Weaving the threads of international criminal justice : The double dialogicity of law and politics in the ICC al-Mahdi case

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Weaving the threads of international criminal justice : The double dialogicity of law and politics in the ICC al-Mahdi case © 2021 The Authors. Published by Elsevier Ltd. Published version D’hondt, Sigurd; Dupret, Baudouin; Bens, Jonas D’hondt, S., Dupret, B., & Bens, J. (2021). Weaving the threads of international criminal justice : The double dialogicity of law and politics in the ICC al-Mahdi case. Discourse, Context and Media, 44, Article 100545. https://doi.org/10.1016/j.dcm.2021.100545 2021 Weaving the threads of international criminal justice: The double dialogicity of law and politics in the ICC al-Mahdi case Sigurd D’hondt a, ⇑ , Baudouin Dupret b , Jonas Bens c a University of Jyväskylä, Finland b CNRS/IEP Bordeaux, France c Freie Universität Berlin, Germany article info Article history: Received 11 March 2021 Received in revised form 2 September 2021 Accepted 10 September 2021 Keywords: Dialogicity Dialogical network International Criminal Court Timbuktu Cultural heritage abstract In this paper, we examine the international criminal trial of Ahmad al-Faqi al-Mahdi, a Malian Islamist who appeared before the International Criminal Court (ICC) in The Hague, charged with the destruction of Islamic shrines during the 2012 jihadist occupation of Timbuktu. Our objective is to analyze the al- Mahdi case as a dialogical network (the destructions occurred in the context of an asynchronous, translocal, press-mediated exchange between jihadists and the international community) and as an event unfolding at a dialogical site (when the commander responsible for the destructions was referred to the ICC four years later). These two dialogical orders exist largely independent of each other but are at crucial points also partly entangled. We conclude by pointing out the relevance of this ‘doubly dialogical’ approach to the broader field of sociolegal studies of international criminal justice. Ó2021 The Authors. Published by Elsevier Ltd. This is an openaccessarticle under the CC BY license (http:// creativecommons.org/licenses/by/4.0/). 1. Introduction Durkheim’s ‘objectivity of social facts’ (1982 [1895]) is, from a praxeological perspective, the outcome of a ‘thematization,’ a complex of situated practices through which the members of a site make a fact recognizable (Garfinkel, 2002). Such thematizations proceed in a constrained fashion, according to the context, purposes, and circumstances of the ongoing course of action (which, together, form its ‘grammar’). This paper argues that thematizations are also dialogical and that the objectivity of social facts is dialogically achieved. Dialogicity, as we understand it, comprises the dialogical texture of a social phenomenon. It is a constitutive feature of social order that connects people, institutions, discourses, times, and places in multiple threads of relevance, and it occurs in accordance with the purposes of the activity in/for which these threads are woven. Dialogue and dialogicity should be understood in the ‘flat’ sense of actions and utterances realized in a sequential and interrelated way, as interacting people address specific topics that are interactionally relevant to them. We use the term ‘thread of relevance’ to refer to the dialogical sequences through which these themes are addressed and treated. These dialogues and threads of relevance are accomplished through ethnomethods that are always specific to the particular social order being established. Dialogicity has been explored most systematically in conversation analysis. For Sacks (1992), a conversation is a sequential and contextualized distribution of interconnected utterances oriented to a set of common relevancies. Order is created through a succession of speech turns, by members who intervene based on what was said in the previous turn and who prospectively establish a basis for the next turn. Institutional sites, too, can be inherently dialogical. Trials and parliamentary deliberations, for example, proceed through successive speech acts, organized and allocated in a temporally ordered and sequentially organized manner. Dialogicity can also extend across sites, as there exist forms of order that do not require the participants to be copresent for their accomplishment. One example of such a distributed dialogical order is media networks mediated by press coverage, in which members interact despite their distance in time and space (Leudar and Nekvapil, 2004, and the various contributions to this issue). In this paper, we examine the international criminal trial of Ahmad al-Faqi al-Mahdi, a Malian Islamist who appeared before the International Criminal Court (ICC) in The Hague on the charge of having directed and participated in the destruction of ancient Sufi religious shrines during the 2012 jihadist occupation of Timbuktu. First, we show that the al-Mahdi case started off as a series of violent acts (the 2012 destructions) that formed part of a https://doi.org/10.1016/j.dcm.2021.100545 2211-6958/Ó2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). ⇑ Corresponding author at: Department of Language and Communication Studies, PO Box 35, FI-40014 University of Jyväskylä, Finland. E-mail address: [email protected] (S. D’hondt). Discourse, Context & Media 44 (2021) 100545 Contents lists available at ScienceDirect Discourse, Context & Media journal homepage: www.elsevier.com/locate/dcm ‘dialogical network’ (Leudar and Nekvapil, 2004): an intensely mediatized, asynchronous, translocal exchange that pitted the jihadist militia against representatives of the Malian government and various international bodies together representing the ‘international community.’ It is through these mediatized exchanges that the demolitions acquired the specific meaning of ‘destruction of cultural heritage.’ Four years later, in August-September 2016, al-Mahdi, the head of the Morality Brigade responsible for the destructions, was tried before the ICC. On this occasion, the dialogical network transformed into an event at a dialogical site, in this case, an international tribunal entrusted with the task of adjudicating war crimes and crimes against humanity. Interactions that take place at a dialogical site typically lack the open-ended quality that characterizes press-mediated dialogical networks. Instead, they are tightly constrained by their institutional embedding and their connection to a particular physical precinct (Dupret and Ferrié, 2008). Thus, a dialogical site like the ICC associates parties and personnel through successive procedural steps that are interactionally and sequentially organized (but across multiple events), where each step involves the production of situationally relevant characterizations that ascribe a particular legal status to a particular selection of events-qua-facts according to situationally applicable rules. The two dialogical orders that we are concerned with in this paper, network and site, thus exist largely independent of each other, but at crucial points they are also partly entangled. Events at a dialogical site, like a trial before the ICC, may become part of a network that extends beyond the boundaries of the judicial site and that also involves non-judicial actors. This happens, for example, when translocal dialogical connections are made relevant through out-of-court, third-party reporting. In the al-Mahdi case, however, networking extended in the opposite direction, from network to site. As we will show, the crimes with which al-Mahdi was charged at the dialogical site (the destruction of the shrines) were originally designed as a move in, and hence derived their meaning from, a mediatized, translocal dialogue, that is, from a dialogical network. 