This article explores how the ad hoc filing of research reports led to the creation of a human rights archive on Africa, with implications for future accessibility and relevance.

By the time Amnesty International received the Nobel Peace Prize in 1977, it was well-established as an international membership organization with a narrow but deep mandate to release prisoners of conscience (POCs), end torture and abolish the death penalty worldwide. Its base, mostly grouped in the global North, consisted of individual members and local groups, coordinated by national Sections and serviced with information from Amnesty’s International Secretariat (IS) in London. Numerous adoption groups still adhered to a ‘rule of three’, taking on cases from East, West and non-aligned countries, but other geographical disparities remained; among these, the IS Research Department recognized that its African casework was overwhelmingly anglophone in character. In September 1977, Amnesty’s governing International Executive Committee decided to increase its francophone research capacity by setting up a ‘Research Attaché’ post in Paris which, then as now, was a crossroads for information from passing visitors as well as established exiled communities, and a concentration of press agencies, academic and publishing centres.

In those days, the work was often referred to as a ‘listening post’, though this description implied a purely passive role that was soon overtaken by the need to follow up on a multitude of sources. This early work, conducted from an attic room in the rue de Varennes offices of Amnesty’s French section, developed on an ad hoc basis, according to opportunities and new contacts that arose. There are some twenty African countries with French as their official language. Djibouti and the Comoros were the first to present themselves; others quickly followed, including Madagascar (Indian Ocean), Burundi (Great Lakes), and the Maghreb countries of North Africa, in addition to the former French territories of West and Central Africa. Some lusophone countries also proved to be ‘francophone’, with sources from Guinea-Bissau, São Tomé and the Cabinda province of Angola all using French to some degree, many as a result of years spent by contacts in exile.

Research on a country would habitually open up after outside sources were contacted by the Paris Research Office (PRO), either following press reports on events in the country concerned, or through existing intermediaries. For a few countries, this process was inhibited by historic suspicions of Amnesty as a ‘Western’ or indeed French NGO; this was the case for Cameroun, for example, where one opposition leader had developed a strong though clearly unfounded suspicion of Amnesty France. It was thus important that the PRO researcher, and from 1986 her assistant, be known to be IS staff, independent of the national Section. In all cases, trust in the debriefing process had to be built and earned by demonstrating that Amnesty could get results, while respecting the confidentiality of contacts who often went in fear of their lives. Moreover, there was a felt need to assist the many informants whose asylum claims were pending, often for years, with referrals through Amnesty’s French section to NGOs assisting refugees and a centre treating torture victims.

A year after the Paris Office opened, the post was upgraded from research assistant to researcher, while continuing to supply raw research material to Amnesty’s researchers in London. This was shortly after the infamous ‘Bokassa affair’ in April–May 1979, which marked a turning-point for Amnesty’s French section as well as for the PRO. A short press statement by Amnesty about the beatings and deaths in custody of dozens of school students in what was then called the Central African Empire prompted a massive response in the French press and led ultimately to the deposition of the self-styled ‘Emperor’ Bokassa by the French state that had helped create him (Power Reference Power2001; Blanchard Reference Blanchard2020). This scoop helped to promote Amnesty’s name and work throughout francophone Africa, at a time when few international bodies showed much interest in monitoring or intervening on human rights issues there.

But while African issues rarely made the international headlines, Amnesty’s work was usually focused on such local human rights issues, on the assumption that every victim of human rights abuse was worthy of concern. Widespread arrests, disappearances, torture and killings might be ongoing for years without ever surfacing in the mainstream international media. Small countries like Benin had large numbers of imprisoned students – Catholics and Communists alike – detained without benefit of due legal process. Larger countries like Mali could deport dissident teachers to its far North,Footnote ¹ and former ministers to work in salt mines in the depths of the Sahara without being noticed by the Paris-Dakar Rally (Garling Reference Garling1989). Sékou Touré’s Guinea held dozens of purported enemies of the state in squalid jails without hope of pardon; many were starved to death when put on a ‘black diet’ by presidential fiat. Throughout francophone Africa, many people – mostly men – were arbitrarily detained in local lock-ups, often at the whim of a local potentate, and with little or no recourse in law.

