A large rose-tree stood near the entrance of the garden: the roses growing on it were white, but there were three gardeners at it, busily painting them red. Alice thought this a very curious thing, and she went nearer to watch them […]. “Would you tell me,” said Alice, a little timidly, “why you are painting those roses?” (Carroll 1865).

Almost ten years after the Tunisian revolution, narratives of the event are still vividly debated. History textbooks are still awaiting new chapters, and memorials and monuments have yet to be inaugurated. The “transition period,” associated with the disappointing process of transitional justice, seems to be extending longer and longer without clearly easing the memorial tensions related to the 2011-revolution. Competing narratives coexist and the disagreements are visible in public spaces all over Tunis. They are perceptible on the walls of the city: on the colorful murals that have stood the test of time since 2011 and under the corrective coat of white paint covering certain facades. They are personified through names: some of them engraved on elegant marble stone in front of the Interior Ministry, others handwritten on banners waved in demonstrations every January throughout Tunisia. They are represented by bodies: proudly posing, victorious, on billboards in downtown Tunis or shown injured or dead on posters displayed during gatherings of associations of victims’ families. In the fight over the revolution’s political meaning, images and terminology are crucial and are strategically used to play one narrative against another. Any post-revolutionary society must deal with the crucial question of the authority to recount what has happened. Who should have the power to choose the words with which the history will be narrated and the names and faces who will symbolize that history? In this chapter, I will focus on the re-configurations of the uses of the word “martyr” (shahīd) and of associated graphical representations. I seek to analyze the processes through which, in post-revolutionary Tunisia, some dead are officially and publicly honored while some others remain invisible and doomed to oblivion. Drawing from a gender perspective, I shall show that rival masculinities each of them associated with different political and moral meanings, compete within the national memory and shape the strategic labeling of “martyrs of the revolution” (shuhadāʾ al-thawra).

After some introductory comments about masculinity studies, the “Arab Spring,” and some methodological remarks about my own position as researcher in the field, I will describe ways in which the issue of martyrs took on major political importance in post-revolutionary Tunisia. Families and local collectives first used the label before authorities took charge of the project of listing the “true” martyrs of the revolution, providing an illustrative case of the selective processes through which national narratives are constructed and implying the marginalization of subaltern experiences. In public spaces and on official memorials, “martyrs” appeared to have no faces besides those of uniformed policemen and soldiers, whom official institutions portray as powerful, muscular, and heavily-armed paragons of masculinity.

Arab Masculinities in Question

There is no need to bemoan the lack of research in gender studies about the 2011 revolutions in the Middle East and North Africa (MENA) region. Some recent academic works demonstrate that sex categories are social constructs and privileged tools for organizing and performing specific social orders; thus, these structures help us to better understand the re-configurations in the MENA region since 2011 (Ryzova 2011; Ghannam 2013; Salime 2015; Kréfa 2016). Yet Tunisian society largely remains understudied in the numerous special issues, papers, and conferences devoted to the topic; the underrepresentation of the Tunisian case among academic work on the MENA region is a preexisting tendency and in fact nothing new (Dakhlia 2011). Of the few insightful papers written on the 2011 Tunisian revolution from a gender perspective (Khalil 2014; Kraidy 2016; Kréfa 2016; Gronemann 2017), an even smaller number deals with masculinity. Existing studies give the impression that gender is synonymous with “women.” This classical misconception, as Maya Mikdashi argues in a recent piece published in the online magazine Jadaliyya, has damaging effects: it creates a false dichotomy between gender and politics, reinforcing “a framework within which sexual or gendered violence happens to women and sexual minorities while political violence happens to men” (2017). Some analyses describe the uprising as an occasion for Tunisian men to prove their “virility” and show the whole world that they are “real men” (Piot 2011). Journalists report that Mohamed Bouazizi, whose self-immolation triggered the Tunisian revolution, was supposedly slapped by a female police officer before he set himself on fire in December 2010, interpreting this sequence of events as an illustration of the over-importance of “honor” for Tunisian men (Belhassine 2011) without mentioning the social and political crisis Tunisia was facing at that time. Whether or not this famous slap ever occurred—it likely did not (Larguèche 2011)—this approach demonstrates the prevalence of the “masculinity crisis” interpretative frame with respect to Arab societies (Amar 2011). Masculinity studies has fortunately shown that its academic and critical ambitions are more sophisticated than simply corroborating stereotypes about Arab men (Inhorn 2012). Researchers involved in that field follow feminist scholars’ invitation not only to pay attention to the power relations between men and women but also to question the hierarchies within each category. They examine men’s experiences and develop intersectional analyses that take into account criteria besides gender, such as class, race, sexual orientation, or age (Crenshaw 1991; Connell 1987; Ghannam 2013). “Multiple masculinities coexist in relations of power, contestation, and negotiation,” Maya Mikdashi asserts, outlining what she considers to be one of the major premises of masculinity studies (2017). Following these approaches, I will explore the gendered dimension of narratives linked to the revolution, investigating the processes through which different models of masculinity have been framed and re-configured in post-2011 Tunisia. My main argument is that they do indeed inform the political challenges faced by the post-revolutionary Tunisian society.