1 Two additional remarks are in order. First, dialogicity does not mean that speech turns necessarily converge, orient to the same viewpoints, and/or express agreement. Turns produced by different speakers routinely address different audiences and may therefore also activate contrasting relevancies. By audience, we mean the public that the speaker is addressing; these publics can be both real (e.g., the other participants in the hearing) or virtual (e.g., the hypostatic addressee of a discourse, like ‘the people’ or ‘the nation’). Relevance refers to the discursive repertoires with which the speakers claim to align and which they purportedly adopt; these relevancies possess no autonomous prior existence of their own and are always activated, actualized, and reinvented in the talk at hand. Still, in spite of the polyphony of audiences and relevancies that characterize mediated exchanges like the one analyzed here, the different contributions out of which they are composed still participate in the same dialogical order: they are subject to similar constraints and are headed toward identical practical purposes. Dialogicity is above all a question of procedure, not of content. Second, dialogicity is a members’ phenomenon. Dialogical connections are always occasioned by members in the context of a particular course of action. In their subsequent takes and retakes, utterances and reactions, questions and answers, they weave together the discursive threads of an object (e.g., heritage, war, civilization, Islam, jihad) for the practical purposes of the task at hand (e.g., legislation, a court ruling, media coverage). In this sense, our approach is radically disjunctive with linguistic anthropologists’ attempts to map out ‘speech chains’ (Agha, 2003) or ‘text trajectories’ (Blommaert, 2005) since we do not trace the subsequent reinscriptions of a text as they are de- and recontextualized (Bauman and Briggs, 1990) across settings and contexts. Reconstructing such trajectories might well benefit from inquiries into how dialogicity is produced (cf. Hodges, 2010, specifically in relation to global textual flows of the kind investigated here), but they privilege a ‘helicopter’s perspective’ on traffic flow, casting aside that of the drivers whose conduct constitutes the flow (Livingstone, 1987). This distinction is not always clearly drawn, and elsewhere we have argued that the notion of dialogical network itself at times retains traces of an ‘overhanging,’ ironic perspective on social order at the expense of members’ practices through which dialogicity is threaded (Dupret and Ferrié, 2015). In this paper, we proceed in two steps. Section two examines the mutually constitutive dialogicity of the network pairing the jihadi Islamists responsible for the destructions and representatives of international bodies like UNESCO and the ICC. Section three elucidates how this broader dialogical network, which started in 2012, became entwined with the dialogical site of the 2016 ICC trial. We map out how these parallel dialogicities (network and site) are mutually constitutive and how the resulting ‘double dialogicity’ impacts the shaping of judicial activities. The conclusion outlines the relevance of this approach for the sociolegal study of international criminal justice. 2. The dialogicity of destroying something: An assemblage of statements, violent acts, and video footage that went viral Three decades ago, Appadurai (1990) observed that the global availability of infrastructure for producing/transmitting information drastically reconfigured local landscapes of image and discourse. Since then, mobile phones and the internet have taken this democratization of image production to an unprecedented level (Schankweiler, Straub, and Wendl, 2018), and the 2012 events in Timbuktu illustrate the conflict-constitutive potential this harbors. Locally produced footage of the tomb destructions played a central role in a series of dialogically linked exchanges connecting actors and audiences across the globe. According to Allais (2017: 67), much of this footage had been recorded by ‘‘a Mauritanian videographer [...] who was embedded with Ansar Dine for months. [The videographer] periodically broadcast them on Saharamedia [...] Eventually they were acquired by producers and reporters from France 2 and broadcast in Envoyé Spécial on January 31, 2013, as part of the report Mali: La vie sous le régime islamiste [...].” The Ansar Dine leadership let this footage of the destructions circulate freely on the internet and social media and did not try to keep its iconoclastic campaign a secret from the outside world. On at least one occasion, they actively invited a television crew (cf. infra), which demonstrates their awareness of the shock value of such imagery and testifies to the performative character of the violence they engaged in. As such, the Timbuktu destructions inscribed themselves in an emergent tradition of jihadist iconoclasticism, which started with the demolition of the Bamyan Buddhas by the Taliban (2001) and the destructions of pre- 1 We are fully aware that the dialogical analysis proposed here can never capture the full totality of the material, embodied, practical, and situated dimensions of the destructions and the way they were experienced by the people on the ground. The tomb destructions were many things at the same time, and they meant different things to different people. At one level, they formed part of a jihadist attempt to restructure Timbuktu’s urban landscape in accordance with the tenets of Sharia, which had started much earlier and also involved the destruction of secular monuments and banning human faces from commercial advertisements (McLaughlin, 2015). The fact that the destructions at one point acquired the additional meaning of targeting ‘cultural heritage,’ however, was largely due to their entanglement in the translocal exchange with the international community reported here. Joy (2016) makes a similar argument but emphasizes that this dialogue was also locally embedded locally in the multiple trajectories of Islam in Mali. S. D’hondt, B. Dupret and J. Bens Discourse, Context & Media 44 (2021) 100545 2 Islamic temples and idols by Islamic State in Hatra and Palmyra (2015). In Timbuktu, the destructions derived their performative character in large part from the fact that they were perceived as dialogically connected to the UNESCO World Heritage Committee’s (WHC) decision to put the northern Malian sites on the List of World Heritage in Danger, announced just a few days prior to the onset of the destructions (on June 28, 2012). This dialogical connection was formulated already in the first report of the destructions that reached the outside world, a press statement released by the Agence France-Presse (AFP) news agency on Saturday, June 30, 2012, which was shortly thereafter republished on the websites of various French and international newspapers (around or immediately after 1 pm CET): [introductory paragraph omitted] ‘‘Ansar Dine is today going to destroy every mausoleum in the city. All mausoleums without exception,” Ansar Dine spokesperson for Timbuktu Sanda Ould Boumama declared assisted by an interpreter, contacted by AFP from Bamako and interrogated about the destruction of mausoleums of Muslim saints in the city that started on Saturday morning. [description of the Timbuktu site omitted] The Ansar Dine spokesperson had shortly before directly expressed himself to AFP on this destruction, suggesting in broken French that it is a retaliation against Unesco’s decision, announced on Thursday, to place Timbuktu, a city that belongs to the world heritage of humanity, on the list of heritage in danger. ‘‘God, he is unique. All this is ‘haram’ (forbidden in Islam). We are Muslims. What is Unesco?,” he said, adding that Ansar Dine was reacting ‘‘in the name of God.” [...] 