The trunks began as oil drums. Metalworkers in Mopti hammered them into chests, and couriers used them to carry manuscripts away from Timbuktu during the occupation of 2012–13. Inside were works of astronomy, poetry, history and law. Saving a written tradition required something as ordinary as a strong box, followed by someone willing to take it through a dangerous journey.[13]
Abdel Kader Haidara helped organise the movement through a network of library owners and custodians. Some collections first went to hiding places within Timbuktu; many later travelled south to Bamako. The chests protected their contents during escape, but the packed books faced another danger in the capital’s more humid air. The work changed from moving manuscripts to keeping them safe.[13][14]
Long before those journeys, families had gathered, copied and inherited the pages. Their libraries held arguments about inheritance, instructions for calculation, religious learning and records of transactions. Those particular uses help explain why custodians took such risks: the trunks held the work of people whose names, questions and obligations could still be read.[1][5]
Books belonged to the trading town
Leo Africanus described a market for books in his sixteenth-century account of Timbuktu. Manuscripts brought from the north, he wrote, could be more profitable than other merchandise. They travelled within a trading world linking the Sahara and Sahel, where acquiring knowledge and acquiring goods could bring people to the same town.[6]
Timbuktu developed beside the Niger’s wider trading geography, with caravan routes reaching the desert and a river port at Kabara. Gold and salt supplied familiar parts of its commercial history. Written learning supplied another. By at least the fourteenth century, important works were being written and copied in the town, whose scholars had connections across the Islamic world.[2][4]
The mosques of Djinguereber, Sidi Yahya and Sankoré belonged to that scholarly setting. Learning also entered family collections, where a text could be copied, studied and passed on. The word “library” here includes collections kept within households, rather than one institution containing everything the town knew.[2][11]
Ahmad Baba’s life crossed the region’s political upheaval. Born in 1556 into a scholarly family, he studied with his father and Mohammed Bagayogo before becoming a prolific writer. Moroccan authorities deported him after the conquest of the Songhai realm. Accounts differ over details of his exile and return, but his later work belongs to the intellectual world the surviving manuscripts preserve.[3][11]
One of those works answers a Moroccan correspondent’s questions about slavery. The Library of Congress exhibited it in 2003, beside other manuscripts from Timbuktu-area collections. A scholar remembered in the name of the Ahmed Baba Institute is also present through a specific argument on a specific page.[5][11]
The text considers slavery within Islamic law and begins from the position that human beings are originally free. That does not remove slavery from the society in which the argument was written. It preserves a scholar’s reasoning within it, available for examination alongside the transactions and acts of emancipation recorded elsewhere in the collections.[5]
The collector who had planned a different life
Abdel Kader Haidara had intended to go into business. After his father’s death in 1981, he was asked to work as a manuscript collector for the Ahmed Baba Institute. He was still a teenager. The new work took him into the difficult business of persuading families to entrust him with books that had belonged to them for generations.[13]
His father, Mohammed “Mamma” Haidara, had gathered manuscripts through his own scholarly travels. The family’s name later became attached to its memorial library. The son’s collection work therefore grew from a household history as well as an institutional appointment. A manuscript could be a research object to an archive and an inheritance to the person opening the door.[13]
The state archive was comparatively recent. Established in the 1970s, the Ahmed Baba Centre brought public collecting and conservation into a much older world of private libraries. Later cooperation with South Africa supported training for Malian conservators and a new building. Preserving the collections involved relationships between owners, local expertise and outside institutions.[6]
The University of Cape Town’s Tombouctou Manuscripts Project, led by Shamil Jeppie, grew from that cooperation. Its account names the Ahmed Baba Institute alongside the Mamma Haïdara Memorial Library, Fondo Kati and al-Wangari collections. The names matter because these are different holdings, with different custodians, even when international reports gather them under “the Timbuktu manuscripts”.[reference needed]
Haidara’s later network, SAVAMA-DCI, brought private library owners into a shared preservation effort. That network became central to evacuation. The trust needed to move a chest during a crisis had roots in quieter years of collecting, negotiating and caring for books.[13][21]