Fieldwork: Investigating a Polarized Context

This chapter draws on empirical research conducted in Tunis during several stays in the field between 2016 and 2018. In addition to observational work conducted at political meetings and demonstrations, I interviewed twenty people representing the main groups engaged in the debates regarding the public recognition of the “martyrs of the revolution,” including representatives of martyrs’ families, human rights associations, political parties, transitional justice institutions, and governmental commissions.

As a researcher, I have been affected by my fieldwork, just as I have influenced it myself. Acknowledging this reciprocal dynamic is one of the premises of social anthropology (Favret-Saada 1977). In the field, my own gender, experiences, and emotions constituted sources of knowledge, and I particularly paid attention to the ways my informants used my presence, understood my requests and interpreted my interactions with them. Despite my frequent clarifications, I was generally perceived as a French journalist whose imagined objectives varied depending on the position of my interlocutors. The institutional representatives saw me as an opportunity to demonstrate their efforts to overcome challenges of transitional justice for an international audience; political leaders tried to defend their own strategies and alliances; human rights activists frequently suggested that there were more important issues to focus on; and members of associations for martyrs’ families saw me as a potential ally—my gender drove them to call me “sister” or “daughter”—in their fight against national amnesia, expecting me to publicize their struggle in Europe. Gaining families’ trust and collecting their stories under such conditions often made me feel uncomfortable, and I repeatedly confessed that I seriously doubted my research could have any positive impact on their situation. As a researcher investigating the issue of the “martyrs of the revolution,” whether I liked it or not, I became a part of the debate. My image was invested with suspicions and hopes. My picture was posted on Facebook walls, combined with different public and private comments. One of my informants had a dream about me and her deceased brother, while another tried to extract details of what other interviewees had told me. All sides warned me about the risk of the others manipulating me. In any field, there is no such thing as a neutral position—and this is probably even truer in a polarized context such as the environment I found in post-revolutionary Tunisia.

The Early Days of the Revolution: The Impulsive Designation of “Martyrs”

The Tunisian upheaval started with the self-immolation of Mohamed Bouazizi in Sidi-Bouzid in December 2010, and his death a few days later. As demonstrations multiplied in several Tunisian cities and social movements intensified in December 2010 and January 2011, the death toll increased. Repression was especially brutal in the marginalized governorates deep inside the country, where the conflict began. The longer the list of victims grew, the more the commemoration of the “martyrs of the revolution” became visible in public spaces all over Tunisia. From the first days of mobilization, memorials were erected in city centers to honor local “martyrs.” Streets, schools, and squares were spontaneously renamed after some of them, usually in their families’ neighborhoods where they had lived (Fig. 1).

Fig. 1
A photo of a label. The text reads Place des Martyrs.

“Martyrs’ Square,” Gabes, April 2016 (P. Lachenal)

These commemoration processes, including the use of the label “martyrs,” reached a national level in March 2011, when the Tunisian postal service released a stamp bearing Bouazizi’s face, commemorating his “martyrdom” and the “revolution of dignity” (thawrat al-karāma)1(Fig. 2).

Fig. 2
A stamp of 600 with the face of Le Martyr Mohamed Bouazizi. The text on the stamp reads, Revolution De La Dignite, TUNISIE.

