Confined knowledge flows in transitional justice

ABSTRACT

The article contributes to literature that critically scrutinizes knowledge production and transfer in conflict and intervention contexts. Drawing on original research on the Tunisian transitional justice process, it contributes an empirically grounded picture to the study of co-production of governance orders and security knowledges through transnational assemblages. These transnational assemblages are formed by complex coalitions of actors from the Global North and South, and the socio-material context they operate in. The article shows how security knowledge is produced, channelled, and steered into confined knowledge flows as transitional justice processes unfold. It then shows the ambivalent nature and different qualities of confined knowledge flows as they may be enabling and limiting, exclusionary and protective, and implicated with power relations. By doing so, it contributes to the understanding of how the (neo-)liberal politics of transitional justice are reproduced.

KEYWORDS:

• knowledge production
• transitional justice
• governing technology
• power
• Tunisia

1. INTRODUCTION

There is an emerging focus in International Relations and Peace and Conflict Studies on knowledge production in and about conflict and intervention. Scholars critically scrutinize how and what kind of knowledge gets produced and/or transferred, by whom, and with what consequences for inter- and transnational governance orders (e.g., Bliesemann de Guevara & Kostić, 2017; Bonacker, 2022; Danielsson, 2020; Distler, 2016; Kortendiek, 2021; Martín de Almagro, 2021). Reflecting on trends in the field, they identify the marginalization (or at least a danger thereof) of knowledge from those who occupy less powerful positions in global politics (see also Anderl & Witt, 2020). Situated within this broader trend, these questions also occupy scholars working on transitional justice (e.g., Ben-Josef Hirsch, 2006; Jones, 2021; Jones & Lühe, 2021; Menzel, 2020; Rowen, 2017; Salehi, 2022a). One central interest lies in examining whose knowledge is perceived to be valid and useful, and how knowledge ‘low down on the hierarchy’ (Foucault, 1980, p. 2) is rendered inadequate.

While it is important to point out these hierarchies and marginalizations, and the above-mentioned studies do so in a careful and nuanced way, doing so might carry the danger of a dichotomous picture emerging between ‘the local’ and ‘the international’, and of knowledge production by, on, and/or for actors from the Global North and Global South (Lopez Lucia & Martín de Almagro, this issue). However, if we zoom in on concrete knowledge production processes in conflict and intervention contexts, the picture gets blurry. Various actors form complex coalitions with ideas and interests cutting across this often-assumed North–South or local–international divide. As this special issue sets out to explore ‘how governance orders and security knowledges (are) co-produced through transnational security assemblages’ (Lopez Lucia & Martín de Almagro, this issue), I aim to contribute one empirically grounded picture of such co-production. Co-production here functions ‘as a way of interpreting and accounting for complex phenomena’ and interrogating the interplay between knowledge and governance (Jasanoff, 2004, p. 3). To do so, I examine knowledge production, transmission and confinement in transitional justice. I ask how security knowledge is produced, channelled, and steered through transnational assemblages – through cooperation, but also through friction and resistance (see Jones & Lühe, 2021; Salehi, 2022a) – into confined knowledge flows as transitional justice processes unfold. The focus is therefore on the process – more so than the substance – of knowledge production, flow and confinement.

Assemblages are temporal and contingent formations that combine the social and the material, shaped by power relations and political struggles (Acuto & Curtis, 2014). They can be used as an analytical tool to help us think through ‘the circulation of modern forms’ and ‘how global elements interact with situated practices beliefs and politics’ [sic] (Sassen & Ong, 2014, p. 21). The notion of assemblage allows us to bring different pieces into a picture, look at governing technologies in different ways, and explore the relationality of things and people. There are various usages of the term assemblage and entry-points to thinking with it, drawing from different strands of scholarship, often building on the work of Deleuze and Guattari, Latour, or Foucault (Anderson & McFarlane, 2011). As a commonality, Colin McFarlane and Ben Anderson understand the added value of thinking with assemblage in terms of critique and orientation: they see the critique assemblage thinking offers in its focus on process and formation, and the orientation in an ‘ethos of engagement attuned to the possibilities of socio-spatial formations to be otherwise within various constraints and historical trajectories’ (McFarlane & Anderson, 2011, p. 162). What sets assemblage approaches apart from other processual-relational approaches is their explicit purpose of bringing ‘materialist characters’ into the analysis and ‘[m]aking the work [these materialist characters] do in global governance visible […]‘ (Leander, 2021, p. 161). As such, they are a suitable analytical tool to explore the production, flow, and confinement of knowledge as I would like to do in this article. This is done, I would argue, through people in an interplay with material aspects, such as infrastructure/the built environment.

Transnational transitional justice assemblages are ‘multi-layered, process-based, and emergent’ formations, in which different domains – for example the judicial, the political, the symbolic, and the affective – are unevenly entangled (Reading, 2019, p. 249). They are formed by a variety of actors from the Global North and South, among them transitional justice professionals, politicians, civil servants, victims and civil society representatives, and the socio-material context they operate in. These assemblages, and their power relations, concurrently characterize the production and flow of security knowledge and the ‘dominant epistemological flows of power’ (Rai, 2003, p. 60). Therefore, this disposition shapes what knowledge is deemed relevant in transitional justice, who gets to produce or access to it, and if and how it may move.