2 The two final paragraphs, and the indirect and direct quotes of Ansar Dine spokesperson Sanda Ould Boumama they contain, explicitly formulate a sequential connection between what would otherwise merely be temporally contiguous events (the UNESCO decision and the tomb destructions). Observe that the AFP reporter is not so much ‘mediating’ a dialogue between UNESCO and Ansar Dine (e.g., by inviting the latter to comment on the WHC decision) but is reporting on the dialogical nature of the destructions and portraying them as a turn in a conversation, of which the decision formed the first part. For this, however, the AFP reporter solicits confirmation from the party responsible for the demolitions; the semiotization of this non-discursive act of violence apparently constitutes an act of interpretation that AFP cannot legitimately perform on its own authority and for which it needs external license in the form of a quote. 3 The report also includes a detailed account of how the quote was obtained and how it was mediated, which conveys a sense of the geographical and linguistic distance that AFP had to cover in reaching out to Ansar Dine: (1) a reporter called Ansar Dine on the AFP’s initiative and (2) caught their press officer Boumama unprepared; (3) the latter initially responded in ‘‘broken French” (4) until he eventually received assistance from an interpreter. 4 This sequential connection is premised upon a delicate interweaving of antagonistic audiences and relevancies. The dialogue reported by the AFP journalist evokes a complex of opposingbut-interlocking master- and counternarratives (Lynch and Bogen, 1996) and us vs. them categorizations (Leudar et al., 2004), through which Ansar Dine’s response to the UNESCO decision selects its audience and threads its relevance. To elucidate this interweaving, let us take a look at the text of the original UNESCO announcement, issued on June 28, 2012: The World Heritage Committee on Thursday accepted the request of the government of Mali to place Timbuktu and the Tomb of Askia on UNESCO’s List of World Heritage in Danger. The decision aims to raise cooperation and support for the sites threatened by the armed conflict in the region. The Committee, meeting in St Petersburg until 6 July, also asked Mali’s neighbours to do all in their power to prevent the trafficking in cultural objects from these sites. There is concern that such objects, notably important ancient manuscripts, be looted and smuggled abroad by unscrupulous dealers. The 21 members of the World Heritage Committee urged the African Union and the international community to do all in their power to help protect Timbuktu, inscribed on UNESCO’s World Heritage in 1988, and the Tomb of Askia, inscribed in 2004. [three final paragraphs omitted] 5 This statement powerfully evokes the master narrative that renders Ansar Dine’s destructive efforts meaningful. It formulates a range of addressees, all state-actors, bound by a shared responsibility to protect Malian heritage sites. Mali itself is singled out as the party that took the case to the WHC, to whom the committee is responding and on whose behalf it is acting. The appeal to an interstate framework and associated obligation to protect inscribes the statement in a normative order that entails a particular way of ‘reading the past’ (Leudar and Nekvapil, 2011), one that is ‘‘self-avowedly historical and global” and embraces ‘‘a genealogical method which, based on pluralistic universalism, values anything that relates the present to its historical roots” (Dupret and Gutron, 2021: 484). The statement refrains from identifying a party renouncing this obligation, but the mere positing of a threat evokes an outsider whose conduct endangers the continued existence of the monuments. Ansar Dine’s response is carefully but inversely coordinated with the framework of relevancies and constitutive-normative order established by UNESCO, in a way that closely resembles the networked categorization practices found in public statements by George W. Bush, Tony Blair, and Osama Bin Laden in the aftermath of 9/11 (Leudar et al., 2004). First, Ansar Dine militants enthusiastically took up the outsider role by acting out that which UNESCO warned against (destroying the monuments). In the ensuing telephone interview with AFP, Boumama formulated an alternative us vs. them opposition, orienting to a competing normative order that justified the destructions on religious grounds (‘‘we are Muslims,” and therefore, ‘‘all that is haram”). This alternative category pair curiously mirrors the insider–outsider framework projected by the UNESCO statement, and the two category pairs are ‘‘coordinated through incumbency” (Leudar et al., 2004: 262), with the religious ‘us’ advocated by Boumama occupying the slot of UNESCO’s outsider-category and vice-versa. The normative order formulated here is connected to a counter-narrative grounded in a ‘‘fundamentalist” method of reading the past, one that is ‘‘self-avowedly ahistorical and specific: it specifically relates the Muslim community to a founding past and denunciates any 2 20 Minutes, ‘‘Nord du Mali: ‘Tous les mausolées’ de Tombouctou seront détruits.” (https://www.20 min.fr/monde/963457–20120630-nord-mali-tous-mausolees- tombouctou-detruits, accessed March 20, 2020). All translations from French are ours. 3 The sequential connection with the UNESCO announcement was probably obvious to any competent observer (because of temporal contiguity and the contrasting but interlocking relevancies sketched below). Hence, the report suggests that AFP actively solicited a response from Ansar Dine, and Boumama’s denunciation ‘‘What is UNESCO” (in French L’Unesco, c’est quoi) comes as an afterthought to an extended turn containing a religious justification, suggesting that the journalist had confronted Boumama with the tombs’ world heritage status just before. Nevertheless, a license from the mouth of one of the perpetrators is still deemed to be required. 4 An AP report issued one day later also relied on an external license for attributing dialogicity but put the formulation of the dialogical connection in the mouths of a local journalist and a local politician. 5 UNESCO, ‘‘Heritage sites in northern Mali placed on List of World Heritage in Danger” (https://whc.unesco.org/en/news/893, accessed March 20, 2020). S. D’hondt, B. Dupret and J. Bens Discourse, Context & Media 44 (2021) 100545 3 deviance from its original purity” (Dupret and Gutron, 2021: 484). Unlike the master narrative espoused by UNESCO, it renunciates any global pretensions, or more precisely, it claims a different imaginary of the global and offers an Islamic version of such universalism (see also Joy, 2016: 68/69). Ansar Dine sources indicated, furthermore, that the militia’s leadership was not only reacting against the framework of categorial relevancies established by the UNESCO decision but also expressing its discontent with the decision’s projected audience. The next fragment is taken from a report published on the France24 website ten days after the destructions started. It states that the jihadists were not just seeking global attention but also responding to the worldwide concern for the fate of Mali’s heritage sites: Pernicious mediatization The Ansar Dine fighters are seeking to publicize their actions as much as possible, and on Tuesday they summoned a camera from the Qatari Al-Jazeera network, according to AFP. The dissemination of the images at the international level [...] in turn serves as a pretext for the destruction of the mausoleums. One of the Islamists’ spokespersons, contacted by Serge Daniel, declared: ‘‘We have heard non-Muslims talking on television about the cultural heritage of Timbuktu. They don’t have the right to do so, and this is our answer.” [...] 