Stamp dedicated to the “martyr” Mohamed Bouazizi. March 11. Screenshot from the author

In downtown Tunis, a square previously known as the “Seventh of November,” the date Ben Ali’s regime came to power, was renamed “Fourteenth of January, 2011,” the day he resigned. The celebration of the uprising implied the material destruction of the previous symbolic system, a phenomenon Bernhard Giesen describes in his major piece on martyrs and heroes (2004, p. 34). In a few weeks, one could observe Ben Ali’s portraits disappearing from the walls of shops, administrations, and restaurants, sometimes replaced with faces of local “martyrs.”

If the vast scientific literature devoted to “martyrdom” insists on the polysemy of the category (Mayeur Jaouen 2002; Verdery 2004; Blunden 2006; Buckner and Khatib 2014; Bozarslan 2015; Mittermaier 2015), the ways it was used in revolutionary Tunisia in fact suggest clear nationalist and secular dimensions.2 Indeed, the word “martyrs” (shuhadāʾ) was usually followed by the phrases “of the revolution” or “of the nation” (al-thawra/al-waṭan). This specific use is not surprising since the word is historically ingrained in the Tunisian national vocabulary (Bendana 2018), especially since the pro-independence struggles. The political function of the martyrs of the “Arab springs” nevertheless differed slightly from that of their predecessors, as Elisabeth Buckner and Lina Khatib assert (2014): the 2011 martyrs were used to unify a people against an oppressive regime, not to construct a collective identity against a foreign occupier. For the family members I interviewed, and in line with what Ieva Zakarevičiūtė (2015) witnessed in 2011 Egypt, the word “martyrs” served as a common reference point with the capacity to lend meaning to violent and unfair disappearances, integrating them into a national liberating narrative. Using the word “martyr,” I was told, implied recognizing that the person did not die in vain but for a cause, and that his or her death ought to be collectively honored.3 The religious dimension of the “martyr” category was rarely invoked during my fieldwork, although, in intimate spaces and informal discussions, my informants sometimes wished that God would accept the deceased as a “martyr” and offer them post-mortem rewards and peace.4

A Strategic Cause on the Post-Revolutionary Stage

In the months that followed the fall of Ben Ali, the issue of the “martyrs of the revolution” acquired major political importance. The first trials of people accused of killing demonstrators were accompanied by both national and international media coverage and street gatherings. Press conferences had to be held in very large hotel conference rooms to accommodate crowds of journalists, and tribunals did not have enough space to welcome all the “martyrs’” families wanting to attend the adjudications. “Martyrs of the revolution” became a strategic source of legitimacy for those who wanted to play a role during the so-called “transitional” period. They served as a rhetorical tool and from 2011 on, any public speech regarding the revolution started and ended by referring to them and reminding listeners of the obligation to honor their memories and sacrifices.5 An activist told me jokingly that, in 2011, “It looked very good to appear among the martyrs’ families and to explicitly support their cause. Each political group wanted to have its own martyr!”.6 This quotation refers to the question of “ownership” of martyrs and their images. “It is not always obvious who ‘owns’ a martyr,” Andy Blunden argues; “A social movement or institution is going to be able to claim ‘ownership’ of a martyr if it shares the icons and symbols of the subject” (2006, p. 51). Contradicting interpretations could challenge each other, as was the case for Mohamed Bouazizi, who was sometimes incorrectly presented as an “unemployed graduate” in order to strengthen the demands of this specific demographic group (Desorgues 2013). Other appropriation attempts were aimed at the family members of martyrs. The sister of a “martyr” told me that as soon as she decided to begin wearing a hijab in 2011, she was approached by several religion-based groups who offered to take on the case of her brother’s death.7, 8 More generally, families had to distinguish potential allies from profiteers who tried, as I was told, to “make money off the cause”9 and to transform the martyrs into a business, as Kinda Chaib (2007) describes in great detail for the case of Lebanon.

During the period that followed Ben Ali’s resignation, several associations of martyrs’ families were established in Tunisia. One of them was Lan Nansākum (We Won’t Forget You), which developed into a powerful mobilizing force and soon became a major participant in the debate over the “martyrs of the revolution”’ cause. It distinguished itself with its creativity, mobilization strategies, and political statements. This was especially remarkable considering that almost none of its members had been politicized prior to 2011 (Mekki 2017). The association organized regular marches in Tunis, occupying the front lines of demonstrations carrying the portraits of their deceased relatives. Mothers dressed in black were particularly visible among them.10 The demands were clear: the truth about the people’s repression had to be voiced and the individuals responsible for the killings had to be judged and punished, both those who had issued the orders to open fire on the demonstrators and those who had obeyed said orders.