I introduce confined knowledge flows as a particular form shaping governance orders. What do I mean with confinement here? Knowledge can only flow within bounds – and sometimes not at all. For confined knowledge flows, I assume that some kind of movement (flow) is possible and desirable, but what kind of movement is restricted (confined). Some knowledge should not get out of certain settings or reach only certain audiences, only certain knowledge should be transferred or even produced. Confined knowledge flows have an ambivalent quality, as they are enabling and limiting at the same time. They may play a crucial part in the (re-)distribution of discursive and material resources necessary to oppose injustice (Fraser, 2012), which can often only be obtained with technical knowledge since other forms of knowledge may be disqualified (Foucault, 1980, p. 85), and access to positions of power and decision-making. But confinement, as a ‘technology of power’ (see Foucault, 1988, p. 18) may be seen as a form of control, as a violent and disciplinary measure, or ‘as techniques for socially engineering people and places […]’ (Khalili, 2012, p. 3). Of course, confinement of knowledge is not the same as confinement of people and I do not want to equate the former with the latter, since the exertion of control and the violence of these instances may be perceived differently. However, we may still be able to draw some overlapping logics from Khalili’s work quoted above with regard to the functioning as governing technology/social and political engineering. This is about whose knowledge becomes dominant or is marginalized, and who has the power to do the confinement of whose knowledge and how. Drawing on Foucault, Li points out how assemblages fulfil a ‘strategic function’ (Foucault, 1980, p. 195): they ‘direct conduct and intervene in social processes to produce desired outcomes and avert undesired ones’ (Li, 2007, p. 264). The confinement of knowledge flows through transnational assemblages is one specific way to direct conduct, to work towards desired outcomes and avert undesired ones. However, since this text is mainly drawing on empirical research in Tunisia, a crucial side note is in order: Ironically, by mostly erasing his experience in Tunisia from his work, and the inspiration he gained from Tunisian revolutionaries, as well anti-imperial and anti-authoritarian struggles, in which he was involved during his time in the country (Medien, 2020), Foucault seems to have done exactly what he criticized: marginalizing knowledge from those who occupy less powerful positions in global politics.

Eventually, by empirically showing confined knowledge flows, the article hopes to contribute to our understanding of how the (neo-)liberal politics of transitional justice that privilege particular forms of knowledge and ‘truths’, of (in)justice and violence, and of visions for the future over others (see e.g., Andrieu 2010; Arthur, 2009; Bowsher, 2018; Franzki & Olarte 2014; Furtado, 2017; Jones, 2021; Miller, 2008; Mullin & Patel, 2016; Nagy, 2008), are reproduced.¹

Following this introduction, I briefly outline the empirical context, as well as how the article came about and the data it is built on. I then turn to a discussion on security knowledge production and transmission in transitional justice. Afterwards, I use empirical material from the Tunisian transitional justice process to show confined knowledge flows and their ambivalent quality. I close with a conclusion.

2. EMPIRICAL CONTEXT: TRANSITIONAL JUSTICE IN TUNISIA

Empirically, I draw on research on transitional justice in Tunisia after the 2011 revolution. Before the ouster of President Zine-el Abidine Ben Ali, Tunisia had been under authoritarian rule since its independence from French colonialism in 1956. Both Ben Ali and his predecessor, the country’s first president Habib Bourguiba, ruled based on violence, repression, and political and economic marginalization of entire parts of the country. Police control wove a ‘net of fear’ over society (Hibou, 2011, p. 81) and a predatory quasi ‘mafia state’ developed (Ayeb, 2011). Quickly after the revolution, Tunisia introduced measures to seek justice and accountability for these atrocities and violations. This started with ad hoc measures based on pre-existing rules and institutions, such as military and civil trials and investigation commissions. These were quickly followed by the development of an institutionalized transitional justice project with strong support of international transitional justice professionals, working for international organizations or NGOs. After a country-wide consultation process that sought input from broader civil society, a technical committee composed of representatives of civil society and the Ministry for Human Rights and Transitional Justice drafted a comprehensive transitional justice law that provided for, among others, the establishment of a Truth and Dignity Commission (TDC), its central institution, specialized chambers in the Tunisian court system, to which the TDC referred cases, and a reparations fund, for which the TDC made recommendations on compensation to be paid to victims.

Because of its comprehensiveness, the law raised high hopes that the corresponding institutions would holistically address physical and structural violence, political, as well as socio-economic grievances, and help to transform state institutions, so that oppressive and corrupt structures would be dismantled. The TDC was operational from 2014 to 2018 and published its final report on its website in 2019, which was then published in the country’s official gazette in 2020. However, so far, the government has not presented a plan for implementing institutional reforms suggested by the commission, which it was supposed to do within a year of the issuance of the report. The specialized chambers took up work in 2018 and at the time of writing have not concluded any case. And while the reparations fund exists in theory, victims have not received reparations from it in practice. This leaves many Tunisians unsatisfied with the transitional justice project, given that neither the promise of accountability, nor of material redress, nor of a way forward for systemic change has materialized.² Given the undemocratic power grab of Tunisia’s current president Kais Saied in July 2021, as well as the dismal state of the Tunisian economy, there are also not many reasons for hope that this will change. However, the transitional justice project has been a site of political struggles from its inception, which makes it a worthwhile subject of inquiry beyond questions of ‘success’ and ‘failure’ (Salehi, 2022b).

3. DATA AND RESEARCH APPROACH

The article is based on field research in Tunisia (almost six months in total between 2014 and 2018) and the United States (12 weeks in total in 2015 and 2019),³ as well as remote data collection by phone and video (2020), interviewing both Tunisian actors involved in the transitional justice process – and deliberately also some who were not directly involved – and international transitional justice professionals. I have also observed situations of knowledge transmission and confinement, such as workshops, conferences, public hearings, and ‘confining’ built environment, in person and digitally. In total, I have interviewed over a hundred individuals, most of them individually and in person, some of them in groups or with another researcher or journalist, some remotely, and some of them more than once. Most of the interviews were conducted in French, some in English, and a few with translation from Arabic (impromptu, non-professional translation for in-person interviews, professional translation for some of the phone interviews). Approval for fieldwork was obtained through local trip authorization procedures at my university. Since the research topic has been politically sensitive and interview partners may have been hesitant to provide written evidence of participation, I obtained verbal consent. If I was allowed to record interviews, which was not always the case, I placed the recording device so that my interview partners could themselves stop the recording at any time. Some made use of this opportunity and re-started the recording when they felt comfortable. Interviews have been anonymized for research ethics reasons. For the purpose of this article, providing the role or function of my interview partners, rather than their personal identities, is sufficient and does not distort scholarly meaning.

I was coming to Tunisia as a researcher from Germany and would introduce myself as such. Given that I was researching an evolving process over several years, I developed familiarity with Tunisian transitional justice institutions, as well as expertise and networks in transitional justice, over time. Therefore, my positionality would change from someone entirely new to the context and subject, to someone who has become acquainted with the context, the subject, as well as some of the actors (see Salehi, 2022a, p. 13). To describe such emerging and dynamic positionalities, Porisky and Glas (2023, p. 51) propose the ‘aspirational status of a ‘credible visitor’’, in which the researcher is not an ‘insider’, but able to show ‘sustained interaction, (…) privy knowledge, and (…) a track record of access’ (Porisky & Glas, 2023, p. 54). Thus, while I have not been an insider, I hope that ‘credible visitor’ aptly describes my positionality as it has evolved over time, while acknowledging that this status does not automatically stick once acquired but requires sustained engagement. Moreover, for interpretive research that builds on longer-term approaches ‘the possibilities of inquiry may change during the process of inquiring itself’ (Salehi, 2022b, p. 6, referring to Elias ((1983) 2006): 383). Therefore, my sustained interaction, privy knowledge and track record of access, as well as the process-concurrent nature of my research, would influence what I would be able to know. In turn, other approaches that rely more on the benefit of hindsight may allow for gaining different kinds of insights.