6 The Ansar Dine source invokes the same religiously grounded category pair as Boumama. Unnamed others had been making public statements about Timbuktu’s cultural heritage. The negative identification ‘‘non-Muslims” suggests that only Muslims are entitled to do so, and this category-based distribution of entitlements transforms the UNESCO decision into a transgression that required a response from Ansar Dine. In addition, the quote explicitly singles out the mode in which the Ansar Dine leadership took notice of the decision (‘‘we have heard non-Muslims talking on television”). At one level, this ‘self-centered,’ experiential way of referring to the decision can be heard as an obstinate refusal to recognize UNESCO’s authority. It also highlights, however, that these transgressive statements were part of a dialogue from which Ansar Dine was ostensibly excluded. In spite of its ‘ownership’ of the sites (both actual, because of the military occupation, and symbolic, as Muslims), Ansar Dine only featured as a ‘figure’ in the UNESCO discourse (Goffman, 1981), rather than being recognized as a ratified, fullfledged participant. Ansar Dine’s acting out (and doing exactly what UNESCO warned against) may therefore be understood as a way to counter the exclusion from the UNESCO inter-state participation framework, and as a means to secure itself a forum for presenting its counter-narrative to the outside world. In this sense, mediated network-like dialogicity is definitely more complex than the standard CA notion of ‘sequence’ (i.e., a set of mutually constitutive paired actions by two speakers who are already on speaking terms), as the need to secure access may exert a dramatic influende on the shaping of the network’s consecutive ‘turns.’ This exclusion of Ansar Dine as an addressed audience continued while the destructions were underway. Take the following statement by a Malian government representative, quoted in an article published on the website of Radio France Internationale (RFI) in the early evening of Saturday, June 30: The Malian government spokesperson strongly condemned the destruction of the mausoleums. [He] specifies that the perpetrators of these acts expose themselves to prosecution at the national and international level: ‘‘The government denounces this obscurantist practice. And we have already decided to take the case to the International Criminal Court. We want to tell the world that we are dealing with terrorists who know neither faith nor law. [...]” Bamako also denounces ‘‘a destructive furor comparable to war crimes.” 7 The government representative’s quote evokes the same normative order and categorial/participation framework as the UNESCO statement. It frames the message as directed to ‘‘the world” and includes a list of ‘‘tuning up predicates” (Leudar et al., 2004: 251) that add flesh to the outsider category: Ansar Dine are ‘‘terrorists who know neither faith nor law”; they engage in an ‘‘obscurantist practice”; and they are driven by ‘‘a destructive furor”. 8 While media reports typically frame such third-party denunciations as dialogically connected to the destructions, the statements themselves systematically ignore the dialogical (retaliatory) character of Ansar Dine’s actions, neither do they recognize the jihadists as a potential partner for dialogue. In fact, the only third party respondent that explicitly addressed the jihadists is ICC Prosecutor Fatou Bensouda, who on Monday, July 2, delivered a statement to AFP, explicitly warning them that the destructions constituted a war crime under the Rome Statute and that her office would open an investigation. She, too, refused to ratify the destructions’ dialogicity, but her ‘addressive’ statement is the only one which elicited an answer from Ansar Dine (or that the mediating agencies considered worth the effort of soliciting such an answer): In radio and television interviews from Senegal, the newly appointed chief prosecutor of the International Criminal Court, Fatou Bensouda, warned the rebels that destruction of religious and cultural heritage could lead to war crimes charges. ‘‘The only tribunal we recognize is the divine court of Sharia,” the Associated Press quoted Ansar Dine spokesmen Oumar Ould Hamaha as saying in response to Bensouda’s warning. The AP said Hamaha justified the destruction as a divine order to pull down idolatrous constructions ‘‘so that future generations don’t get confused, and start venerating the saints as if they are God.” 9 3. The al-Mahdi trial as a branch of the network: Truncating facts, audiences, and relevancies for practical legal purposes As indicated, the ICC dialogical site associates parties and participants through a succession of procedural steps distributed across multiple hearings. The integrity of this site is guaranteed not only by its connection to a particular physical location (though its role cannot be underestimated, see D’hondt, 2021a) but also by the participants’ joint orientations to a set of shared relevancies. One set of such relevancies is provided by the constitutive order of the trial. Through their shared orientations to this constitutive order, the parties situate the trial within the framework of international criminal law and the Rome Statute, the ICC’s foundational treaty that defines the competence and jurisdiction of the Court. This constitutive order determines the teleology of the trial and determines the sequence and the pace of relevant actions, opening up affordances while simultaneously closing down others. Part of this constitutive order is procedural, comprising a set of institu- 6 France24, ‘‘Ansar Dine détruit les mausolées de la grande mosquée de Tombouctou” (https://www.france24.com/fr/20120710-mali-ansar-dine-attaque-grande-mos- quee-djingareyber-tombouctou-mausolees, accessed March 20, 2020). 7 RFI, ‘‘Mali: la destruction des mausolées de Tombouctou par Ansar Dine sème la consternation” (https://www.rfi.fr/fr/afrique/20120630-mali-bamako-tombouctou- islamistes-ansar-dine-mausolees-charia-france-hamadoun-toure, accessed March 20, 2020). 8 On the relevance of attributing emotions to human and non-human bodies in the al-Mahdi case, see Bens (2018). 9 LA Times, ‘‘Islamist rebels in Mali destroy Timbuktu historic sites” (https:// latimesblogs.latimes.com/world_now/2012/07/radical-islamic-rebels-in-mali- destroying-timbuktu-treasures.html, accessed November 12, 2018). S. D’hondt, B. Dupret and J. Bens Discourse, Context & Media 44 (2021) 100545 4 tional constraints to which members of the site must at all times demonstrably abide, at the risk of their action being subject to objection or appeal. Another part of it is thematic, stipulating certain repertoires that are deemed acceptable while dismissing others as irrelevant. There are yet other sets of relevancies that accountably came into play in the al-Mahdi trial, such as universalism and world heritage, which exceed the narrow legal qualifications provided by the Rome Statute (e.g., Art.8(2)(b)(ix) on attacks against ‘‘buildings dedicated to religion [...] or historical monuments”) and related sources of relevant law (such the 1907 Hague Regulation on the protection of cultural goods in case of armed conflicts, or the case law of the Yugoslavia Tribunal). In the previous section, we saw that these notions of heritage and universalism formed part of a translocal dialogical network, in which a UNESCO-centered master narrative encapsulating a universalistic conception of heritage (and its associated categorizations) clashed with jihadi and fundamentalist counter-narratives (Dupret and Gutron, 2021). Here, we examine how the al-Mahdi trial, as an event situated at the ICC dialogical site, came about as a new ‘node’ of this heritage-in-Mali network. 