Memories of internal slavery and resistance to slavery in Mali tend to disappear because of continuous public shaming of the people considered “descendants of enslaved populations” and the disappearance over time of the few indirect witnesses still able to tell the story of their enslaved grandparents [Rodet 2010]. Yet understanding this history is crucial in a context where tensions surrounding descent-based slavery have arisen in Mali since 2018.

Slavery existed in the Sahel before the Transatlantic slave trade and endured beyond its abolitions. In the 19th century, Kayes in western Mali was a major transit zone of slave caravans and experienced an expansion of internal slavery through war and conflict. French authorities abolished the internal slave trade in their West African colonies in 1905 as they searched to recruit liberated enslaved people for forced labour. They soon turned a blind eye on the continuation of what they called “domestic slavery”, pretending it had simply transformed into salary contract work thanks to the colonial legislation [Camara et al. 2021a]. Descendant of those enslaved people continued to inherit “slave status” and their labour remained controlled by the local historical ruling class with the complicity of the colonial and postcolonial state, a system that allowed historical hierarchies to persist to present day in the form of what is henceforth called “descent-based slavery”.

Despite the patchy implementation of the 1905 abolition decree on the ground, some formerly enslaved people managed to take their destinies into their own hands, escaping a violent institution to live freely in independent communities such as Bouillagui in western Mali. But this history has remained largely silenced as the colonial and local ideologies prevailing at the time of abolition endured in the postcolonial period which has prevented the stigma of slavery from disappearing. On one hand, although the region experienced successive waves of emancipation in the colonial and postcolonial eras, descent-based slavery continued under disguised forms of kinship, marriage, and adoption/fosterage [Rodet 2013]. On the other hand, the historical ruling class remained largely in power despite abolition and independence and thus continued to exercise control over defining what a society of “freemen” had to be [Rodet 2010] [Rodet 2013]. In such a context, confining the history of slavery to a “public secret” well into the postcolonial era was most likely a necessary step in conveying an authoritative national identity. Today, poverty and discrimination continue to exclude Malians with ascribed “slave status” from social mobility and to restrict their agency about how their history is being told in the public space. Conversely, the elite continues to benefit from their ostracization and exploitation by ruling the moral economy of honour and shame [Camara et al. 2021b] [Iliffe 2005].

Postcolonial Mali has never criminalized descent-based slavery, despite the numerous advocacy campaigns conducted by Malian human rights and anti-slavery organizations. Mali has certainly signed up to the major international conventions banning slavery (including the UN International Declaration on Human Rights) and passed a law criminalizing international trafficking in 2012. More recently, slavery has been made a criminal offense in the criminal code in Mali, but its definition is still based on the idea of property and thus cannot capture the complexities of descent-based slavery as a web of socio-economic subjugation and control that does not necessarily manifest in overt physical violence. Perpetrators of descent-based slavery can thus still only be prosecuted if physical violence has occurred. The continued amnesia on the history of slavery and fraught political environment in today's Mali contribute to the state's inability or unwillingness to understand descent-based slavery as a contemporary manifestation of slavery [Camara et al. 2021b].

New technologies have the potential to complement fading oral history transmission, offering new ways to engage, share, and understand collective heritage histories for the longer term. In 2018, we began working on the digital public history research project “Visualising Liberté”, which was intended to explore the history of the village of Bouillagui in western Mali. Bouillagui has a very specific history: its founders liberated themselves from slavery at the beginning of the twentieth century. Despite the general taboo still surrounding slavery in West Africa, the people of Bouillagui are very proud of their history, and they were particularly keen to share it within the framework of our digital history project recounting the specific histories of the struggle for freedom and resistance and rebellion against slavery.

“Visualising Liberté” was conceived as a transmedia research project: the villagers were given mobile phones and trained to make short video reportages about their village history, which would become the narrative basis for a graphic novel, an animation film, and a web documentary. The web documentary itself was also conceived as a transmedia digital object, in which the new videos made by the villagers themselves were added to the associated fragments of this history of resistance to slavery that had been collected over the previous decade of research conducted by Marie Rodet in multiple media (audio, textual documents, video, photographs, etc.). All this was pieced together into a single new digital media platform and made accessible to various audiences, in the form of the 2020 web documentary titled Bouillagui: A Free Village (https://bouillagui.soas.ac.uk/) [Rodet and Doquet 2021].

In Mali, new technologies are often blamed for the disappearance of the intergenerational oral modes of knowledge transmission. Creating digital media on the history of resistance to slavery was therefore a crucial opportunity to reflect further on how media may be used widely and efficiently to convey a difficult and still traumatic past, which could be capitalised upon to fight against discrimination, exploitation, and modern slavery today.

In the past fifteen years, historians and anthropologists of slavery and post-slavery have paid great attention to retrieving the voices of (formerly) enslaved populations and to analysing oral and written slave narratives [Bellagamba, Greene, and Klein 2013]. However, they have been slow to reflect on the potential ethical implications of such endeavours, including the broad political dimensions of such research and the intricate power relationships enabled by knowledge production about (formerly) enslaved populations [Brannelly 2018].

“Visualising Liberté” was concerned with “listening to the history of those who don't forget” and making these voices — which had often been silenced in a form of continuous epistemological violence — widely heard on their own terms [Rodet 2013]. As with many other public projects dealing with contentious issues, the digital was to be used as a “moral technology” to influence and modify thoughts and behaviours regarding slavery in Africa [Cameron 2008], while also supporting knowledge production by a marginalised community that fits with their own ways of knowing, through visual participatory research with mobile phones [MacEntee, Burkholder, and Schwab-Cartas 2016]. Here, the participatory visual research was used to explore new ways of publicising a neglected history of resistance to slavery, with a commitment to both social justice and an “ethics of care” [Brannelly 2018].