The argument presented in this article emerged out of a broader research project on the Tunisian transitional justice process. In general, the research has been based on an interpretive approach, going back and forth between theory and empirics. This approach aims to grasp ‘what is going on’ (Schwedler, 2013, p. 28) and contribute to the conceptualization thereof. I developed the argument based on an attention to knowledge orders, production, and transmission that emerged out of the empirical research. Aiming at casing the emerging study (Soss, 2018, p. 21), I attempt to analytically grasp something that crystallized over time through interview data and (participant) observations. For the observations, attendance of conferences, workshops, public hearings, as well as visits to relevant institutions, were particularly important. However, the data I draw on here emerged from both these purposive observations, as well as from what Fujii (2015) terms ‘accidental ethnography’. For some of the instances of knowledge flows and confinement I describe in the article, I only realized later that they would turn into relevant data.

4. SECURITY KNOWLEDGE PRODUCTION AND TRANSMISSION IN TRANSITIONAL JUSTICE

Following Frowd (2018), Lopez Lucia and Martín de Almagro (this issue) define security knowledge as both social and technical. Security is understood here not as a ‘state-centric idea (…) focused on ‘objective’ material threats’, but rather as an ‘intersubjective process of construction and negotiation’. For Frowd, security knowledge also refers to ‘the standards, norms, best practices and objects that transmit understandings of how security (…) should be done’ (Frowd, 2018, p. 9). Both properties, the social and the technical, also mark security knowledge as it is understood in this article.

Knowledge produced and transmitted in transitional justice relates to security in different ways. For the purpose of this article, it is useful to differentiate between two kinds of knowledges: first, the knowledge that should be produced in and through transitional justice processes, in particular in truth commissions, which is related to ‘ways of knowing atrocities’ (Jones, 2021) (the social) and second, the knowledge that is produced and transmitted as a kind of governing technology (Dezalay & Garth, 2011) (the technical) and which may be instrumental in the former.

The first kind of knowledge is knowledge about conflict, violence and repression. Producing this knowledge and establishing a historical record, a ‘shared truth’ (Bowsher, 2018), is one of the central tasks of transitional justice processes in general, and truth commissions in particular (Buckley-Zistel, 2014). This knowledge should establish what kind of atrocities were committed, who committed them, and which structures enabled these violations. This knowledge could and should then be used to hold those responsible accountable, offer some kind of reparation to those who suffered, and to build a basis for transforming repressive structures enabling state violence, often including the security sector. As one of my interview partners put it with regard to the Tunisian truth commission: ‘The TDC is about documenting. I can use the system to document, to mitigate harm, and to have a public conversation about power. And the last point – this is how we can uproot the system.’⁴

The second kind of knowledge relates more to discursive resources or governing technologies that are necessary to conduct these processes of knowledge production and transmission, as well as for the confining of knowledge flows. This knowledge includes, for example, training on what transitional justice is supposed to be according to the dominant, professional understanding. This ‘socio-technological offering’ (Salehi, 2022a), or the ‘supply mechanism’ of transitional justice (Ben-Josef Hirsch, 2006, p. 185), includes dominant knowledge on how things should be done in transitional justice projects, and therefore also how the first kind of knowledge should be produced.

Both kinds of knowledge are therefore pertinent for security knowledge production and negotiation through transnational assemblages (Lopez Lucia & Martín de Almagro, this issue). The first kind of knowledge more so with regard to substance – that means, producing knowledge about violence and repression – and the second kind of knowledge more with regard to governing technologies, and therefore also concerning what kind of security knowledge should be produced and how. International transitional justice professionals accompanied the transitional justice project through its initiation, design and performance phases, sharing their knowledge and expertise, but also advocating for the project when political support declined (see Salehi, 2022a). Their knowledge, and how it is transferred and channelled, plays an essential part in the knowledge flows and their confinement this article is concerned with. So does the knowledge that is produced within transitional justice institutions. For the latter, the focus here is on the operations of the TDC, because of its centrality in the project, and also because truth commissions, as liberal interventions, are co-constructors of knowledge (Jones, 2021, p. 166). They mobilize, and produce, particular forms of knowledge: ‘shared truths’ with the simultaneous function of amassing a society and individualizing subjects (Bowsher, 2018, p. 99).

Thus, in contrast to Frowd (2018, p. 9), security knowledge in this article is not assumed to be ‘usually moving from north to south’. Dominant knowledge – as governing technology – in transitional justice is actually significantly shaped by knowledge from the Global South anyhow (see, e.g., Jones & Lühe, 2021), and experts from the Global South play a significant role as part of institutional Global North actors, such as the International Center for Transitional Justice (ICTJ), and to some degree also United Nations (UN) institutions.⁵ In Tunisia, for example, the office of the ICTJ in the country has been run by Tunisians, and also other consultants that played a significant role were not all from the Global North. In addition, transnational knowledge assemblages also include domestic elites (Dezalay & Garth, 2011) that are not necessarily part of the so-called justice industry (Subotić, 2012). In Tunisia, for example, the transnational assemblage that has shaped the flow and confinement of transitional justice knowledge has encompassed staff members of the ICTJ and other international NGOs working in the field,⁶ representatives and staff members of UN organizations,⁷ representatives of Tunisian civil society organizations, truth commissioners, and sometimes even politicians. Moreover, while the focus of this article has been on dynamics in Tunisia, Tunisians are of course not only conceivable at the receiving end of knowledge flows. Their expertise has been travelling elsewhere, not just since the end of the TDC. Truth commissioners, for example, have been speaking internationally at panels and roundtables or acted as advisors for institutions in other countries.⁸ And Tunisian transitional justice professionals, who worked in and on their home country, have moved on to work in and on other contexts as well, just as other experts from the Global South who work for institutional Global North actors described above.⁹