10 We show how ICC actors align with particular positions within the network and how, in doing so, they recognize certain forms of dialogicity while rejecting others. In this way, the analysis elucidates how international criminal justice values a very specific conception of society, history, spirituality, and memory (see also Bens, 2022). We already saw that the Malian government, on the first day of the destructions (June 30, 2012), issued a statement that threaded its relevance in a particular way: it excluded the possibility of entering into negotiations with Ansar Dine, characterizing them as terrorists who ‘‘know neither faith or law” and denouncing the destructions as ‘‘a destructive furor akin to war crimes.” This use of the legal category of war crimes created a new affordance, opening up the possibility of referring the case to the ICC. This affordance was amplified by ICC Prosecutor Fatou Bensouda’s warning, on July 2, 2012, that the destruction of buildings dedicated to religion could result in a war crime charge. Both statements can be described as dialogical connectors. They are dialogical because they function within a sequential frame of remote speech turns, and they are connectors since they trigger the intervention of the ICC. As such, they set in motion a procedure before the ICC that became a new branch of the mediatized heritage-in-Mali dialogical network. The crucial role of these dialogical connectors is confirmed in the 141-page document outlining the charges, which the ICC’s Office of the Prosecutor (OTP) submitted in December 2015. The following fragment comes from the ‘‘Legal conclusions” section, summarizing the legal principles that apply to the case. It shows how the OTP inscribes the charges in the ICC’s constitutive order (the legal framework established by the Rome Statute) by referring to Mali’s ratification of the Rome Statute (par. 235) and to the letter by which the Malian government seized the court (par. 236): 235. Mali signed the Rome Statute establishing the International Criminal Court on July 17, 1998 [...] The International Criminal Court therefore has jurisdiction for crimes under the Rome Statute committed on the territory of the State of Mali or by Malian nationals from July 1, 2002. 236. On July 13, 2012, Mali referred to the Prosecutor of the International Criminal Court the most serious crimes committed on its territory since January 2012. The referral letter sent by Mali to the Office of the Prosecutor referred in particular to ‘‘summary executions [...], the massacres of civilian populations [...], the destruction [...] of [m]ausoleums and [m] osques.” 11 The integration of this dialogical connector into the Court’s constitutive order is itself highly selective: only the government’s letter is mentioned, not the complex of exchanges (including a statement by the ICC Prosecutor herself), in which that letter dialogically inscribed itself. The dialogical connection established here sets the legal process in motion and initiates a new procedure at the ICC dialogical site. The latter articulates a multiplicity of audiences and relevancies, activating networks related to the Court’s general purpose (the implementation of international criminal law) and establishing connections with the events that justified the seizure of the court. Now that the case has been referred to the ICC, the heritage-in- Mali network transforms into a collection of data and evidence specifically oriented to the site’s practical purpose (the trial of an international criminal case). This selection operates in three directions: (1) putting in place the parties to the trial and settling their status; (2) addressing the relevant audiences according to their status in the trial; and (3) threading the legitimate legal relevancies of the case. (1) The ICC Prosecutor’s intervention of July 2, 2012, already projected a particular distribution of roles onto the dialogical network and pre-characterized relevant ‘actants’ (acting entities, including people, objects, and events, cf. Greimas and Courtès, 1979). Hence, objects were identified as ‘buildings dedicated to religion’ and ‘heritage’ (thus paving the way for a legal characterization according to the Rome Statute); events were selected as putatively constitutive of ‘facts’ (thus justifying the seizure of the ICC); and aggregates of people were designated as perpetrators and as affected by the crime, preempting their legal designation as ‘accused’ and ‘victim’ respectively. This pre-characterization determined which ‘turns’ in the dialogical network were legitimate (e.g., the statements of Malian authorities, but not Ansar Dine’s Sharia-based justifications of the destruction) and pre-allocated moral and normative value to relevant actants (tombs are sacred, destroying heritage is evil, local Timbuktu residents are victims, tomb destructors are criminals). As such, these connective triggering statements represent the first steps of what Latour (2009) calls the ‘passage to law.’ This passage to law truncates the intrinsic dialogicity of the heritage-in-Mali network for the practical purposes of its own legal rephrasing. This process transforms the network into a plentiful, fact-finding trove, with the ‘facts’ for which al-Mahdi is tried miraculously corresponding to the legal categories and provisions of the Rome Statute. This is exemplified by the mirroring structure of the OTP’s charging document discussed earlier, of which the first section reviews ‘‘Factual elements” and the second ‘‘Legal conclusions.” The heritage-in-Mali network provides the input for this process of legal characterization, and the trial procedure at the ICC dialogical site hence comes to incorporate elements from the network. Importantly, this legal characterization also erases the dialogical character of the destructions documented in section two. Take the following account of the events leading up to the destructions, taken from the Legal conclusions section (where the factual elements detected by the OTP in the fact-finding trove have already been converted into hard legal currency): 10 In line with the critique of the reification of the notion of a dialogical network formulated in the introduction, the term ‘heritage-in-Mali network’ should be interpreted as a shorthand for the dialogical potential of the mediated interactions analyzed in the previous section, rather than as an actual corpus of networked statements. 11 ‘‘Document présentant les conclusions factuelles et juridiques du Bureau du Procureur au soutien du Chef d’accusation dans l’affaire contre Ahmad AL FAQI AL MAHDI,” (December 17, 2015), https://www.legal-tools.org/doc/2ac52a/pdf/. S. D’hondt, B. Dupret and J. Bens Discourse, Context & Media 44 (2021) 100545 5 4.2. Context of the attack carried out in Timbuktu in June and July 2012 against historical monuments and buildings dedicated to religion [...] 90 Buildings dedicated to religion and historical monuments other than those referred to in these writings were also attacked in Timbuktu in 2012. 91. A first wave of attacks occurred in April/May 2012. It resulted in acts of degradation and partial destruction, such as the ripping out of the doors of certain mausoleums. The affected buildings included: the Al Farouk monument, desecrated at the end of April/beginning of May 2012; the Sidi Mahmoud mausoleum, desecrated around May 4, 2012; the Cheick Mohamed Mahmoud Al Arawani mausoleum, also desecrated around May 4, 2012; and finally the Martyrs’ Monument, desecrated around May 23, 2012. 92. This first attack prompted the government of Mali and UNESCO to meet on May 24 to ensure better protection of the cultural heritage located in Timbuktu and in northern Mali (note 368: UNESCO statement). As of June 28, 2012, the city of Timbuktu was inscribed by UNESCO on the list of world heritage in danger (note 369: partly blackened; reference to the UNESCO lists). 