Several events during the project reminded us of the unpredictability of the research process and of several ethical implications. Even as we tried to mitigate ethical issues from the conception of the research project to its end, the COVID-19 pandemic still strongly impacted the co-production dynamic. More importantly, during the project and until today, the violence against people who are ascribed locally with the status of “descendants of slaves” and who refuse to be called “slaves” increased in their home communities. In 2018, the anti-slavery group Gambana or Ganbanaaxun (which means “All Equal” in the Soninke language) was formed by Soninke migrants in Paris to fight systematic discrimination and violence mainly among their ethnic group, which is present in several West African countries, including Mali. Activists denounce the anachronistic and demeaning use of the word komo and the worst forms of exploitation this use results in [Camara et al. 2021a][1]. Their protest against descent-based slavery in their community generated a violent backlash: since 2018, more than 3,000 Malians with ascribed “slave status” have fled slavery-related violence in western Mali [Cissé, Pelckmans, and Rodet 2024]. Activists and members of the movement have also been attacked and threatened in several localities of western Mali and beyond, yet the state and global community have remained largely silent [Rodet et al. 2021].

This situation changed the ethical stakes of “Visualising Liberté” and its potential impact: from retrieving the history of people who were proud of having successfully liberated themselves from slavery to rethinking what terms to use and safeguarding measures to add to ensure that exposing people's history would not endanger them. The core question thus became how to conduct a digital public history project that supports social justice, by bridging the gap between endogenous historical resistance to slavery and modern anti-slavery activism, without doing harm to the persons concerned.

In this article, we will focus on the ethical issues at stake in co-producing digital historical knowledge about resistance to slavery in a web documentary that involved identifying and naming marginalised populations in Mali. The other two media produced by “Visualising Liberté” — the graphic novel and the animation film — have not been officially released yet and are closer to fiction and less documentary in their nature, using fictional names. They thus do not involve the same ethical dimensions, although those related to co-production, shared authority, and equitable partnership could be approached in similar ways. The ethical and safeguarding issues raised by this project will not be resolved exhaustively in the body of this article, especially since unexpected issues may continue to arise due to the fast-changing political and digital landscapes worldwide. Nonetheless, we would like to confront here most of those that have crossed our research paths since the start of our collaboration with the Bouillagui villagers.

Here, we are attempting to uncover some of the blind spots in postcolonial digital historical practices and writings. As the initiator of “Visualising Liberté”, we built individual careers in recent years based on research material collected in West Africa in collaboration with marginalised populations and with the support of the Malian organisation Donkosira (lit. “the road to knowledge”, in the Bamanan language), an association working on the dissemination of local community knowledge in West Africa whose founding members are Mamadou Séne Cissé, Moussa Kalapo, and Mariam Coulibaly.

From a decolonial perspective, there is a pressing need to pose rather uncomfortable self-reflecting questions about our own professional practices as historians of Africa and cultural practitioners conducting digital public history together. This is a crucial first step towards confronting the potential (negative) impact of our digital interventions. Indeed, we must question whether a project about decolonising historical knowledge production — through digital media, based on equitable partnership and co-production, and “decreed” from the outside — is not ultimately doomed to failure. And in the case of research on slavery, is it not at risk of exposing and even doing harm to the marginalised populations involved? To what extent have we succeeded in making the digital and interpretative “shared authority” and the historical knowledge co-authorship recognised in this project [Frisch 1990]? Have we contributed to the creation of a “safe space for unsafe ideas” [Gurian 1995] [Anderson 2018]? How have we been able as a team to work around equitable digital shared authority against a moving political landscape in both Europe and Mali, while the process also entailed naming populations at potential risk? What is the effect of our intervention of digital transmediation or remediation on the voices publicised? How can we position ourselves as researchers and practitioners in such contexts where radical asymmetries in terms of safety increase? How can we ensure that, despite these radical asymmetries, the project works for everyone and guarantees individual safety?

Few researchers and practitioners dare to openly confront these issues of “North–South divide” and their ethical implications.[2] Despite all its promises, the digital poses sepcific challenges in terms of funding, authority tensions, and asymmetrical relations. Without claiming to propose a protocol, we reflect on a case of dialogue among researchers, practitioners, and villagers, and we expose digital research and citizen intervention mechanisms that attributed singular places to each.

As a public history project, “Visualising Liberté” involved more than using new media — the web documentary — as a dissemination strategy. Since first coined in the United States some fifty years ago, public history works to make history-making largely accessible to large audiences while increasingly using various forms of collaborative, participatory research and communication processes, including digital ones, thereby fostering a more “democratic ethos” in historical practices [Frisch 1990] [Rosenzweig 1998] [Noiret 2022] [Basaraba and Cauvin 2023]. This latter aspect of collaborative, participatory history-making was at the heart of “Visualising Liberté”: using the digital as a powerful tool to accelerate the potential impact of little-known histories of African resistance to slavery.

“Visualising Liberté” emerged after a decade of research focused on the history of slavery and emancipation in the Kayes region (1890–1940) in Mali and following a first experience of public history through a film in which Bouillagui's history was already partly featured [Rodet 2020]. Rodet's first interviews in Bouillagui date back to 2009. When we began “Visualising Liberté” in 2018, the history of internal slavery had long been a taboo subject in West Africa, but there was a demand from local communities in Kayes (western Mali) to know and understand better this difficult past. Populations with ascribed “slave status” are still discriminated against and stigmatised today, though the younger generations are less aware or understanding of the underlying reasons. Many of the victims of descent-based slavery are also among the most vulnerable to economic uncertainties and thus at continuous risk of being subjected to further exploitation, with new forms of servitude strongly overlapping with the legacies of historical slavery [Rodet 2016] [Rodet et al. 2021]. The main goals of “Visualising Liberté” were to capitalise upon historical resistance to slavery in Kayes to produce visual material for the young generations and to raise awareness about slavery and its complex legacies in Mali, while also bridging the gap between historical resistance to slavery and contemporary anti-slavery activism.