5. CONFINED KNOWLEDGE FLOWS

In the following section, I show confined knowledge flows and their ambivalent quality through original empirical material. I commence with interview material that shows that technical knowledge of how to do things in transitional justice has not only been imposed ‘top-down’ from transitional justice professionals. Rather, there are processes of ‘co-production of governance orders’ (Lopez Lucia & Martín de Almagro, this issue) through transnational assemblages, and the situated and the ‘globally portable’ (Danielsson, 2023), at play. I then proceed to illustrate the ambivalent nature of confined knowledge flows, showing when they are limiting and knowledge does not flow further or gets blocked. Socio-technological knowledge on transitional justice, as governing technology, is transmitted for example through workshops, trainings, consultations and conferences. The knowledge is transferred to different kinds of actors – for example politicians, government officials, or civil society representatives – who have different interests in what to do with it.

In Tunisia, members of the TDC were part of transnational security knowledge assemblages since they were trained and supported by members of international organizations and NGOs. Therefore, they received access to tangible and intangible resources, such as funds and technical knowledge, that they could then use for the institution to produce security knowledge about violence and repression itself. However, knowledge flows are also confined for both kinds of knowledge in several ways. To give an example, for consultations that should increase inclusiveness and epistemological buy-in of a variety of actors, people who are consulted get trained beforehand on what transitional justice is – or should be (Andrieu, 2016).¹⁰ Thus, consultations are rarely entirely open, but the options on the table are already somewhat pre-determined, reflecting the normative preferences, dominant dogma and socio-technological offering at a particular time (Salehi, 2022a), shaped by liberal interventionism and neoliberalism (Andrieu 2010: 544; Bowsher, 2018; Jones, 2021, p. 171). Such pre-determination of consultations also exists in other fields shaped by similar logics, as Anderl (2022, p. 69) for example shows for the development field.

Turning to a different empirical setting, I afterwards show the different qualities of knowledge flows that are confined with the help of the built environment, respectively infrastructure, which may be exclusionary, curated, and sometimes protective. In all that, they reflect political struggles and power relations.

5.1. Confined knowledge flows as enabling – knowledge as resource

In the following section, I show that knowledge was not only transferred – or even imposed – from the so-called justice industry to domestic actors. Rather, knowledge was co-produced, steered and confined through (dynamic) transnational assemblages. To start with, knowledge provision and transmission through experts/transitional justice professionals was presented to me as very much welcome by many of my interview partners. To give some examples, one of my interview partners, who was part of the technical committee drafting the transitional justice law, described the support they received from international experts – and the co-production of knowledge as governing technology – in the abovementioned consultation process that informed the drafting of the transitional justice law:

In this committee, in fact, what happened is that in Tunisia, several associations and organisations are interested in transitional justice, but they have several different points of view. I can even say contradictory. For example, the point of view towards the victims. One agrees that the victims should be in the process. The others, they refute that they are at the same time the judge and the victim. And therefore, this committee, it allowed in fact the gathering of these various points of view, and we tried … we found as a solution to return towards the Tunisian people, towards the active ones in the civil society, towards the families of the martyrs. What do they want? And that was the dialogue we conducted. Our question was: ‘What do you expect from transitional justice and what do you want from transitional justice? ‘We even distributed a questionnaire, a long questionnaire I think 11 pages … 

Me: Yes, that’s very long.

And it was done by experts. And we tried to gather the results of the questionnaire and the reports of the dialogue (…). In the end, we had all these reports and the results of the questionnaire, and we tried to proceed to the drafting of the project. In fact, we were supported by three international organisations - UNDP, OHCHR, and ICTJ. So, in our drafting, we were almost always assisted by an international expert, provided by one of these partners and we drafted the bill.¹¹

There are several things that we can learn from this excerpt. First of all, knowledge flows were described as enabling, because the transfer of knowledge – as governing technology – provided my interview partner and their colleagues with resources to, in their assessment, better fulfil their task (for a discussion of expert knowledge as resource in law-making see also Lesch & Reiners, 2023). However, knowledge production, flow, and confinement are complex and entangled processes. The second kind of security knowledge, as governing technology, was used in order to produce knowledge that would then be used to decide who may produce the first kind of security knowledge, knowledge about violence and repression (here: should victims be part of the process or not). Therefore, knowledge flows are confined through an assemblage of transitional justice professionals, a variety of domestic actors and the socio-material context. Whose knowledge and how knowledge becomes authoritative as a governing technology is therefore a question of power (Foucault, 1980, p. 81ff).

In a similar vein, a civil society representative – who was actually unsatisfied with the institutionalized transitional justice project and had withdrawn from engaging with it – equally emphasized the enabling function of knowledge flows:

Transitional justice today totally differs from transitional justice 20–30 years ago. Now we can develop transitional justice with the help of experts.¹²

Another member of the technical committee emphasized that receiving expert advice did not mean that they would copy and paste knowledge:

International standards … that’s not copy and paste. So, it is only, we used the experiences that there are in the world on transitional justice. We have tried to take good practices from these experiences and to express them in our law for transitional justice. That's why you … you see that there is everything.¹³

And a civil society representative told me how ICTJ provided them with technical knowledge that they could then adapt and transfer:

We had international supporters, like the ICTJ. They made sure to inform us. We learned more about the process and how to work exactly. […] ICTJ helped us to hold meetings and to learn about the concept [transitional justice] and the purpose of the concept. We worked on applying the concept and make it clearer for victims.¹⁴

Those examples show that ‘international standards’ or the knowledge of ‘experts’ is seen as somewhat authoritative. However, this knowledge is not perceived as being set in stone, but rather as enabling for the co-production of governance orders and security knowledge.

Training was described to me as very important in shaping the transnational assemblage as well as the flow of knowledge. One member of the technical committee described to me how they, working together with transitional justice professionals, shaped what knowledge got transmitted and to whom (see also Salehi, 2019 for a more detailed description of the resulting harmonization of rhetoric):

Yeah, it's true that they all got the same training … which is true I mean. At first, we were a group of twelve … And then we selected a hundred candidates out of four hundred … let's say moderators of debate out of the four hundred candidates that we received, and they all went through the same training on transitional justice, on moderation of debate, on writing reports, on all of this. So, at some point of time, yes, some of them got the same training and they get the same rhetoric but some of them are probably better than others in terms of transitional justice. Some of them are much better than others. Some of them are very new to the subject. So, it depends.¹⁵

For truth commissioners, training – and the financing of it – by NGOs and donors was perceived as assistance (not imposition):

This is one of the partners. ICTJ. So, everything that is training.