93. A second attack on historic buildings and monuments dedicated to religion took place between approximately June 30, 2012 and approximately July 11, 2012 (note 370: blackened). [...] Paragraphs 91 and 92 report a series of relatively minor incidents that happened shortly after Ansar Dine took control of Timbuktu, including the desecration of the secular, post-independence Al Farouk and Martyrs’ monuments. These events set in motion an interchange between UNESCO and the Malian government (with a meeting on May 24, 2012), which eventually resulted in the WHC’s June 28 decision to place the Malian sites on the endangered heritage list. The initial dialogue prompting the UNESCO decision is thus extensively documented, but the report ignores subsequent dialogical connections between that decision and the tomb demolitions, in spite of their temporal contiguity (the demolitions started only two days later) and the fact that such connections had been amply commented on in the local and international press. Instead, a new paragraph begins immediately after the UNESCO decision is introduced. The charging document also contains quotations from Ansar Dine spokesperson Boumama to the international press (although not from the AFP interview discussed earlier) but, again, without acknowledging their dialogically networked character (and the role they played in semiotizing/networking the destructions) (par. 129–131). (2) Once this ‘passage to law’ has been initiated, the actants for the legal process are designated, and the heritage-in-Mali network is truncated accordingly, the dialogue in the courtroom can start, which involves trial actors taking turns according to an institutionally constrained dialogical format. This does not mean, however, that trial actors exclusively address each other. They also reach out to out-of-the-courtroom audiences, although in terms, roles, and identities that have been legally re-specified. In the al-Mahdi trial, at least three such audiences are particularly relevant: the victims, the international community, and the jihadi nebula, of which the last two also featured in the dialogical network. This entangling of internal and external audiences is most obvious in the apology that al-Mahdi made on the first trial day, immediately after the charges had been read out to him (see also D’hondt, 2021b). The apology is interesting for various reasons. Trial parties’ shared orientation to the trial’s constitutive order does not determine conduct in a purely mechanical fashion. They orient to applicable procedural constraints, legal relevancies, and the trial’s overall teleological (verdict-oriented) nature, but they always do so creatively. The body of laws and principles opened up by the passage to law is ‘‘putatively shared,” but the resulting accounts and interpretations are often ‘‘radically disjunctive” (Jayyusi, 2015: 274). Agreement on applicable terms and categories does not necessarily imply a consensus on the interpretation of these terms or on who qualifies as the right incumbent (Kertzer, 1988). In the al-Mahdi case, however, there existed such a consensus. Before the trial started, the defense had indicated al-Mahdi’s wish to plead guilty and had negotiated a settlement with the OTP, accepting the latter’s presentation of the facts and of applicable rules. 12 This strategic positioning on the trial chessboard dialogically materialized in the apology, where al-Mahdi assumed a range of alignments (and associated reciprocities) that looked completely different than if he had rejected the charges. In the apology (in Arabic, simultaneously interpreted into English and French), al-Mahdi addresses audiences in and outside the courtroom: [7 lines omitted] Ladies and gentlemen, it is with deep regret and with great pain I had to enter a guilty plea and all the charges brought against me are accurate and correct. I am really sorry. I am really remorseful and I regret all the damage that my actions have caused. I regret what I have caused to my family, my community in Timbuktu, what I have caused my home nation, Mali, and I’m really remorseful about what I had caused the international community as a whole. My regret is directly -- or, is directed particularly to the generations, the ancestors of the holders [sic, the French interpretation mentions ‘‘descendants of the Saints”] of the mausoleums that I have destroyed. I would like to seek their pardon, I would like to seek the pardon of the whole people of Timbuktu, I would like to make them a solemn promise that this was the first and the last wrongful act I will ever commit. I seek their forgiveness and I would like them to look at me as a son that has lost his way and consider me part of the social fabric of Timbuktu and must not forget what I have contributed in the past to Timbuktu. It is my hope that in accordance with the noble Islamic principles be able [sic] to forgive me and to accept my regret [...]. 13 al-Mahdi first addresses the Court, demonstrating awareness of the specific dialogicity characteristic of courtroom talk: he responds to charges brought against him by the prosecution but addresses this response to the Court, not the accuser (cf. Komter, 1994). Next, he lists those for whom he feels remorse: his family, the Timbuktu community, the Malian nation, and the international community. At the moment of the apology, ICC Prosecutor Fatou Bensouda had not yet presented her own submission, but apart from ‘‘my family,” al-Mahdi’s list perfectly mirrors the three abstract victim constituencies that Bensouda had evoked six months earlier at the March 1 Confirmation of Charges hearing (see D’hondt, 2019). None of the members of the list participated in the trial. In the second paragraph, al-Mahdi returns to the dialogical site, expressly directing his remorse to the so-called ‘victim-participants’ recognized by the Court and represented at the trial (nine Timbuktu residents, members of families considered the guardians of the tombs), anticipating an equally conciliatory response on their part. Finally, just 12 The settlement included a list of mutually agreed upon facts and an agreement on a sentencing range (9 to 11 years) within which the other party would not appeal the verdict. This arrangement, however, was not binding to the Chamber. 13 Transcript ICC-01/12–01/15-T-4-Red-ENG WT 22–08-2016 (https://www.legaltools.org/doc/14e68e/), p.8, lines 11–25. S. D’hondt, B. Dupret and J. Bens Discourse, Context & Media 44 (2021) 100545 6 before wrapping up, al-Mahdi again turns to an external audience with a warning to fellow Muslims not to commit the same mistake: In conclusion, I would like to make -- to give a piece of advice to all Muslims in the world not to get involved in the same acts I got involved in because they are not going to lead to any good for humanity. 14 Here, al-Mahdi aligns with the image the ICC presents of itself as a justice institution that not only seeks restorative justice but also possesses a deterrence value (‘‘The shadow the Court casts ahead,” in the words of former ICC prosecutor Moreno Ocampo). 15 (3) Trial parties thus also speak to audiences external to the ICC dialogical site and to actants in the network, including the Malian nation, the international community, and other intangible, abstract entities. They establish relationships to these audiences and actants through the threading of relevancies. The latter involves the identification and delineation of relevant themes, such as the Rome Statute and the relevant body of international criminal law, the UNESCO and the heritage discourse it promotes, notions of universalism, and so forth. It also entails the exclusion of alternative discourses, such as fundamentalist understandings of heritage. Through this threading of relevancies, the parties to the trial, collaboratively but not symmetrically, specify the limits of acceptable arguments, select and interpret the sources for such arguments, and establish the basis of the future uses of these arguments. This process is at once pro- and retrospective, and it represents a site-specific instance of what Garfinkel (1967) calls the ‘documentary method of interpretation.’ One example of this is the OTP’s fleshing out of the meaning of what Art. 8(2)(e)(iv) of the Rome Statute refers to as ‘‘buildings dedicated to religion” (which takes several pages in the charging document, par. 265 to 272), which paves the way for its subsequent characterization of Ansar Dine’s destruction of the shrines as ‘‘an attack on buildings dedicated to religion.” In essence, this amounts to a double-bind process: facts are processed so as to fit legal categories, while legal categories are interpreted in a way that makes them relevant to the processed facts (what Hacking (1995) refers to as a ‘looping effect’). If the Trial Chamber confirms the OTP’s interpretation of this legal category in the verdict, this can, in turn, serve as a basis for characterizing similar acts elsewhere as ‘‘attacks on buildings dedicated to religion.” This does not happen automatically, and the elevation of a ruling to the status of precedent requires a subsequent ruling that acknowledges the prior decision as the basis for the new decision (see also Mertz, 2007; D’hondt et al., 2021). Occasionally, this logic of precedent, too, is mediated by a dialogue with third parties that are not members of the site. This is exemplified in the following excerpt, which comes from the OTP charging document in the al-Hassan case, the second trial related to the 2012 jihadist occupation of Timbuktu, which started in July 2020: 688. The term ‘‘attack” in article 8–2-e-iv has an autonomous, distinct meaning from the more frequent use of the term ‘‘attack” in article 8 (note 1719 Al Mahdi Judgment, & 12, 14– 16): it does not oblige the author to have directed ‘‘acts of violence against the adversary, whether these acts are offensive or defensive” (note 1720: additional protocol of 1949 Geneva Convention...), in other words to have acted in the ‘‘conduct of hostilities.” On the contrary, for the purposes of article 8–2-e-iv, it is sufficient that the author directs any violent act against the protected property, regardless of which belligerent party exercised control over the object to the time of the facts (note 1721: Contra W. Schabas, ‘Al Mahdi has been convicted of a crime he did not commit,’ [reference...], particularly pp. 76– 77, 83). It is therefore a war crime to direct acts of violence against cultural property even when it is already in the possession of the party to the conflict to which an accused is affiliated. 16 In this excerpt, the Prosecution argues that the notion of an ‘‘attack directed against buildings dedicated to religion” does not require that that the attack take place in the context of military hostilities for the war crimes charge to be applicable. Footnote 1719 refers to the al-Mahdi ruling to buttress the validity of this interpretation (although the ruling did nowhere specifically elaborate the meaning of ‘‘attack”), while footnote 1721 frames this line of reasoning as a rebuttal of the position endorsed by William Schabas. In a criticism of the al-Mahdi judgment provocatively titled ‘‘Al Mahdi has been convicted of a crime he did not commit” (Schabas, 2017), Schabas argues that a persecution charge (i.e., a crime against humanity) would have been more applicable for precisely this reason. Schabas is a reputed international criminal law scholar and a figure of authority in the field, but not a member of the ICC dialogical site. 4. Conclusion In this article, we have argued that international criminal justice procedures are characterized by a double dialogicity. An international criminal trial is an event at a particular dialogical site but is also embedded in a broader dialogical network. Hence, the 2016 al-Mahdi trial proceeded through successive speech acts allocated in a temporally and sequentially ordered manner and oriented toward common relevancies. In a broader way, however, the trial was also part of a network of dialogically threaded exchanges mediated by international press coverage, of which the Timbuktu 2012 mausoleum destructions formed the center. Analyzing this double dialogicity of international criminal trials opens up fresh perspectives on many longstanding sociolegal debates on international criminal justice, the most important probably being that of how law relates to the political. Since their inception, international criminal trials have been criticized for promoting a ‘spectacularization’ of justice (Lynch and Bogen, 1996; specifically in relation to the ICC in Africa, see Clarke, 2009), which transforms international criminal proceedings into political trials (Kirchheimer, 1961; Arendt, 1963; Shklar, 1964). From a performative angle, such spectacularization is of course unavoidable, and it may therefore be more fruitful to investigate instead exactly how these trials are political. Here, the double dialogicity of international criminal proceedings provides a useful point of departure. The 2012 destructions revolved around the articulation and denial of particular dialogicities, which involved making a selection between competing relevancies and audiences. The 2016 trial resumed certain of these dialogues but closed down others, acknowledging the dialogicity of certain actors’ conduct while refusing to do so for others. These dialogues were of a political nature, and in acknowledging and closing down dialogicities the ICC was thus navigating a political landscape. Our analysis, then, seeks to demonstrate exactly how it navigated this political landscape. 17 14 Ibid., p.9, lines 16–18. 15 Two days later, on August 23, the legal representative of the nine victimparticipants (the guardians of the tombs) refused to accept the apology, arguing that it was calculated and insincere (D’hondt, 2021b: 72; on the role this refusal had for a rhetoric of sentimentalizing, see Bens 2018: 81). Those parts of the apology directed to court-external audiences, however, remained unanswered within the dialogical site of the trial but became the object of further networking: in their coverage of the trial, the media regularly included reports of how the apology was received by Timbuktu residents, representatives of civil society, and government spokespersons. 16 ‘‘Version amendée et corrigée du Document contenant les charges contre M. Al HASSAN Ag ABDOUL AZIZ Ag Mohamed Ag Mahmoud” (May 11, 2019), https://www. legal-tools.org/doc/1e4aac/. 17 For an alternative account of the ICC as inevitably political (albeit from an ironic, ‘overhanging’ position), consider Nouwen and Werner (2010). S. D’hondt, B. Dupret and J. Bens Discourse, Context & Media 44 (2021) 100545 7 In addition, the analysis sheds new light on the framing of the al-Mahdi trial as contributing to deterring future destruction of cultural heritage. At the March 2016 Confirmation of Charges Hearing, for example, ICC Prosecutor Bensouda explicitly stated, ‘‘We must protect our common heritage from the desecration, ravages and long-term effects of such destructive acts.” 18 This deterrence effect of international criminal proceedings is the subject of an enduring debate (Mullins and Rothe, 2010), with some authors arguing that trials dissuade future perpetrators (e.g., Smidt, 2001; Dancy, 2017) and others being more skeptical or even outright denying any such effect (e.g., Rodman, 2008; McAuliffe, 2010; Cronin-Furman, 2013). Elucidating the double dialogicity of international criminal trials demonstrates that deterrence is not a unilateral, unidirectional communicative act that ‘emanates’ from the ICC and ‘addresses’ potential perpetrators. In fact, our analysis reveals that the