With this project the intention was not only to retrieve local narratives and memories of resistance to slavery, but also to bring these histories to a new dimension, in close collaboration with the community, including school children, through the organisation of participatory research activities with strong social impact. It was both a response to specific historiographical questions about African resistance to slavery and a research-action programme raising awareness among younger generations and contributing to the fight against slavery's discriminatory legacies and modern exploitation. As already discussed elsewhere, what is often misunderstood is that descent-based slavery sustains modern slavery, in which descent-based slavery creates a terrain for a continuum of violence. It is the banalization of descent-based slavery and its violence which trivializes next-door exploitation and allows overall social acceptance of modern slavery [Camara et al. 2021b].

With an “ethics of care” valuing longer-term, ongoing relationships and the needs and wishes of the community [Brannelly 2018], this project was the second participatory research project led by Rodet in partnership with Donkosira in which Bouillagui was involved. Bouillagui villagers have since been involved in three other participatory research projects led by Rodet and Donkosira.

Bouillagui's history demonstrates that formerly enslaved populations were not just recipients of colonial abolition of slavery, but rebelled against slavery on their own terms, in some cases violently, and obtained equality through liberating themselves. Similar self-liberation movements would also be followed by other enslaved populations, initiating a series of migratory waves to escape slavery during the twentieth century [Rodet 2015]. This history needed to be named and taught, to help in the fight for human rights and against other forms of modern exploitation. The timing of the project was particularly crucial, as Mali had been encountering at the time deep political, economic, and social crises for almost a decade at the time, a situation that risked increasing further discrimination and exploitation of already marginalised communities. What we did not predict was that the rapidly deteriorating political situation in Mali would effectively prove our hypothesis of increased risk for these populations immediately in the first year of our project (though independently of it). The history of slavery has since been more widely discussed in Mali in the public space, following peaks of violence against these populations since 2018 and for which the victims have regularly been blamed until recently. We will return to the specific impacts of this latter aspect on our project in Section 4 below. For now, we will just add that other research-action interventions within the framework of a follow-on research project (also led by us) on “Slavery and Forced Migration in Western Mali” (2020–2023) have since contributed to legitimising Malian grassroots anti-slavery activist organisations such as Gambana RMFP and to ending the usual “naming and shaming” of victims in the public debate [Rodet et al. 2021].

In the “Visualising Liberté” project, we worked with five Bouillagui villagers: Hawa Cissoko, Salou Diarra, Diangou Diakité, Aboubakar Traoré (deceased in 2023), and Wally Traoré, who had all been designated by their home community as knowledgeable in the history of their village and available for the project. We trained them in a workshop to collect local historical knowledge via mobile phones (provided by the project) and to translate this into non-fiction visual narratives for the broader public, with the help of the Donkosira team. Once the storyboard was conceived, it was shared with the populations of Bouillagui, who gave feedback and further input for the production of the full graphic novel and animation. The school children from Bouillagui also worked with Donkosira on drawings about the history of their village and on collecting family histories. These activities were ultimately integrated into the script of the graphic novel.

By making visible in an accessible way little-known historical resistance against slavery in their own community, the villagers thus co-produced educational material about human rights, citizenship, and social justice. They further contributed to raising awareness among young generations about the importance of fighting against all forms of discrimination and exploitation, including modern slavery, by making the history of resistance to slavery better known and recognised.

Overall, the project had as a key goal to develop a model of equitable partnership with Bouillagui villagers. The project team worked continuously and closely with the villagers to ensure that they remained in control of the narratives about their village and how experiences of slavery were recounted and transmediated. These strategies of inclusion and participation ultimately aimed at reintegrating marginalised historical experiences into the global production of legitimised knowledge, while using the digital media as powerful civic storytelling to spark public dialogue for cognitive justice, democracy, and social struggle [Piron, Régulus, and Dibounje Madiba 2016] [Hall Budd and Tandon 2017] [Borum Chatoo 2020].

Let us now turn to unpacking the ideas of “equitable partnership” and “co-production” of digital data in this project. While these are increasingly stated imperatives of funders to support projects, especially in the United Kingdom, and to decolonise knowledge production on the Global South, surprisingly many historians of Africa are still struggling to confront the asymmetrical relationships of the researcher with “their” fieldwork (the possessive is extremely revealing) and to recognise the complex positioning of “intermediaries” in this mechanism, who are often “erased” [Bertho et al. 2022]. In such contexts, how do we involve villagers in a co-production using different media (graphic novel, animated film, web documentary)? How can we invite them to be actors in their own right, guiding the preservation of local history that is part of a personal, intimate, and political agenda but will also be remediated for dissemination in a digital community? That is, what does it look like to create opportunities for villagers to become public historians through digital media beyond their own community?

As Jennifer Hart, Victoria Ogoegbunam Okoye, and Joseph Oduro-Frimpong point out, it is important to distinguish between co-writing (exchanging and writing together to produce a text), collaboration (generating knowledge with the participation of diverse social actors and forms of knowledge), and co-production (including research participants as active stakeholders throughout the research production process).