Because up until then, ICTJ has provided us with various training sessions, so we have been able to see other experiences with truth commissions. So, we have … 

The dossiers assist us in a lot of our work.

Me: Yes

Not only about money but also training and … 

Just the trainings, the field visits, so … 

Experts who also bring us along … 

The experts, the sharing of experiences, so … 

Me: So, there is a lot of knowledge transfer like … 

Germany also helped us.

Me: Germany, yes.

Through the UNDP.

Yes, through the UNDP, of course. They made a good donation, a big donation.

Me: Oh, right.

To support the truth commission. And this money for example is used there through the UNDP for training, to organise seminars, to assist us in what we are going to do in the commissions.¹⁶

5.2. Not just north-south transfer: joint knowledge production and confinement

Knowledge as governing technology provides discursive resources and interpretive schemes, not only to make injustice visible, but also to formulate demands and gain access to resources. As Dezalay and Garth (2011, p. 277) note, domestic elites ‘are therefore able to draw on the universal legitimacy of the technology or governing device – the imported text – while at the same time using it instrumentally for their own purposes and competitive struggles’. Thus, different actors may also acquire means to draw on to assert themselves vis-à-vis those in more powerful positions, or to exert power over those in less powerful positions.

Domestic actors, as part of the transnational assemblage, actively contributed to shaping what kind of knowledge would be transmitted. To give a concrete example, in autumn 2014, thus during a time period when truth commissioners were already nominated, but the commission had not started its work yet, I took part in a conference¹⁷ in Tunis that brought together practitioners and researchers from Tunisia and abroad. One evening, after the content-related programme had ended for the day, I was driven in a bus together with other conference participants to a reception. During that ride, another participant – an expert from the Global North who had served as truth commissioner in a country in the Global South – told me that he had just been in a meeting with Tunisian truth commissioners, set up by a United Nations agency. It was an interesting meeting, he recounted, but it did not go as expected. The Tunisian truth commissioners were interested in knowing other things from him than the representative of the UN agency had previously asked him to talk about. He stated that he was fine with the shift in topic, since the truth commissioners should know best what they could use from him.¹⁸ Although we did not get into detail about the exact shift in topics, this instance of ‘accidental ethnography’ (Fujii, 2015) shows that knowledge was not only transferred subject to preferences of professionals from the Global North – and therefore solely confined by them. Rather, domestic actors also played a part in what knowledge they would receive, and which then consequently could be used in their own knowledge production practices and flow further. Therefore, the dominant epistemological flows of power were steered and confined by a transnational assemblage.

I could observe a similar dynamic at a workshop of Tunisian and international experts on how to deal with sexual violence within the framework of the TDC.¹⁹ The international experts initially planned to talk about issues very concretely linked to the operations of the truth commission – for example, how to establish evidence for sexual violence for the purpose of truth commission work, when establishing forensic evidence is not possible. Tunisian experts diverted attention from the topics and steered it away from concrete operational questions of the truth commission – how to produce knowledge on sexual violence under repressive rule – to broader questions of flaws in the Tunisian penal system that would not recognize certain forms of sexual violence as such. They thereby also confined the production and flow of the first kind of security knowledge, here knowledge about sexual violence. While I cannot certainly establish the logics behind the steering of knowledge production at the abovementioned workshop, this confinement of knowledge on sexual violence was also pointed out by interview partners after the work of the truth commission had ended. For example, a truth commissioner told me:

The computer application that was used for statement taking in the secret hearings, there was an order from [the president of the Truth and Dignity Commission] that statement takers don’t open questions on sexual violence. They should just open questions for torture.²⁰

Thus, sexual violence was logged as torture and knowledge production and flow on sexual violence was confined. This had the potential consequence of women receiving less reparations than they would be entitled to if the crimes had been logged correctly (Salehi, 2022a, p. 154). In this case, the confinement was done by domestic actors, in assemblage with the governing technologies and infrastructures that were set up together with international experts. Eventually, the chapter on sexual violence ‘disappeared’ from the TDC’s final report without explanation. My interview partners hold the president of the Truth and Dignity Commission responsible, allegedly having been under political pressure.²¹ This was perceived as particularly perfidious, since women testified on their experiences with sexual violence as part of public hearings (see below).²² Here, we can see the simultaneity of amassing and individualizing that Josh Bowsher (2018, p. 85), drawing on Brown (2015), identified for transitional justice’s neoliberal politics: victims should contribute to establishing a particular form of ‘shared truth’, while at the same time being individualized and responsibilized as victims.

5.3. The ambivalent nature of confined knowledge flows

Transnational assemblages do not only influence what security knowledge flows further, what is confined and how, but also the production of security knowledge itself. The simultaneity of enabling and limiting that characterizes confined knowledge flows, therefore, manifests itself also in situations, for example, in which civil society representatives produce knowledge that is then used by or flows further through transitional justice professionals. A transitional justice professional from a UN organization explained to me in an interview that they worked with civil society organizations to do an evaluation on the (ad hoc) reparations programme, as well as a pre-mapping of victims, which would be handed over by them to the yet-to-be-established truth commission: ‘An evaluation through civil society took place for the reparations programme. We will give it to the commission. And a pre-mapping of the victims, which we will also give to the commission’.²³ The international organization therefore does not impose knowledge and ideas ‘top-down’ but transfer and ‘solicitation’ of knowledge are intertwined. This is opening up options to shape the ‘dominant epistemological flows of power’ (Rai, 2003, p. 60).

The task of producing knowledge that would then flow into official discussions and recommendations of the transitional justice project was also described to me by civil society representatives. For example, during the same fieldwork visit/time period, a member of a civil society organization for political prisoners, considered to be close to Ennahda, explained to me how they were doing focus groups with their constituencies, potential victims, on transitional justice and their needs, demands, and expectations.²⁴ They described how this was part of their engagement in the official transitional justice process, supported by UN organizations and the ICTJ.