An inheritance calculation beside a drawing of the heavens
A volume of commercial law could use verse to help students remember their obligations. An inheritance commentary explained who should receive what after a death. An astronomy manuscript copied in 1733 carried a diagram of the heavens. These were among the objects lent to the Library of Congress, where their subject matter unsettled any simple division between sacred books and practical knowledge.[5]
The collection included medicine and religious remedies, arithmetic, legal reasoning and records of commerce. A certificate freeing an enslaved woman stood beside evidence of the slave trade. The surviving papers do not deliver one flattering account of the past. They preserve different actions and arguments, including evidence of both bondage and emancipation.[5]
The exhibition drew on the Mamma Haidara collection and the library of Cheick Zayni Baye in Boujbeha. Those loans made a small selection available abroad, while the larger collections remained with their owners. A public display could show the range without becoming a complete inventory of what families held.[4][5]
Arabic connects much of this scholarship with a wider written world. Ajami adds another relationship: African languages written with Arabic script. Fallou Ngom’s study of West African manuscripts describes these traditions and the knowledge still held in collections that are little catalogued or studied. The script on a page does not, by itself, tell us which language its words belong to.[21]
Copying also gave a work another material life. An author’s argument could travel through later manuscripts, each made from particular paper and ink. A conservator studying the surviving object must therefore attend to its physical history as well as the words it transmits. That attention leads back to the people who knew how to make the ink.[20]
The ink recipes were still spoken
When Oliver Hahn and Eva Brozowsky investigated Malian writing materials, their team needed reference inks whose ingredients were known. Written recipes were not available to them. They interviewed marabouts and scribes in Djenné instead, gathering the practical knowledge needed to make comparison samples. The study began with people who still knew the craft.[20]
Their 2015 report describes recipes using plant parts, gums and minerals. Bark from Vitex chrysocarpa, called koronifin, appeared alongside indigo, hibiscus and other materials. Gum helped bind ingredients; laterite supplied a mineral red. A manuscript’s colour could carry products of local vegetation and ground into the work of a scribe.[20]
Paper followed longer trading routes. Mary Minicka’s account describes European paper made for Islamic markets, including watermarks acceptable to buyers there. The page and its writing material could therefore bring imported paper together with locally prepared ink. The physical book held connections that extended beyond its text.[6][20]
Under the microscope, the research team distinguished different kinds of black ink. In their reference work, iron-gall ink showed a brown halo around the stroke, unlike the comparison carbon and plant inks. Other analytical methods helped identify materials without taking large samples from fragile objects. Knowing the mixture could guide decisions about treatment.[20]
That is a practical question because ink can damage its own support. Corrosion associated with some ink compositions can weaken the paper carrying the writing. Preservation cannot be reduced to protecting a book from fire or theft; a harmful process may already be at work along the line of a letter.[20]
The chests leave the libraries
Armed groups took Timbuktu in April 2012. During the occupation, people faced restrictions and violence, and historic religious sites were attacked. The destruction of mausoleums threatened part of the town’s Islamic heritage alongside the danger to its books. The crisis reached into worship, education and ordinary life.[2][11]
The Ahmed Baba Institute’s new building became quarters for fighters. Haidara went south to Bamako, but the manuscripts’ journey required work within Timbuktu too. His network organised movement into safe houses before the longer evacuation. Couriers carried trunks in small loads, rather than gathering a conspicuous convoy of libraries.[10][13]
Joshua Hammer’s account describes the oil drums made into chests at Mopti and repeated journeys between the autumn of 2012 and the following January. Each trunk was only part of a collection. The scale of the rescue came from repeating a manageable movement, with couriers, owners and helpers making it possible across different stages of the route.[13]
The Prince Claus Fund helped finance the emergency work. Stephanie Diakité worked with Haidara in the rescue network. Their contribution joined the efforts of custodians whose names did not all enter international accounts, but whose choices let the texts leave a house, cross town or continue south.[15][1]