To what extent does working with audio-video digital data — and in this specific case transmediated or remediated historical narratives — change the asymmetric research relationship? Oral and public histories have continuously disrupted the primacy of the written document and the intellectual and interpretive authority of trained academic researchers, while encouraging broader participation and inclusivity in historical debates and producing new historical perspectives from below [Frisch 1990] [Dorn 2013]. The digital permits experimentation with alternative publishing platforms and processes that make research more broadly accessible and allow audio-visual sources to be published directly online.

Interpretive historical authority in the digital age dominated by online mass media often involves remediation. It makes one medium, such as oral historical narratives and memories that have already been refashioned in audio-visual interviews during the research process, “remediated” on a new online platform [Bolter and Grusin 1999] — in our case, a web documentary. This immediately invites us to question whether such remediation changes the political meaning of the shared narrative, given that it is no longer intended for an intimate political community agenda but for the wider world. We will explore these questions more carefully in Section 3 of this article.

Digital technology, especially the mobile phone, was used to help strengthen the power of the Bouillagui villagers to narrate their history on their own terms without the immediate intervention of the researcher. Indeed, the wider use of digital technology (phones) was intended to amplify the impact we already attempted to achieve with the free-access version of the 2014 documentary film The Diambourou [Rodet 2020]. The first part of the process involved a familiarisation phase — with the project, the various team members, and the mobile phones — so that the villagers could reappropriate the project fully. Participants also drew up their own lists of historical knowledge to be preserved as a priority through their video recordings, such as the song that announced the self-liberation plans and how their ancestors managed to survive the post-liberation famine they experienced. The villagers' video reports and additional research conducted by Marie Rodet and Mamadou Séne Cissé became the basis for the web documentary Bouillagui: A Free Village.

Bouillagui's history of resistance to slavery is a complex one to narrate, since people categorised as “descendants of slaves” are still stigmatised today in Mali. Honour is still associated with the historical ruling class — that is, the so-called nobility [Rodet 2010] [Rossi 2013]. Associating the history of slavery with resistance in the web documentary was to offer a counter-discourse to the public shaming of victims of slavery. The “unsafe ideas” here go beyond recounting the history of resistance against slavery, which has been “whitewashed” by national, hegemonic discourses [Rodet 2010] [Lloyd and Moore 2015]. By co-producing a “safe space” for alternative historical voices, the web documentary gives authority to Bouillagui villagers who had been denied the right for their history to be integrated into public, official Malian history.

Naming Bouillagui villagers in the webdocumentary credits as co-producers in a digital public history project that recounts the history of their ancestors' self-liberation was to inevitably identify them publicly as “descendants of self-liberated slaves”. However, the Bouillagui villagers did not seem to have an issue with this, as what mattered to them and seemed to erase all other aspects was the “self-liberated” part of the designation. It is their history, a history in which they have cultivated a fierce pride and that they wished to publicise and name in order to defend full and effective equality and citizenship in the longer term in Mali.

The project aim was to further nurture collectively this discourse of pride around the notion of resistance to slavery and the promotion of Liberté (the name given to several villages and neighbourhoods founded by freed enslaved people in the region with the support of the colonial authorities), especially within the new generations that did not always know the exact circumstances of a village's creation and its links with the history of slavery. We thus carried out extensive educational work, collecting family histories with Bouillagui's primary school pupils. This work was integrated into both the web documentary and the script of the graphic novel.

Thus, the digital technologies used in this project were not about showcasing a new fancy digital dissemination strategy; instead, they operated as tools to deeply shape scholarly collaboration, ethical practices, citizen intervention, and communication of research-action results from within the Bouillagui community to a wider audience and academic community [Wyatt 2012].

Much has been written about the (unintended) consequences of e-research [Wyatt 2012] [Christensen and Larsen 2020], participatory visual research [Cahill 2007] [Walsh 2016] [Dickens and Butcher 2016] [Marzi 2023], and digital storytelling for the people involved [Edmonds et al. 2016] [Gubrium, Fiddian-Green, and Hill 2016]. However, the ethical impact of using digital technologies in historical public research has been little studied, although ethical issues in digital history, including concerning slavery, have been of increased scholarly concern in recent years (see for example [Lovejoy, et al. 2022]).

Digital technology made available to everyone through a simple smartphone often gives the illusion of a catch-all solution that will level up all communication and technological asymmetries while facilitating democratisation, equitable participatory research, and digital co-production for social change, awareness-raising, and educational purposes — especially through the production of what have been termed “cellphilms” [MacEntee, Burkholder, and Schwab-Cartas 2016]. However, the notion of “co-producing” itself remains fraught with a symbolic power relationship that still gives pre-eminence to the researcher, who eventually remains in charge of the budget that also conditions the research process [Bertho et al. 2022]. “Radical trust” seems possible only when the participants have at least the same technical savvy as (or even more than) the researchers, which is rarely the case when working with marginalised populations from the Global South — a context in which visual participatory research led by a researcher from the Global North may actually increase hierarchical power relationships [Gardner 2010] [Walsh 2016].

With “Visualising Liberté” — despite a carefully thought-through digital data production protocol with the villagers and the crucial building of mutual trust to facilitate the remediation of Bouillagui historical narratives into a digital public history project — the asymmetry was never totally neutralised, since the researcher retained ultimate authority regarding technological decisions that guaranteed, to the funder, achieving production of the project deliverables. This symbolic power was also exacerbated during encounters when we were unable to completely break the domination of the use of French, despite Soninke–French live translation. It especially affected the integration of female villagers, most of whom are non-French speakers. Ultimately, the digital gap was underestimated. It would be naive to think that the digital could have eliminated socio-geographical inequalities once and for all. If a targeted audience is not already widely using a specific medium in their daily life, it can be more difficult to reach them by encouraging them to participate in a medium that they are not familiar with. Despite initial training in cellphilm and historical data collection, the first videos produced were often of poor quality and had to be remade with the support of Mamadou Cissé.