However, the confined nature of the knowledge flows becomes clearer when we add another piece of information. The abovementioned civil society organization for political prisoners is close to the Islamist part of the political spectrum, and therefore close to the strongest political party at that time (a large number of political prisoners were supporters of the Ennahda party, which was prohibited under dictatorship). Civil society organizations from the more secular part of the political spectrum, however, had withdrawn their engagement in the official, institutionalized transitional justice process, due to disappointment with political decisions, for example the exclusion of civil society from the nomination procedure of truth commissioners.²⁵ Therefore, those organizations that produce knowledge and feed it into the official process are limited and the flow of knowledge is confined through political conflicts and power.

5.4. The ambivalence of knowledge flows confined by built environment/infrastructure

Let us now take a closer look at the material characters of the transnational assemblages that produce, transfer, steer, and confine security knowledge. Lauren Berlant points out that while all times are somewhat transitional, there are times in which infrastructure ‘defined by the movement or patterning of social form’ (Berlant, 2016, p. 393) needs to be particularly taken care of. There is a need ‘to determine the terms of transition’, since they ‘provide conceptual infrastructures not only as ideas but also as part of the protocols or practices that hold the world up. To attend to the terms of transition is to forge an imaginary for managing the meanwhile [… .]’ (Berlant, 2016, p. 394). This perspective allows us to think about transitional justice infrastructures – that are set up in particular moments of transition where change seems possible if properly attended to – in both their material sense and their social sense, and thereby in what they can contribute conceptually to forging such an ‘imaginary for managing the meanwhile’ (ibid.). Jeffrey and Jakala, for instance, critically assess for the example of courts the centrality of materiality, including buildings, for transitional justice processes and its implications. The ‘material and embodied nature’, as they argue, ‘illuminat[e] the forms of comportment, categorization, and exclusion through which [transitional justice] establishes its legitimacy’ (Jeffrey & Jakala, 2014, p. 652). Against this backdrop, I will now turn to another, albeit differently situated, example for the ambivalent nature of confined knowledge flows, which is related to the ‘confining’ built environment and infrastructure of the Tunisian TDC as both a producer and transferer of knowledge. The following sub-sections show three different qualities of confined knowledge flows related to the built environment and infrastructure: the exclusionary quality, the protective quality, and the curated quality.

5.5. Exclusionary quality

Before visiting the Tunisian Truth and Dignity Commission’s building for the first time, I expected there to be some kind of publicly accessible space where knowledge would be imparted and people could inform themselves about the commission and its work, such as an information centre. I expected to find at least leaflets and brochures, explaining the commission’s mandate, organizational structure, and procedures to submit files, maybe even staff members who would explain these things to visitors. However, my expectations were wrong. The main commission building was not publicly accessible. One had to go through a security gate and – at least this was my impression – could only enter the building if one had an appointment. Once in the building, there was no information centre, just a waiting room with not even leaflets or brochures lying around.²⁶ As a researcher who usually had appointments,²⁷ I had to sit in the waiting room until the person I was supposed to meet was free to see me, when I would go to an office on an upper level. Once up on the office floor, it was actually possible to pop into other offices for further conversations. At one visit to the TDC, I attended an event for civil society representatives, which was held in a conference room at the commission building. During this event, one of the participants brought up the inaccessibility of knowledge. She asked why the commission’s numéro vert – a phone number that could be called free of charge – was not working.²⁸ Therefore, one can see from these examples that the TDC, as a producer of the first kind of security knowledge (and arguably as transferer of the second kind), was confining the flows of knowledge through the built environment and infrastructure. Knowledge was channelled into confined flows to those who already had privileged access, such as civil society representatives or researchers, and not made accessible to a broader public. This confinement was therefore reproducing hierarchies and preventing an ‘insurrection of knowledges against the institutions and against the effects of the knowledge and power […]’ (Foucault, 1980, p. 87). Trying to find out more about the seclusion of the truth commission, I was told that the unwillingness to communicate to a broader public – and therefore the confinement of knowledge – was a deliberate decision. For example, an interlocutor told me ‘I was a communications consultant with the TDC. They did not want to communicate, so I quit.’²⁹

5.6. Protective quality

Similar to the main building, the ‘annexe’ building of the TDC,³⁰ a few minutes by foot away from the main building, was also not easily accessible for the public and there was no display of information there either. However, in the annexe building, the confinement of knowledge that was produced in the building not only had an exclusionary, but also a protective function for those who shared their experiences of violence and repression with the truth commission.³¹ The annexe building was where the victims’ statements were taken. I was led through the annexe building and the departments located there by two statement-takers, who explained the commission’s work and the procedures at the different departments. After an initial check of admissibility of their file (see Salehi, 2022a, p. 188) every potential victim should have been entitled to a closed hearing/statement-taking session. As it was explained to me, the statement-taking sessions followed a prescribed protocol.³² The hearing with the victim or, in rare occasions, with a legal representative was conducted by a team of two statement-takers, one lawyer and one sociologist or psychologist.³³ The victims could decide whether they would agree to audio or video recording, and the statements were logged in a specific software.

However, the knowledge about violence and repression that was produced here was also sensitive and the commission staff put great emphasis on their efforts to keep the knowledge safe and confined. While I was led through the building by the two statement-takers, they made a demonstration to show me that the rules in place to ensure confidentiality, and confine the knowledge, were actually working. They told me that statement-taking sessions under no circumstances would be disturbed, and no person could just enter the room, so that the victims would not be interrupted in their session – for reasons of treating them sensitively – and the knowledge produced would not be influenced by other people and would not get into false hands. To demonstrate, one of the statement-takers told me that they would knock at the door of a session in progress, which was indicated by a sign at the door, and nobody would answer or open the door, since this was the protocol. They knocked and indeed, nobody answered or opened the door. They were visibly relieved that it worked out as it should have and joked about it.³⁴

However, given that how knowledge is produced in these settings is determined through governing technologies and the dominant epistemological flows of power, the protective quality is also disciplining in the regard that those who should be protected might not have the power to resist their protection and the confinement of knowledge. Through the confinement of knowledge, they were also at the mercy of the statement-takers who in turn may have been subject to myriad political pressures (see above), and amassed as contributors to a shared truth, while at the same time ‘being individualised as “responsibilised” and self-sufficient entrepreneurial units’ (Bowsher, 2018, p. 85).