Suspicious Masculinities: In Search of the “Fake” Martyrs of the Revolution

Given the sensitivity of the topic, official measures were soon taken regarding the “martyrs of the revolution.” In October 2011, a decree was published in an effort to regulate the ways in which post-revolutionary Tunisia would approach them (Ben Ghozi 2012). The decree’s first five articles focus on the “national memory” and on the collective duty to honor the people who died during the revolution with memorial sites. The decree also introduced the project of publishing an official list of the “martyrs of the revolution.” Having the names of their relatives on that list would be a prerequisite for the families to benefit from some compensation measures. This project of listing the “martyrs” immediately appeared problematic since it implied that, among the 338 people who were reported dead during the revolution,11 not all of them would be recognized as such. Samir Dilou, as the head of the Ministry of Human Rights and Transitional Justice, asserted this in 2012, publicly expressing doubts about the validity of the first list and the veracity of some victims’ stories. A few years later, when I asked the head of the General Committee devoted to the “martyrs” about the necessity of creating such a list, she expressed a similar suspicion mentioning that “not everyone told the truth” during the initial investigation into repression by the police and the army:12

There are people who died, but not in the context of the revolution, and their families would like to benefit from the social advantages accorded to the martyrs’ families! … This revision will take some time because we have to check every single case in order to identify the fake martyrs. I understand the families’ impatience; it has already been five years since the revolution occurred and the names of their children are still not public. For the families of the true martyrs, this must be disappointing. [Emphasis added].

Later on in the same conversation, my interviewee once again stressed the importance of identifying the “true martyrs” and removing the names of those who do not deserve the title, calling them not only liars but also “thugs”: “We have to tell the truth and who the true heroes are. There is no way thugs’ names should appear on the list” [emphasis added].

If the decree specified the composition of the commission in charge of the martyrs list, which included representatives of civil society and of seven ministries, it remained silent about the concrete criteria according by which a person should or should not be recognized as a “martyr of the revolution.” Only the chronological criterion is publicly known: a “martyr of the revolution” should have died between December 17, 2010 and the February 28, 2011. The few interviews I conducted with some members of the commission did not help me to better identify the other criteria. Indeed, the selection process seemed largely to be shaped by the commission members’ subjective perceptions of heroism and heroic masculinities. For example, the circumstances of death played a major role in the selection process and it seemed problematic, as I was explained, to recognize as “martyrs” any young men who were killed while attacking the police or who died during a fire at the prison in which they were detained.13 Actors’ social status determined the ways their violence and political commitment could be publicly depicted, and the meaning of their death. Beyond the political dimension, which one could have reasonably expected in the context of a revolution, it seemed that ultimately the social actors’ “good morals” or lack thereof was the decisive argument determining whether or not they would be labeled “martyrs.” If the young men who died while in jail or fighting the police were not publicly remembered as “martyrs,” it is because they embodied models of masculinity that were suspicious due to a lower class status. Because their social and gender performances were framed as immoral and violent, and their actions were perceived as threats to the social order by those in charge of determining the official revolution narratives, their names would remain unknown and their faces invisible.

The focus on the morality of the “martyrs” was particularly visible in the discourses surrounding the cases of the many prisoners who died in 2011 during the destruction and burning of their prisons. These accounted for almost a quarter of the casualties of the Tunisian uprising.14 As various people in charge of generating the “final” list told me in interviews, it seemed problematic to publicly assign prisoners, who were probably sent to jail for committing crimes, the prestigious status of martyrs. The same reservation was expressed regarding the young men who were killed at the margins of the protests. In these cases, the disqualification was often based on the probable immorality of the youngsters, embodied through the suspicion that they had died while drinking. When I met him in his office, in September 2016, one of the members of the Bouderbala commission explained to me:

Someone who was reported dead, burned alive inside the Magasin général, a supermarket, at midnight, was not expressing any political discontent—in the middle of the night, in a supermarket, in the spirits section! (laughs) With someone who died outside the trade union center or the Interior Ministry, it is totally different.15

Using gendered rhetoric as a tool, the authorities contributed to removing entire social groups from official Tunisian history without having to acknowledge any embarrassing exclusion process. Their choices were justified through terms such as “troublemakers,” “drinkers,” “vandals” and other references to problematic and suspicious models of masculinity.