The digital work undertaken was nonetheless facilitated by continuous communication via social media between the researcher and the Donkosira team. At the same time, issues of internet access in such a marginalised community remained one of the major challenges of the project. Aside from the initial training workshop with the villagers conducted in Bamako in December 2018 and the follow-up visit to Bouillagui by the project team, direct communication with the villagers remained the prerogative of the Donkosira team — and, in particular, Mamadou Séne Cissé, who visited Bouillagui every three months on average over almost two years (2019–2020).

The impact of the COVID-19 pandemic and the resulting series of lockdowns exacerbated beyond our expectations the ethical tensions at the heart of co-production. Lockdowns increased, to some extent, what we initially wished to avoid — namely, the risk of delegating the production of research materials and the collection of data to “invisible third parties” [Bertho et al. 2022]. Marie Rodet was prevented from travelling again to Mali before the end of the project, which made face-to-face activities crucial to the co-production of the web documentary impossible. Many of the project's other activities were then led by Donkosira — albeit in close collaboration with Marie Rodet via social media — while an effort was made to prevent Donkosira's members from becoming “instrument-bodies” used remotely to access difficult terrain [Bisoka 2020]. This autonomy, reinforced by the constraints of the pandemic, still enabled Donkosira to experiment with research techniques and strengthen their knowledge of local history, both theoretical and practical. Importantly, the designation from the outset of the project of Donkosira as the main partner, with their own allocated budget in a legally binding collaboration agreement with the School of Oriental and African Studies (SOAS) at the University of London, nonetheless greatly mitigated the risk of the Donkosira team's invisibilisation in the process. Mamadou Séne Cissé conducted two workshops in Bouillagui with school children, further supported the villagers in making their video reports about the history of Bouillagui, and conducted additional interviews with elders in the community, while also regularly organising large-screen screenings on the main village square of the historical material produced within the framework of the project. Ultimately, the web documentary was also screened with live translation when necessary in Bouillagui in November 2020. Everyone was able to recognise themselves or their relatives and friends and further discuss the different aspects of Bouillagui's history.

The web documentary is thus the result of research only partly co-produced with Bouillagui villagers. While the authority was shared in producing the public history narrative at the origin of the web documentary, the project struggled to escape completely the classic mechanisms of practical and symbolic inequality in researcher–villager relations, conditions exacerbated by the consequences of the COVID-19 pandemic. The villagers could not be directly involved in the final stages of the production of the web documentary, which was thus not designed by them but instead was produced online between the United Kingdom and France by Marie Rodet — with the editing support of Cosimo Maximin and the team of the Paris-based creative industry firm Narratives, which specialises in the production of web documentaries. The web documentary remediated and translated historical narratives co-produced by Mamadou Séne Cissé and the villagers, but also older filmed interviews conducted by Marie Rodet for the documentary film The Diambourou with Fanny Challier [Rodet 2020].

If there is no universal recipe for escaping the epistemic violence of knowledge production, the devices — the provision of mobile phones, the attempted insertion into a network of digital family practices, the passage through Donkosira with which the villagers could easily identify — worked to at least nuance and enrich the bilateral relationship between investigator and respondent [Bertho et al. 2022].

Creating a web documentary was not an immediate choice when we started the project in 2018. Our initial main concern was to explore how to engage with different audiences using the digital media and narratives produced with the villagers within the framework of this project. As the media had been co-produced by different authors, we aimed at offering an overarching narrative piecing together polyvocal perspectives while making the material accessible to a large audience. Yet, a specific event made us feel strongly about the pressing need for a rapid alternative publishing platform solution. The Kayes public archives, which we had helped salvage a decade earlier with Mariam Coulibaly (and which had not been digitised), were reduced to ashes at the height of a very restrictive COVID-19 lockdown that fueled widespread discontent and large-scale mob protests in Kayes in the spring of 2020 [Rodet and Doquet 2021].

In such a context, and unsure how the rest of the visual participatory research would unfold in the conditions of the COVID-19 pandemic, we decided to make a web documentary with digital storytelling professionals who would produce such output in a timely manner. In addition, the web documentary seemed compatible with the initial idea of writing bottom-up history with the community. Its architecture could easily accommodate transmediation (multiple media use) and remediation (change of media), while acknowledging the intrinsic shared authority of the narratives that had been produced through the active dialogue between historical expertise and experience [Frisch 1990] [Frisch 2011]. Indeed, the web documentary offered a polyvocal transmedia narrative of perspectives often missing from conventional history.

While offering a polyvocal narrative, we did not aim at integrating all voices about slavery in this web documentary. Indeed, we did not give space to the memories of former enslavers, as the web documentary aimed at changing the mainstream discourse in Mali — which still mainly featured the negationist voices of the former enslavers at the time [Camara et al. 2021b]. Although objectivity is often argued to be a pre-requisite for the search for historical “truth”, what is hardly considered under such an imperative is the harm that it can produce by providing “equal” space to all sides of the story, including the aggressors, while not acknowledging how it can reinforce and reinscribe existing systems of oppression and inequality. People with ascribed “slave status” continue to face systemic and real violence in present-day Mali, and they should feel safe when co-producing and using digital media in relation to their history. To this end, the recognition of the root causes of harm, past and present, is essential so that digital media used for research do not reinforce or reproduce intergenerational trauma and historical harm. On the contrary, they should be transformative by acknowledging and publicising histories of resilience to slavery. Thus, the project never intended to research conflict narratives or integrate them into the web documentary, unlike what has been done in other transmedia storytelling in digital public history [Basaraba and Cauvin 2023] [McRoberts 2016]. Instead, it prioritised acknowledging a history of marginalisation, valuing experiential knowledge, and sustaining research praxis review, following key principles of an ethics of care: attentiveness, responsibility, competence, responsiveness, and solidarity [Brannelly 2018].