5.7. Curated quality

The public hearings of the Tunisian TDC are relevant for the networked production and transmission of security knowledge in different ways. I will concentrate here on the first public hearings in November 2016, since these were the ones which I observed in person, and they received the most attention.

After the TDC had ended its work, the public hearings were often described to me as the most important achievement of the TDC. For example, a truth commissioner told me in an interview: ‘The best gain, the fruit of our work were the public hearings. The victims made it a success. They gained sympathy.’³⁵ Thus, the knowledge on violence and repression produced and transferred in these hearings, and especially its publicness and the emotions conveyed, were essential.

Although the hearings were broadcast at primetime on national television, and therefore widely accessible, the location of the hearings was moved from a large venue in the centre of Tunis to a smaller venue in the suburb Sidi Dhrif. The official reason for the move were renovation works at the original venue. And the truth commission president presented the new venue, the ‘Club Elyssa’ that belonged to former President Ben Ali’s wife Leila, as a symbolic choice for re-appropriating spaces of the corrupt old regime (see Salehi, 2016). However, the move meant that much fewer people could attend the hearings in person. Attendance was only possible after obtaining accreditation and given that the room in which the hearings were conducted could only fit approximately 300 people, these spots were mostly occupied by stakeholders, decision-makers and ‘VIPs’. There was an extra room in the same building with live broadcasting for accredited journalists, and a tent directly next to the venue, also with live broadcasting, which hosted for example civil society representatives and interested members of the public.³⁶

Moreover, in the run-up of the hearings, as well as on site, interlocutors told me that the moving of the hearings to the suburbs did not actually happen for the reasons claimed by the authorities (the renovation). Rather, with the ‘Tunisia 2020’ investment conference coming up, then-President Essebsi wanted to remove the hearings from the city centre to divert attention from the issue.³⁷ At this conference, Tunisia was to present its development plan to international investors, hoping to mobilize funds for large-scale infrastructure projects. The investors therefore allegedly should not learn about violence and repression during dictatorship,³⁸ at least they should not have access to non-curated, immediately transferred knowledge (see Salehi, 2022a, p. 141).

Here we can clearly see the interplay between the material and the social in the knowledge production assemblage. The confined knowledge flows through built environment and infrastructure have been curated at the public hearings to the extent that first, it was chosen carefully who could tell their stories at these hearings. Public hearings exemplarily covered particular topics, reflecting the broad temporal and substantive mandate of the TDC. For example, at the first public hearings those who gave testimony included mothers of ‘martyrs of the revolution’ whose sons were killed during the 2010/11 uprisings, the wife and mother of a ‘disappeared’, as well as two political prisoners from two different time periods, Sami Brahem who recounted how he was tortured during the Ben Ali era and leftist writer Gilbert Naccache who gave testimony about imprisonment and torture during Bourguiba’s rule. Persons were also chosen based on psychological considerations – those who were seen as able to bear the psychological burden of giving testimony under such exposure – and/or societal importance as in the case of Naccache.³⁹ The confinement of knowledge was therefore subject to hierarchies of topics that were deemed relevant, but also of the ability to perform knowledge production in a desired manner (see Foucault, 1980, p. 82).

Second, due to the limited space and accreditation procedure, it was curated – and strongly limited – by the truth commission who got to get the immediate experience and who had to watch a broadcast, either live or on television.⁴⁰ And third, knowledge was curated for those watching the broadcasts through camerawork and cuts. While in this example much of the curation – and therefore confinement – was done by Tunisian actors, both from politics and from those who have been part of the transitional justice figuration, the assemblage was still transnational. The ICTJ and UN agencies advised the TDC on how to do public hearings. This relates to both the content of the hearings as well as the set-up and technicalities.⁴¹

Therefore, we can see again how knowledge as governing technology has been instrumental for the production – and confinement – of knowledge on violence and repression.

6. CONCLUSION

In this article, I introduced confined knowledge flows as a particular form shaping governance orders, reflecting political struggles and power relations. Drawing on interviews and observations related to transitional justice in Tunisia, I showed how knowledge is confined through transnational assemblages, a disposition that includes actors from the Global North and South, as well as the socio-material context they operate in. I conceptualized confined knowledge flows as having an ambivalent quality, since they are enabling and limiting at the same time. In these different properties, confinement of knowledge functions as a form of control, a technique for social engineering. It contributes to the reproduction of the (neo-)liberal politics of transitional justice that privilege particular forms of knowledge and ‘truths’, of (in)justice and violence, and of visions for the future over others.

Governing technologies are transferred through transnational assemblages, which shape security knowledge production, flow and confinement. In this vein, the empirical illustration showed how in practice, knowledge flows are confined for example through workshops and trainings, but also through infrastructure and the built environment. It also showed that while transitional justice works to amass societies, it still individualizes and responsibilizes victims in their struggle for structural transformation. So, commitments to the latter remained largely performative. In privileging certain visions for the future, the formation described here may not be most suitable for achieving structural change (Salehi, 2022a; see also Menzel, 2019). However, in line with the ambivalent quality of confined knowledge flows, this does not mean that institutionalized transitional justice cannot be of use to those crafting more radical visions of justice, accountability and structural transformation. For instance, Rigg shows how the Manich Msamah (‘I do not forgive’) movement made use of ‘discourses and ethics of transitional justice that were circulating within Tunisia and assembled them into new arrangements’. This allowed the movement ‘to intervene in a national political debate (…) while also retaining a space for more radical discussion over the meaning of the revolution and its realization’ (Rigg, 2023, p. 187; 178).

Finally, Jones and Lühe (2021) remind us that a reflection on the knowledge politics of the transitional justice field necessarily includes a reflection on scholarly practice. However, with all these critiques of knowledge politics and charting the dangers of establishing certain truth(s) that are inherent to transitional justice, there is some danger of falling into a nihilistic trap (Brown, 2023, p. 9). Privileging ‘failure’ may also not do justice to those struggling for justice in various ways. Wendy Brown suggests a way out if this trap: a ‘post nihilist affirmation of the complex nature of political struggle’ (Brown, 2023, p. 49). This affirmation would be the basis for ‘academic contributions to post nihilist world making rooted in popular struggles over values reflectively established and also harnessed by accountability’ (Brown, 2023, p. 92).