perpetrators of the tomb destructions were at least partially motivated to commit these acts precisely because they were considered an international crime, not in spite of it (see Dupret and Gutron, 2021). Destroying objects and structures that the international community in its ‘euroniversalist’ modernist frame of relevancies had designated as cultural heritage enabled Malian jihadists to engage in a dialogue on who possesses the prerogative to assess cultural and religious value. In this sense, deterrence, too, must be thought of as a highly complex dialogical process. This project of systematically exploring international criminal trials in their double dialogicity represents one way to operationalize the basic social scientific instinct that context matters for understanding the workings of law. As such, we identify with the larger endeavor of investigating the practice of international criminal law, which means, in a broad sense, that we analyze how legal actors perform the law, in a context of structural determinants made relevant by actors in and outside the courtroom (Meierhenrich, 2014). Event and structure are thus equally relevant to the study of law, and the analysis must take into account courtroom proceeding as well as the larger social forces in which they are embedded (Scheffer, 2007), insofar as they inscribe themselves into the legal performance. It is our contention that close attention to the dialogical nature of legal interactions, both in concrete proceedings and in their broader translocal structural embedding, provides an indispensable starting point for a more nuanced analysis of the often ambivalent and messy workings of international criminal justice proceedings. Funding source Academy of Finland, project number 325535; German Research Foundation, project number DFG-SFB1171-B04; A06. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. References Agha, A., 2003. The social life of cultural value. Lang. Commun. 23 (3–4), 231–273. https://doi.org/10.1016/S0271-5309(03)00012-0. Allais, L., 2017. Amplified humanity and the architectural criminal. Future Anterior 14 (1), 51–69. Appadurai, A., 1990. Disjuncture and difference in the global cultural economy. Theory, Culture & Society 7 (2–3), 295–310. Arendt, H., 1963. Eichmann in Jerusalem. Viking, New York. Bauman, R., Briggs, C.L., 1990. Poetics and performances as critical perspectives on language and social life. Annual review of Anthropology 19 (1), 59–88. Bens, J., 2018. Sentimentalising persons and things: Creating normative arrangements of bodies through courtroom talk. Journal of Legal Anthropology 21, 72–91. https://doi.org/10.3167/jla.2018.020105. Bens, J., 2022. The Sentimental Court: The Affective Life of International Criminal Justice. Cambridge University Press, Cambridge. Blommaert, J., 2005. Discourse. Routledge, London. Clarke, K.M., 2009. Fictions of justice. Cambridge University Press, Cambridge. Cronin-Furman, K., 2013. Managing expectations: International criminal trials and the prospects for deterrence of mass atrocity. Int. J. Transitional Justice 7 (3), 434–454. Dancy, G., 2017. Searching for deterrence at the International Criminal Court. Int. Criminal Law Rev. 17, 625–655. D’hondt, S., Pérez-León-Acevedo, J.-P., Barrett, E., 2021. The indeterminacy of precedent: Negotiating admissible victim participant testimony before the International Criminal Court. International Criminal Law Review 00 (00), 1–26. https://doi.org/10.1163/15718123-bja10082. Dupret, B., Ferrie, J.N., 2008. Legislating at the shopfloor level: Background knowledge and relevant context of parliamentary debates. Journal of Pragmatics 40 (5), 960–978. Dupret, B., Ferrié, J.N., 2015. The practical grammar of law and its relation to time. In: Dupret, B., Lynch, M., Berard, T. (Eds.), Law at work. Oxford University Press, Oxford, pp. 27–48. Dupret, B., Gutron, C., 2021. The concept of heritage and the grammar of Islamic fundamentalism. Memory Studies 14 (2), 483–500. https://doi.org/10.1177/ 1750698020921435. Durkheim, E., 1982 (1895). The rules of the sociological method. Free Press, New York. D’hondt, S., 2019. Humanity and its beneficiaries: Footing and stance-taking in an international criminal trial. Signs and Society 7, 427–453. https://doi.org/ 10.1086/705279. D’hondt, S., 2021a. Why being there mattered: Staged transparency at the International Criminal Court. Journal of Pragmatics 183, 168–178. https://doi. org/10.1016/j.pragma.2021.07.014. D’hondt, S., 2021b. One confession, multiple chronotopes: The authentification of an apology in an international criminal trial. Journal of Sociolinguistics 25 (1), 62– 80. https://doi.org/10.1111/josl.12447. Garfinkel, H., 1967. Studies in ethnomethodology. Polity, Cambridge. Garfinkel, H., 2002. Ethnomethodology’s program. Rowman & Littlefield, Lanham. Goffman, E., 1981. Footing. In: Forms of talk. University of Pennsylvania Press, Philadelphia, pp. 124–159. Greimas, A.J., Courtès, J., 1979. Sémiotique, Vol. 1. Hachette, Paris. Hacking, I., 1995. The looping effects of human kinds. In: Sperber, D. (Ed.), Causal cognition. Clarendon, Oxford. Hodges, A., 2010. Discursive constructions of global war and terror. In: Coupland, N. (Ed.), The handbook of language and globalization. Wiley-Blackwell, Maiden, pp. 305–322. Jayyusi, L., 2015. Discursive cartographies, moral practices. In: Dupret, B., Lynch, M., Berard, T. (Eds.), Law at work. Oxford University Press, Oxford, pp. 273–298. Joy, C., 2016. UNESCO Is what? World heritage, militant Islam and the search for a common humanity in Mali. In: Bruman, C., Berliner, D. (Eds.), World heritage on the ground. Berghahn Books, New York, pp. 60–77. Kertzer, D., 1988. Rituals, politics and power. Yale University Press, New Haven. Kirchheimer, O., 1961. Political justice. Princeton University Press, Princeton. Komter, M., 1994. Accusations and defences in courtroom interaction. Discourse & Society 5 (2), 165–187. https://doi.org/10.1177%2F0957926594005002002. Latour, B., 2009. The making of law. Polity, Cambridge. Leudar, I., Marsland, V., Nekvapil, J., 2004. On membership categorization: ‘us’, ‘them’ and ‘doing violence’ in political discourse. Discourse & Society 15 (2–3), 243–266. Leudar, I., Nekvapil, J., 2004. Media dialogical networks and political argumentation. J. Language Politics 3 (2), 247–266. Leudar, I., Nekvapil, J., 2011. Practical historians and adversaries: 9/11 revisited. Discourse & Society 22 (1), 66–85. Livingston, E., 1987. Making sense of ethnomethodology. Routledge & Kegan Paul, London. Lynch, M., Bogen, D., 1996. The spectacle of history. Duke University Press, Durham. McAuliffe, P., 2010. Suspended disbelief? The curious endurance of the deterrence rationale in international criminal law. New Zealand J. Public Int. Law 10, 227– 262. McLaughlin, F., 2015. Linguistic warscapes of northern Mali. Linguistic Landscape 1 (3), 213–242. Meierhenrich, J., 2014. The practice of international law. Law and Contemporary Problems 76 (3–4), 1–83. Mertz, E., 2007. The language of law school. Oxford University Press, Oxford. Mullins, C.W., Rothe, D.L., 2010. The ability of the International Criminal Court to deter violations of international criminal law. Int. Criminal Law Rev. 10, 771– 1768. Nouwen, S.M., Werner, W.G., 2010. Doing justice to the political: The international criminal court in Uganda and Sudan. Eur. J. Int. Law 21 (4), 941–965. Rodman, K.A., 2008. Darfur and the limits of legal deterrence. Human Rights Quarterly 30 (3), 529–560. Sacks, H., 1992. Lectures on Conversation. Blackwell, Oxford. Schabas, W., 2017. Al Mahdi has been convicted of a crime he did not commit. Case Western Reserve J. Int. Law 49, 75–102. 18 Transcript ICC-01/12–01/15-T-2-Red2-ENG (https://www.legal-tools.org/doc/ 1a7bdc/), p. 17. S. D’hondt, B. Dupret and J. Bens Discourse, Context & Media 44 (2021) 100545 8