dissenting opinion by Judges Sajó, Zagrebelsky and Tsotsoria]. Although the phrasing used by the majority could have been more nuanced, the Féret ruling does mark a welcome attempt to move beyond the abstract notion of “reasonableness,” in order to consider the (often unreasonable) effects that jokes or slogans can have in a particularly heated political context:

If, in an electoral context, political parties must enjoy a wide freedom of expression in order to try to convince their voters, in the event of racist or xenophobic speech, such a context contributes to stirring up hatred and intolerance because, by necessity, the positions of the candidates in the election tend to become more fixed, and stereotyped slogans or formulas come to take precedence over reasonable arguments. The impact of racist and xenophobic discourse then becomes greater and more damaging. [13 at para 76]

In literary-theoretical terms, the Féret majority decided not to focus exclusively on a hypothetical reasonable audience, but rather on the “presumed addressee”—namely the public among which the author can rightfully expect their work to be circulated [43]. As previously suggested by Godioli and Little [19], the concept of the presumed addressee is inherently more sensitive to context than the reasonable reader, and can therefore usefully complement the latter when assessing the plausible reception of a given contentious utterance.

With this in mind, the Z.B. decision could have taken into greater consideration the presumed addressee of the disputed joke, since—as ascertained by the first instance court of Avignon—the child only wore the T-shirt only “on one occasion” which was “limited in time (the afternoon of 25 September 2012) and space (the nursery class),” and “only two people had been able to see the words on the T-shirt while dressing the child” [55 at para 8]. Needless to say, a class of 3-year old children is far from constituting a politically inflammable audience, prone to being radicalized by dubious inscriptions on a T-shirt. As to the preschool director and the teacher, i.e. the two adults who saw the T-shirt while dressing the child, they might well have been shocked by the joke; but they could hardly have been turned into jihadists, which is the scenario evoked by the criminal charge of apology for terrorism. Nevertheless, the fact that the T-shirt was only worn at preschool was exclusively acknowledged as an aggravating factor by the Nîmes Court of Appeal, on the grounds that schools are a highly symbolic “forum reserved for learning and the transmission of knowledge” and that the words on the T-shirt “could only have the effect of deeply shocking the staff of the establishment.” The Z.B. judgment only mentions these arguments without problematizing them, despite the ECtHR’s general commitment to also protect expressions that “shock, offend or disturb” (as famously stated in Handyside v. United Kingdom). To be sure, the Court is right in pointing out that it “cannot speculate on the exact nature of the applicant’s intentions” as to when and where the child would have been made to wear the T-shirt after that one afternoon in September 2012 [55 at para 62]. However—along with the textual features of the joke, the prior ethos of the applicant and the ambiguity of the message (see Sects. 3, 5 and 6 respectively) –, the presumed addressee would have deserved a closer examination from both the Court of Appeal and the ECtHR, with a view to ensuring a thorough analysis of the contested expression.

8 Conclusion

By discussing and comparing our ECtHR cases, the previous sections have illustrated how linguistic, semiotic and literary-theoretical research can assist courts in the various phases of the interpretive process; i.e., the analysis of the textual features of the joke (Sect. 3), the latter’s dialogue with previous texts (4), the role of context in its multiple dimensions (5), meaning reconstruction (6) and engaging with actual or presumed reception patterns (7). However, the interpretation of the disputed expression—namely the reconstruction of its possible meaning(s) and their reception—does not usually coincide with the end of the proceedings. After determining whether a joke can reasonably be classified as disparaging humor, courts still have to assess the extent to which this actually amounts to hate speech, and potentially results in legally cognizable harm for the target. As mentioned above, disparaging humor often falls in the grey area between mere offense on the one hand, and objectively harmful expression on the other. Empirical research on humor reception can provide courts with some useful input in this sense. For example, a recent article on the effects of sexist humor showed how exposure to this type of jokes can act as a “releaser” of prejudice within a given community, by facilitating the expression and acceptance of discrimination or violence towards the target group [15]. Further research in this direction is certainly desirable, and is likely to have significant implications for juridical debates on the line between offense and harm in disparaging humor. However, this exceeds the scope and methodological framework of the present article.

Another promising avenue for future research concerns the specific interpretive challenges posed by humor in the digital age, namely an era in which jokes, cartoons or memes can easily circulate worldwide at unprecedented speed. This raises new questions regarding, for instance, (1) the increased difficulty of mapping the contexts in which a joke may circulate; (2) consequently, the ways in which online communication tends to blur the boundaries of the presumed addressee; and (3) the importance for courts to be aware of widespread conventions within relatively recent genres of online humor, such as memes [44]. Although a few references to online circulation are made in Féret v. Belgium [13, dissenting opinion] and Bonnet v. France [5 at para 43 and 52], these aspects are not directly relevant to our corpus cases. Nonetheless, due to the increasing regulation of online space, future ECtHR jurisprudence will inevitably have to engage more with the specificities of online humor in relation to free speech and hate speech. This should of course be accompanied by further interdisciplinary research on the subject, in dialogue with the growing body of scholarship on related phenomena such as trolling and cyberbullying [7, 14, 53]. For the time being, the present contribution has sought to shed new light on recurring issues and challenges underlying different ECtHR cases concerning humor and hate speech, as well as suggesting how insights from the humanities can set the basis for a more consistent and nuanced analysis of humorous expression in courts of law.