It was often these victims’ families who were obliged to bring them food and pay when they could for small mercies (corruption had its uses). These might take the form of letters smuggled out of prison and despatched by an intermediary (priests, medics, released cellmates or others) to their Amnesty adoption group or directly to the PRO and thence to London. Such letters, as with the first-hand accounts by released prisoners, provided vital information about the real state of repression and prison conditions in each country – information not covered by the local media. They formed the raw material of Amnesty’s work, alongside the testimony afforded by post-detention debriefings, or the rare prison visits during country missions where these were allowed. Sifting through the huge quantities of such raw information – what the French called ‘un travail de fourmi’ or working like an ant – was time-consuming but very rewarding. It all needed assessing for its veracity and comparing with the output of secondary sources – key informants, church and medical workers, journalists, academics and others, situating in context and, at the end of the day, writing up for onward transmission to London. There, the broader picture for each country was being worked on in similar fashion, resulting in campaigns conducted via Amnesty’s many national sections and groups worldwide, or in more pressing instances urgent interventions by letter, telex, fax and (latterly) email, combined with press releases, international reports and direct appeals to influential personalities likely to influence events. These campaigns, like the urgent actions focused on just a few local people, would in turn help reinforce the trust built up among primary sources when they saw that perseverance could bear fruit, prisoners be released, torture halted, and death sentences suspended.

However, those with information to impart often expressed frustration at the lack of media attention. Their press conferences would habitually attract only their own supporters (plus the PRO researcher and a man from the DGSE (Direction générale de la sécurité extérieure, French intelligence service). While the Bokassa Affair had snowballed because President Giscard d’Estaing himself had backed the self-styled Emperor, only exceptional events – a coup or a French hostage – might justify mainstream French or international press coverage. Other local stories were regularly spiked by editors (one news source would post discarded agency tapes in our letterbox following the night shift). Otherwise, what went on in the countries known as ‘la Françafrique’ were largely confined to business news and official visits. Local media were usually controlled by those in power and any opposition voices raised at their peril. With a few exceptions, before the national conventions and nascent multi-party politics of the early 1990s, ‘Big Man’ systems of patriarchal power ruled over cowed and subjugated populations, whose main recourse to uncontrolled local information would be ‘radio-trottoir’, or popular rumour (Ellis Reference Ellis and Kelsall2022).Footnote ² The need to sift and evaluate incoming information made the work of the Paris Office all the more vital.

With the benefit of hindsight, the record of these times afforded by the PRO’s work covering forty years provides an insight into the unspoken reality of people’s lives under various dictatorships across Africa and thus forms an important counterweight to the official versions of their history. While its first thirteen years (1978–91) cover the pre-multi-party period, the PRO continued its work for a further twenty-seven years, with two researchers and support staff now more fully integrated into the IS’s work, i.e. able to undertake missions and initiate actions in their own right, but on West Africa alone. This also meant, however, fewer written reports to inform their London counterparts, and less available time for interviewing local Paris-based sources.

Amnesty’s Paris Research Office was closed in 2018, following the untimely death of its last researcher, Gaëtan Mootoo.Footnote ³ In the intervening years, Amnesty International as a membership movement had changed drastically, widening its mandate far beyond that of its early years and adopting research and campaign methods more in line with the public ‘naming and shaming’ used by other major human rights groups, and the requirements of instant online communication and social media (Girot Reference Girot2011). The extensive archives accumulated over forty years were stored at the French Section premises in Paris, until a decision was taken by the IS in 2022 to repatriate them to London. This removal took place in 2023, with the intention of digitalizing the entire content, while the original research materialsFootnote ⁴ remain part of the IS archive and are due to be stored at Amnesty’s external storage facility. Some sixty boxes of secondary newsprint remain in Paris in search of a repository.

The matter of access to these archives, and to the broader Amnesty research archives with which they are now integrated, begs a number of questions. The first of these concerns the strict confidentiality promised to primary sources who were willing to speak to the PRO researchers for the duration of its work. Some of these sources spoke in fear of their return home or for their family members remaining in their home country; some others went on to hold high office, including half a dozen future heads of government. Over time, however, it is safe to say that those survivors and their family members have a right to know what they had said, what was said about them, and by whom, since such information would have been crucial to the eventual resolution of their plight; moreover, they might not be aware of those Amnesty groups and individuals who had intervened on their (or their relatives’) behalf. At a wider level, others caught up in the same stream of events – as political detainees, witnesses, activists, journalists, academics and others – could also claim an interest in accessing such information.Footnote ⁵ While there is always a need to balance confidentiality against access, over time the former need will necessarily diminish. There is a good case to be made for the initial period of PRO activity, running from 1978 until the reformulation of its work in the early 1990s, to be opened to limited public scrutiny, at least for the categories of interested party listed above. This would correspond to a general ‘thirty-year rule’ such as now applies even to French colonial archives.

The corollary to such access would, in the author’s view, consist of a proactive role for Amnesty’s archivists in both vetting access and in ensuring that all likely interested parties were aware of their option to view such material, and of their responsibility towards it. This role could also involve networking with other archives in Britain, France, Africa and elsewhere including, if need be, official government records. Such work would greatly enhance the chances for local African writers and historians to recover parts of their own country’s immediate past that might not otherwise be available (Guimendego Reference Guimendego2015).