The web documentary was certainly based on historical research and ethical standards that required fact-checking. However, it still departed from the usual historical monograph narratives: the artistic and interpretative approaches were defined by the imperatives of using digital storytelling techniques to get the social message across. It included aesthetic choices and necessary simplification of the message to reach a broad audience. The choice of having an aesthetically pleasing layout reinforced the value of telling a marginalised history and aimed at making clear that such a neglected topic deserved as much attention as any other “major” historical subjects and a dedicated “safe space” for this purpose.

Digital media change our practice of history but also our relation to the past and the future, as they alter “how we value, access, and use pasts and histories” [Tanaka 2013]. For example, a web documentary allows the presentation of multiple temporalities and voices while working around historical memories and consciousness where the past continues to be present in our lives — what Tanaka calls the “practical past” [Tanaka 2013]. Nineteenth-century positivism, by approaching the past as a linear, chronological structure and professionalising history as based on objectivity, scientific methodologies, and truth, relegated the “practical past” to the realm of “tradition” and colonised societies were perceived as “a-historical”. The web documentary — by involving Bouillagui villagers as historical knowledge producers — contributed to decolonising such positivist negation of the multiplicity and heterogeneity of pasts. By offering new narratives to link the impact of neglecting such history in recent years in Mali to the renewed violence against people with ascribed “slave status”, the documentary bridges the gap between endogenous historical resistance to slavery and modern anti-slavery activism.

While historical digital storytelling by “non-professional” historians is still rarely considered “proper history writing” in the academic world, historians have nonetheless a responsibility (duty of care) to work as allies with communities to make neglected histories heard and legitimate [Rosenzweig 1998]. They can use their expertise and authority to create with communities digital “safe places” that speak for all the people involved in the process and that are grounded in an ethics of care with responsibility to produce change. Moving away from their authoritative (narrow) expertise, they become curators of collected material that they help contextualise through increased collaboration with the broader public for a better understanding of human experience and for socio-political change [Gardner 2010] [Santana 2022] [Tebeau 2022]. The web documentary Bouillagui: A Free Village is closer to a curated online exhibition with a guided path for the visitor than to a co-produced online primary source repository. The videos have been edited and subtitled in English and French. The navigation follows a specific narrative path, although the visitor can jump from one section to the other, back and forth, without disrupting the understanding of the content.

While historical digital storytelling may be collaborative, including with the creative industry, the risk still remains of blurring the different voices and contributions in general end credits, in similar ways to those in other digital humanities projects that can be explicitly disentangled only in an accompanying academic article, as highlighted by [Lovejoy 2020] and which is attempted in the present article. Indeed, digital humanities projects are the result of multi-layered and multi-partner collaborations, with each partner having their own practices, constraints, and understanding of their role. To turn the conducted historical research about Bouillagui into a narrative for a broad audience, we worked with the Paris-based creative industry firm Narratives, which has longstanding expertise in audio and visual storytelling. It was their first audio-visual project concerned with African history. As they were not involved in the initial research process of co-production, their understanding of the project was primarily centred on the transmediation process and thus about sharing the historian's knowledge and methods (to be narrated in the form of a voice-over by the historian, which they insisted was needed) with the broader public that — we realised too late in the process — they perceived mainly as a Western-based audience. It was Rodet's first experience with the storytelling creative industry and business-format constraints, which unfortunately allowed little leeway and no trial-and-error process.

In addition, the production took place during the first 2020 COVID-19 lockdown, preventing the direct inclusion of the community in the production process. In these circumstances, we realised only after its production that the web documentary was not usable by the Bouillagui villagers themselves because of its non-accessibility on mobile phones (no app version) and its non-intuitive digital design (for villagers with low literacy). These unacknowledged initial misunderstandings ultimately turned the web documentary into more of a teaching tool for Western universities, with university colleagues acknowledging how the web documentary gave access to voices that otherwise are difficult for teachers to transmit to students [IFPH Public History 2020]. We are currently working on a Bambara and Soninke video version of the web documentary as well as on a mobile app to mitigate accessibility issues for Malian audiences.

One may well wonder whose voices are really to be heard in this web documentary and for what social change. Who owns the past and how is it used? How can we ethically represent others? How do we choose between competing agendas without turning such narratives into history for Western consumption and compromising on the ethics of care [Pickering and Kara 2017]? This requires a strong commitment to integrity in the storytelling and editing process [Borum Chatoo 2020].

The history of Bouillagui here is not just transmediated but has also been remediated. The narrative has been edited and curated by a historian and a professional storyteller, creating post-storytelling that can raise concerns about how testimonies are mediated and presented to audiences [McRoberts 2016] [Pickering and Kara 2017]. As already mentioned, the villagers could not take part in selecting or interpreting the video material for the web documentary, which remained in the hands of the researcher. The historian and the professional storyteller are the two named editors of the web documentary and may thus appear as the primary benefiters professionally of this controlled shared authority, with a risk of appropriation, tokenism, and commodification of research [Rose 2014] [Lloyd and Moore 2015] [Brannelly 2018] [MacEntee 2016]. Although all contributors to the research project are named in the final credits, such credits are different from acknowledged authorship and copyrights [Lovejoy 2020].