Subaltern Voices and Alternative Narratives

Confronting the state’s attempts to monopolize the public making of martyrs in post-revolutionary Tunisia, other narratives were constructed and circulated through unofficial channels and family networks. The more visible the suspicion regarding the “good” morality of the young men who died during the revolution became, the more the “martyrs’ families” publicly stressed the political dimension of their relatives’ deaths. Intentionally or unintentionally, they argued, those young men died while participating in the Tunisian revolution and took part, albeit indirectly, in its successful ending: the resignation of the former President. Contesting the top-down nature of the labelling of “martyrs,” “martyrs’ families” resisted the depoliticization and the moralization processes of the revolutionary subaltern experiences. The interviews I conducted with some of them stressed the decisive nature of the rhetorical dimension, expressing concerns about how the Tunisian revolution was starting to be publicly depicted. Terminological debates had clear political meanings, as I was told; for example, the word “victim” was strongly contested by the “wounded of the revolution.” According to Bachir, a 28-year-old man who was seriously burned in 2011 and partly lost the use of his hands, being called a “victim” is insulting:16

I am one of the “wounded of the revolution” (jarḥā al-thawra). To say “wounded” is to recognize that we were actors of the revolution, we were fighters who got injured. But “victim” (ḍakhiyya) is not a political category. The word is pejorative; it implies passivity.

Bachir’s words expressed the opposition of agency versus victimhood, defending the political rationality and intentionality of working young men who took part, sometimes violently, in the movement against the previous regime and were injured or killed.

These alternative narratives had graphic translations. Martyrdom’s political dimension was for example aesthetically asserted through the portraits “martyrs’ families” and relatives carried with them and waved in public events, illustrating the determination to link the killings of their relatives to the revolution. Such portraits could be seen as tools to build a “durable biography,” an expression used by Farha Ghannam about the “technologies of immortality” aiming to shape the public afterlives of martyrs (2015, p. 640). The following picture, taken in January 2018 during a gathering of martyrs’ families, the face of Qaïs Al-Mazlini, killed on the 13th of January 2011, appears on an expressive background: an image of a demonstration on Bourguiba Avenue, a Tunisian flag and the sentence “Tunisia is free” (Fig. 3).

Fig. 3
A photo of the face of Qais Al-Mazlini. The text is in a foreign language.

“The martyr Qaïs Al-Mazlini. Martyrized on the 13th January 2011″. Families’ exhibition, Tunis, January. (Foto by author)

These symbolic references are tools to engrain the individual story into the national one, if there was any doubt about that connection. Taoufic Haddad identifies similar aesthetic choices when analyzing a large corpus of posters depicting Palestinian martyrs (2016): it is common for the families or the poster designer to make the deceased person appear as an armed fighter, even if he was not, so as to include him in the core of the national struggle.

In post-revolutionary Tunisia, the state was not the only actor to build and broadcast martyrs’ iconographies. The role of non-state actors in the production of alternative narratives and collective memory is partly based on practical considerations such as the democratization of access to technologies for designing and printing. If disagreements have always plagued memorial practices, this recent technological shift has given them exceptional visibility (de Georges 2013), and has demonstrated major re-configurations in the public making of martyrdom.

“Martyrs of the Nation”: The Consecration of Militarized Masculinities

While families continued to request the release of an inclusive list remembering their own children as “martyrs of the revolution,” other kinds of martyrs were being publicly identified. Their faces were displayed by the Tunisian army and police in public spaces and their names carved on commemorative plaques: these were casted as the “martyrs of the nation” (shuhadāʾ al-waṭan), meaning the members of the Tunisian security forces who lost their lives not only during the 2011 revolution but also in the years that followed. The terrorist threat that emerged in the country in the wake of 2011 and the concomitant state of emergency were incorporated into the process of re-configuration and securitization of the national narratives, paving the way for the celebration of martyrs from the police and military. Progressively, uncertainty seemed to have helped set the stage for celebrating victorious and heavily armed masculinity models that were perceived as reassuring.

In 2016, the name of the governmental commission devoted to the “martyrs and wounded of the revolution” was significantly extended to include the words “and of the terrorist attacks.” Madjouline Cherni, a sister of a soldier killed in a 2013 terrorist attack, was chosen to direct it. Since 2011, intensive campaigns have glorified the commitment of the security forces to the Tunisian nation. Advertisements popped up all over Tunis. The picture in Fig. 4, which was distributed mostly in 2017, illustrates the process just described.