ACKNOWLEDGEMENTS

I thank my interview partners for sharing their insights and perspectives, and everyone who facilitated access to the settings I could observe to draw on in this article. Previous versions of the paper or the argument were presented at EISA 2018, the Political Ethnography thematic group of the German Political Science Association, the European Workshops in International Studies 2021, at Hanna Pfeifer’s research group at the Peace Research Institute Frankfurt, as well as the German Middle East Studies Association Congress 2022. I thank the participants, Anne Menzel, as well as the reviewers and editors of the journal and the special issue, for critical comments and valuable suggestions.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Bundesministerium für Bildung und Forschung [grant number 01UC1301].

Notes

1 Many thanks to Reviewer 2 for pushing me on this point.

2 For a comprehensive analysis of the Tunisian transitional justice project until 2020, and how it interplayed with the political developments in the country, see Salehi (2022a).

3 In Tunisia, I was mainly based in Tunis, but also did research in the central regions of Gafsa and Kasserine. In the US, I was based in New York in 2015, interviewing staff members of UN organisations and NGOs, such as the International Center for Transitional Justice. In 2019, I was based in Washington, DC, with short research stays in New York and the San Francisco Bay Area. I interviewed transitional justice professionals who (at that time or previously) worked in transitional justice for the US government or for NGOs, most importantly the ICTJ. Some had experience in both sectors.

4 Video interview with activist, November 2020

5 In general, there are several examples of former domestic ‘participants’ in transitional justice processes that later have become internationally working experts (see e.g., Eduardo González, who, in his own words, participated in Peru’s Truth and Reconciliation Commission and now works as an expert on truth commissions: https://gonzalezc.com/; last accessed 15 August 2022). There is also some struggle over who can be an expert where. For example, one domestic transitional justice professional complained to me in an interview that they ‘can’t be experts’ in their own country (personal interview, Tunis, August 2016).

6 The ICTJ is the largest NGO in the field, therefore it deserves particular mentioning here, but other NGOs include for example Impunity Watch.

7 Notably, the United Nations Special Rapporteur, United Nations Development Programme (UNDP), and the Office of the High Commissioner for Human Rights (OHCHR).

8 I personally observed one panel discussion in Berlin in September 2015, where two truth commissioners were speaking. For other examples see the women’s commissioner participating at an event convened by ICTJ and UNICEF in New York (https://www.ictj.org/news/ictj-unicef-education-peacebuilding; accessed 21 December 2022) or the TDC’s president becoming an advisory board member for an American NGO (https://www.iri.org/news/luminaries-join-iris-international-advisory-council/; accessed 21 December 2022).

9 For example, the former (Tunisian) director of ICTJ’s Tunisia office moved to lead the Nepal office at some point, and later to work in other contexts as a consultant. Several personal conversations, the latest in November 2022 in Tunis.

10 Interview with civil society representative/member of the technical committee, Tunis, May 2014.

11 Personal interview, Tunis, May 2014; my translation.

12 Personal interview, Tunis, April 2014; my translation.

13 Personal interview, Tunis, 2014; my translation.

14 Phone interview with civil society activist, September 2020; professional translation.

15 Personal interview with member of the technical committee, Tunis, April 2014.

16 Group interview with three truth commissioners (interrupting each other), Tunis, March 2015; my translation.

17 Regional Expert Conference on ‘Integrating Economic, Social, and Cultural Rights in Transitional Justice Processes: A Vehicle for Reform in the Middle East and North Africa?’, Tunis, September 2014. I participated at the conference in my capacity as a researcher and was invited by the German organizing institution.

18 Personal conversation, Tunis, September 2014.

19 Personal observation, Tunis, March 2015.

20 Phone interview with truth commissioner, December 2020; my translation. See below for more details on the statement-taking process.

21 Several interviews and conversations via phone, video and in person (the latter in Tunis and New York) with international and Tunisian actors, 2020-2022. I did not get the chance (yet) to ask the TDC’s president for her perspective on the issue.

22 Phone interviews with a truth commissioner and a victim’s representative/activist, September and December 2020.

23 Interview with transitional justice professional (UNDP), Tunis, May 2014.

24 Tunis, April 2014. Unfortunately, my notes here are not very precise with regard to the exact nature of the focus groups and their findings.

25 Interview with civil society representative, Tunis, May 2014.

26 I visited the main commission building for the first time in March 2015.

27 Appointments were usually made beforehand by telephone. At the truth commission building, there were usually also people with other concerns than research in the waiting room. I do not know how the procedures were for them.

28 Personal observation, Tunis, March 2015.

29 Personal conversation, La Marsa, September 2016; my translation. I discussed this instance with a transitional justice professional who did not work constantly on site but came to Tunisia occasionally as a consultant. They expressed surprise that the truth commission hired anyone at all for that role.

30 I visited the annexe building for the first time in September 2016.

31 This may also be true to some degree for the main building, but the assessment I draw from my various visits is that the exclusionary confinement of knowledge prevailed there.

32 Procedures in the regional offices of the TDC differed sometimes. Personal observation in Kasserine, August 2016.

33 There may have been three statement-takers when one or two were still in training. Personal observation with explanation from statement-takers, Tunis, September 2016.

34 Personal observation, Tunis, September 2016.

35 Phone interview with truth commissioner, December 2020; my translation.

36 Personal observation, Tunis/Sidi Dhrif, November 2016. Obtaining accreditation was also somewhat cumbersome, respectively, rules only became clear with a short lead-time before the hearings, which may have prevented those from attending who needed to make arrangements to be able to attend.

37 Personal conversations, Tunis, November 2016.

38 Personal conversations, Tunis, November 2016.

39 Personal observation and side conversations, Tunis/Sidi Dhrif, November 2016. See also Salehi, 2016.

40 From my own experience, as a researcher based in Germany, it was not easy to obtain accreditation, or to find out how to do so, until very shortly before the hearings took place. I was in touch with an ICTJ staff member in Tunisia, the Truth and Dignity Commission via various channels (my research assistant also got in touch with them via Facebook), the German Embassy in Tunis, as well as the UNDP Tunisia office. It was through a combination of the latter two that I eventually found out how to obtain accreditation for the event. Once in Tunis, I only found out by chance and through personal connections when and where I needed to pick up my badge.

41 For the latter, personal observation at the first public hearings, Sidi Dhrif, November 2016.