Despite these significant pitfalls, making a web documentary based on a co-production research project on resistance to slavery is an important ethical step in critically reviewing history practices, making the web documentary more than just an argument about the past [Dorn 2013]. Co-production in public history is about making history come alive, thereby forcing us to actively remember the past rather than existing in a mode of passive consumption, which in turn produces social awareness that could lead to social change, especially with the promises of digital virality. At the same time, socio-political contexts of reception change and electronic data remains ephemeral, even when in the case of a web documentary it cannot be erased or modified by users.

In September 2019, a mobile phone video from a village in western Mali showing a man being beaten and tied up started circulating on social media in Mali.[3] The man had been humiliated by a village mob for having dared to oppose descent-based slavery practised in western Mali. Since then, similar scenes have been repeated in villages across the region. The violence culminated in several horrific murders, all taking place in western Mali — in September 2020 in Diandioume, in September 2021 in Bafoulabé, and again in July 2022 in Lany Tounka, where Diogou Sidibé, a 71-year-old grandmother, was hacked to death with a machete on her farmland for having refused to be called a slave. Yet, until the latter murder, the Malian state and global community remained largely silent, despite the important demonstrations organised against slavery in Kayes city [Camara et al. 2021a]. In August 2022, eighteen persons were arrested following Diogou Sidibé's murder. In February 2023, the Kayes Criminal Court sentenced seven people to the death penalty and one to five years in prison for crimes linked to the practice of descent-based slavery in western Mali, notably those crimes committed in Diandioume in September 2020.[4] They were convicted of criminal conspiracy, murder, looting, assault and battery, and ethnic discrimination. Yet, as recently as May 2024, another man fell victim to descent-based slavery: Oumar Konate from Nematoulaye in western Mali died following a violent dispute with the village's enslavers on the grounds that he had refused to accept his ascribed “slave status”.[5]

Against the backdrop of this unsettling socio-political context of exacerbated violence against populations with ascribed “slave status”, naming slavery and making this, until recently, “invisible” history henceforth “visible” obviously had new political ramifications that needed to be re-evaluated, including the potential risks for the villagers we had involved in our project. As the project's principal investigator and co-investigator, we were indeed accountable for the conditions under which the villagers were to become publicly visible through the research process in general and the web documentary in particular [Dickens and Butcher 2016]. This changing socio-political situation posed complex ethical dilemmas. Acknowledging this changing context, we were torn between, on the one hand, our responsibilities as researchers and practitioners to make the complex legacies of slavery recognised and the long history of resistance known, and, on the other hand, the need to avoid causing any harm to the villagers — especially as the violence had not stopped at the time the web documentary was being produced (June–November 2020).

Ganbanaaxun Fede initially began in 2016 among the Mauritanian Soninke diaspora in France. The group was formed in response to audio recordings that had circulated on cassette tapes since 2005 and were later posted on YouTube. These recordings featured a Mecca-based Soninke Islamic preacher justifying slavery using hadith literature.[6] The movement, using primarily the social network WhatsApp, soon began to involve many of the West African Soninke diaspora — including the Malian diaspora, as well as a growing number of militants based in West Africa — in exchanging thousands of audio and video recordings a month, with a virality effect never achieved before on such an issue. WhatsApp also facilitated the viral circulation of mobile videos, photographs, and audio recordings of exactions committed against Gambana militants since 2018.

As the saying “a picture is worth a thousand words” seems to imply, the use of audio-visual media often makes social messages far more effective and appealing, including those concerning historical narratives [Salvatori 2022] [Theibault 2013]. The power of remediation of information into audio-visual narratives such as those in documentary films can sometimes be so effective that viewers tend to forget the medium (due to its “transparent immediacy”, per [Bolter and Grusin 1999]) and fall under the illusion that they directly access and experience unmediated raw information. The (moving) image is usually associated with “truth”, as the camera has supposedly captured things as they happen, resulting in the illusion for the viewer of almost having witnessed for themselves what has been recorded, even though this presence is no longer there [Spence and Navarro 2011]. A web documentary allows one to draw attention to the materiality of oral sources, making more transparent the oral history process. Although all representation or remediation is partial transformation [Spence and Navarro 2011], the illusion of the “authenticity” of visual media is probably where lies its greater power, in comparison with writing, in influencing understandings and conceptions about specific topics — and even more so in the digital era when videos can be easily shared on social media. Because the Gambana have been able to widely circulate on social media video evidence of the exactions that their militants experienced, it alerted the international media and increasingly pressured the Malian government to ultimately acknowledge that slavery practices did exist in the country.

It is in such a context of unprecedented “abundance” of audio-visual material related to slavery in Mali that the production of the web documentary and the related risks must be situated [Fickers 2012] [Rose 2014]. In 2020, the Gambana movement members, who fought for the recognition that slavery still exists in Mali, were paradoxically perceived as being the perpetrators of violence (which they were actually victims of) and of disrupting a purported peaceful social order [Camara et al. 2021b]. Even in Bouillagui and despite their history of resistance to slavery, some villagers were reluctant by then to be associated with the movement.

As highlighted by [Cahill 2007], one must ask oneself: What is safe to share in this context? Who is at risk of harm? Should we edit out some of the research because the context of reception might not be ripe for it? How can we contextualise our research and make it still relevant for the present when the present changes? In other words, how do we present our research so that it is “favourably received”, impacts people, and facilitates change, while carefully assessing the risks for the concerned community? How can we name slavery and resistance while preserving our ethical commitment towards the Bouillagui community in (re)presenting their history? What about the risk of the narrative being used or misinterpreted by certain sectors of the public? Once the web documentary has been made public, it is a story that will potentially define Bouillagui villagers for many years to come.