Fig. 4
A photograph frame depicts an armed man wearing glass and holding a gun in his hand. The text on the frame is in Arabic.

You are the hero. We fully support you. Terrorism “Dégage” Mohamed V Avenue, downtown Tunis, September 2017. (Photo by the author)

By mixing together different repertoires, applying the well-known slogan “Dégage”—an iconic fixture of the “Arab Spring”—in reference to “terrorism,” the fight against terrorism was integrated into the story of the revolution itself and military “heroes” were placed at its forefront. Doing so seemed to absorb the revolutionary moment into state hegemonic narratives. These kinds of catchy advertisements contributed to the processes of depoliticization and securitization—as well as virilization—that attended the commemoration of the Tunisian revolution. According to Daniel Gilman, a state’s ability to build and glorify conservative martyr figures, such as those embodying order and security, mainly depends on a prior depoliticization process that is necessary to “set the stage” (2015, p. 693).


By determining those who should be officially honored and those who should be forgotten, drawing on gendered repertoires of morality and respectability, the Tunisian state sought to frame not only the history but also the meaning of the revolution, and the reasons why it occurred. In this chapter, I have shown that the public designation of “martyrs” influences how the 2011 uprising is narrated and understood, but also, more importantly, affects its political direction. Narrating and labeling is never only about the past. Behind the objective of writing history lies the question of the present and future exercise of power.

As I complete the final version of this paper in August 2019, the final list of the “martyrs of the revolution” has still not been published in Tunisia, eight years after Ben Ali resigned and several months after the responsible commission delivered it to the government. The issue of the “martyrs of the revolution” obviously remains delicate. Omitted deaths continue to haunt the post-revolutionary society through the mobilization of their families, relentlessly weakening official attempts to standardize revolutionary narratives.


  1. 1.

    Bouazizi was sometimes referred to as a “Pan-Arab martyr” in the sense that his image widely circulated and was used as a shared signifier during the social movements that shook many countries after Tunisia, appearing for instance on Tahrir Square and on the walls of Cairo (Buckner and Khatib 2014).

  2. 2.

    The nationalist and secularist acceptation of martyrdom has largely been examined in other societies of North-Africa and West-Asia (Mayeur-Jaouen 2002; Chaib 2007; Haddad 2016). Comparative studies are nevertheless missing, although they could be useful to show similarities and to undermine the conventional idea that martyrdom is “somehow more central to Arab or Muslim cultures than others” (Hayoun 2012).

  3. 3.

    Personal interviews in Tunis, January 2018.

  4. 4.

    Home visits in Tunis, May 2016.

  5. 5.

    Observations, demonstrations in Tunis, January 2017 and 2018.

  6. 6.

    Personal interview in Tunis, May 2016.

  7. 7.

    Personal interview in Tunis, September 2016.

  8. 8.

    Similar ideas about feminine performances of piety influenced the use of Sally Zahran’s images in revolutionary Cairo. Several controversies surrounded the story of this young women who died in January 2011, one of them being her seemingly problematic refusal to wear a headscarf. Some of her photos were manipulated to cover her hair, bringing her appearance in line with Egyptian Islamic standards and enabling social groups to use her portrait (Armbrust 2013).

  9. 9.

    Personal interview in Tunis, May 2016.

  10. 10.

    The “grieving mother” is an emotionally and politically vested figure, commonly constituting a powerful mobilizing force (Latte-Abdallah 2006; Lorber 2002).

  11. 11.

    The Bouderbala commission was the first official attempt to identify those who died and were injured during and following the Tunisian uprising. It was commissioned by the interim government in 2011 to conduct investigations in jails, hospitals, and several governorates in order to register and document abuse committed during a specific timeframe, from December 2010 to October 2011. The commission’s 1,041-page report was published in 2012.

  12. 12.

    Personal interview in Tunis, May 2016.

  13. 13.

    Personal interview in Tunis, September 2017.

  14. 14.

    A total of 86 cases according to the Bouderbala commission (report on line:

  15. 15.

    Personal interview in Tunis, September 2016.

  16. 16.

    Personal interview in Tunis, September 2017.