Yet, for the villagers involved it seems that the “ethics of recognition” offset the risk of being wrongly perceived by the consumers of the web documentary [Dickens and Butcher 2016]. They had won their own battle against slavery a long time ago and wanted it to be widely known and publicly acknowledged. The greater risk for Bouillagui so far had been for their history to be continuously invisibilised and their populations to be unrecognised as full citizens. Despite, or, rather, because of Bouillagui's history of fierce struggle for freedom, the socio-political environment in the region remained hostile to and discriminatory against them, while the slavery ideology continued be upheld in many villages around them. The founders of Bouillagui had already entered the struggle for recognition more than one hundred years ago. In a sense, this digital public history project was just an additional step in a longer history that they had already decided for themselves when they rebelled against their former enslavers in 1914. Here, the risk of naming and public visibility appears necessary and potentially beneficial in the fight for recognition through the gain of increased respect and solidarity in the Malian public sphere and beyond [Dickens and Butcher 2016]. During the preview screening in Bouillagui in November 2020, Bouillagui's notables praised the initiative to produce this web documentary, which they believed would teach the younger generations a great deal in terms of citizenship, enable Bouillagui's children to get to know their village and, above all, the history of its foundation, and make Bouillagui known throughout the world.

Our position as recognised researchers and practitioners has also helped to mitigate the risk of misrecognition, as the shared authority fostered a safe space for and an alliance with the Bouillagui community. We nonetheless attempted a compromise, in that we committed to revisit the documentary as the situation for the Gambana movement in Mali evolves. In one of the videos of the web documentary, Rodet's voice-over mentions the context of increased violence against people with ascribed “slave status” and Gambana activists in Mali and how it affected the research. Yet this specific video was not shown at the first official screenings in Mali. This was suggested by Mamadou Séne Cissé, who was in charge of the web documentary screenings in Mali and who assessed that the situation at the time with the Gambana movement in the region was far too tense to run the risk of having its name mentioned at a public screening. All this took place against the backdrop of a very unstable political situation, with increased surveillance and arrest of “dissident” voices, alongside systematic social media suspensions and the diffusion of fake news to control Malian public opinion. Since then, Gambana's public perception as “troublemakers” flooding WhatsApp with “irresponsible”, “anti-social” discourses has nonetheless changed markedly. They have been recognised by the Malian authorities as a legitimate partner in the fight against slavery. We henceforth play the full version of the web documentary at public screenings.

Such a situation clearly illustrates how research is an ethical process and digital ethics a moving target: conflicting narratives can arise as research unfolds, and the end-user's environment may unpredictably change despite regular risk assessment in the course of the project [Ess 2012]. In the same way, the public for such a web documentary itself can also change. Although we produced this web documentary with specific audiences in mind, as discussed previously, until the web documentary is released it is difficult to predict which audiences will really access and be sensitive to it. Even then, because it is web-based it allows access by a wide range of publics, heterogenous in terms of class and economic status and political commitments — although, as mentioned earlier, the digital gap crucially affects the web documentary's effective accessibility by all publics, most notably in Mali. With digital content, the “politics of reception” is often unpredictable, with no guarantee that the content will travel well not only within, across, and between cultures, but also over time [MacEntee 2016].

This also raises questions about the future of such digital public history projects. On one side, there is increasing funding pressure for open data — for example, amongst UKRI (UK Research and Innovation)-funded projects in the United Kingdom, since 2023. On the other side, there is only open-access funding for published papers. There is no funding for alternative digital publication platforms or any research website or digital collection beyond the lifetime of the funding project. Funding their continued maintenance often requires applying to craete new research projects that would contribute to enhance parts of the original. We managed to get the web documentary hosted on the SOAS server with the hope that this would secure its future, but as universities increasingly sub-contract their data storage to expensive GAFAM “giants” (Google, Amazon, Facebook, Microsoft), while experiencing public budget cuts, this becomes an issue of cost-effectiveness, where the university decides which data to maintain, which to store in the cloud, and which to delete — decisions concerning which the researcher is not always consulted.

Furthermore, an online transmedia project such as a web documentary remains fixed in time; it is not meant to be modified, though the end-users might live in different political, cultural, and social environments that are fast-changing. Leaving aside the illusion that the digital will allow us to record and retain everything, we should be mindful as historians that electronic data and the cloud actually require physical storage space and can thus disappear, just as historical evidence and archives themselves regularly do. If a digital public history website becomes no longer fit for purpose, as digital ethics also evolve — while the technology to maintain data can easily become obsolete (our web documentary uses Klynt, an online software to create and publish interactive narratives for the web)[7] — we should accept letting it go and take the webpage or website down. We could also consider archiving it with public digital archives, such as with UK data archives.

With the aim of engaging the past in the present, our digital history project took two main risks at its outset: conducting a public history project on the still-taboo topic of resistance to slavery in Mali; and betting on research co-production and shared authority with Bouillagui's villagers through participatory visual methods, despite the digital gap. As the socio-political context in Mali and also worldwide rapidly changed over the two years of the project, we did not anticipate all the project's twists and turns. Despite some clear pitfalls, this was an important step in realising that just as research is a process, so too are ethics: digital history and storytelling require responsibility toward co-producers and audiences, not only during the research process but also with the online publication, especially when working with communities who run the risk of sharing their history of resistance to slavery in contexts that have never properly allowed for these voices to be heard. It is through such an ethical process, an ethics of care that relies on long-term engagement and commitment, that the transformative potential of a digital public history project naming slavery can reveal itself. The greater risk for the populations concerned may not actually lie in telling and publicising such history but in allowing the hegemonic discourses to continue to silence them. As highlighted by Gardner, engaging in public history is about taking risks and dealing with unsafe ideas, even in unsafe times, but it is also a risk we should commit to, especially when we have been entrusted by the concerned communities to name slavery and publicise their history of resistance to it [Gardner 2010].