French and Malian troops entered Timbuktu on 28 January 2013, after the armed groups withdrew. Reports of destruction quickly travelled beyond the city. The empty spaces left by evacuation were easily mistaken for proof that everything once kept there had been burned.[22][8]
What was missing from the burning-library story
Alex Crawford’s report from the new Ahmed Baba building described more than 25,000 manuscripts burned or missing. Shamil Jeppie and colleagues in Cape Town were trying to reach people connected with the collections. An interrupted communications network made that difficult. The first international accounts formed before the custodians’ fuller information could catch up.[8][9]
Mohamed Diagayeté explained that much of the institute’s collection had been kept in its older building, across town. Other manuscripts had been hidden or moved. The new building’s empty cabinets did not account for every item in the institute, still less every family library. Reports collected by Foreign Policy also described digital material retrieved from the occupied premises.[8][10]
Loss was real. UNESCO’s June 2013 mission estimated that 4,203 manuscripts from the Ahmed Baba research centre had been lost, while another 300,000 had been taken out, mainly towards Bamako. These figures describe different holdings and outcomes. They cannot establish one survival percentage for all manuscripts owned throughout Timbuktu.[23]
Other contemporary accounts used different numbers. The World reported a count combining manuscripts burned or stolen at the institute; HMML’s December account offered a lower estimate of destruction. The differences should remain attached to the institutions and dates that produced them. Turning them into a simple range obscures what each organisation was trying to count.[12][14]
The rescue story also has several scales. Hammer’s account follows the evacuation of many private libraries; the institute’s holdings have their own history. A manuscript hidden in Timbuktu, one taken to Bamako and one later photographed are different events in a collection’s life. None can substitute for a complete census of the town’s written heritage.[13][8]
In Bamako, the boxes had to open
The capital’s damp air made the packed metal chests a new problem for the people caring for the books. In late 2013, HMML warned about mould and mildew threatening manuscripts stored in lockers. The boxes that had helped their escape were not a permanent archive. Paper needed a different environment after the journey.[14]
The University of Hamburg’s Centre for the Study of Manuscript Cultures joined preservation work in 2013. Its project combined conservation with inventories and the preparation of collections for study. Later funding extended that work beyond the first emergency period. Getting books to safety had left years of quieter tasks.[17][18]
Eva Brozowsky’s training workshops made those tasks concrete. Surface dirt could be removed with soft brushes and special dry-cleaning sponges. Tears could be supported with lightweight Japanese paper and a prepared glue. Mould and insect residues required careful removal with tools including scalpels and aspirators. Each kind of damage called for its own repair.[19]
The aim was to halt deterioration and prolong an object’s life, rather than to make it look newly written. A repaired page should remain the surviving page, carrying its history as well as its words. That approach sets limits on repair: care is directed towards what can be preserved, not towards an imagined flawless original.[19]
Cleaned manuscripts needed suitable folders or boxes, with acid-free material replacing harmful packaging. Storage required regular maintenance, including cleaning shelves and controlling humidity. The key work could be as plain as keeping dirt from building up again. A successful rescue had to become a routine.[19]
A photograph becomes another way to read
Columba Stewart and Walid Mourad helped establish digitisation studios in Bamako in 2013, working with SAVAMA-DCI and local librarians. Training accompanied the equipment. The photographs would provide another route to the manuscripts’ contents, while the physical objects still needed care and stayed connected to their owners.[14]
HMML’s arrangements kept a copy of each digital set with the library owner and stored further copies elsewhere. The owner did not need to surrender the manuscript for a distant scholar to examine its pages. Digitisation could widen access while preserving that relationship, although describing and cataloguing the images remained substantial work.[15]
The work reached manuscripts that had stayed in Timbuktu too. Sophie Sarin helped develop a project with families whose collections survived there, and local technicians carried out photography. HMML’s later account identifies Badji Cissé at work on manuscripts of a Sankoré mosque library in 2019. The continuing story includes those doing the careful work beside the pages.[16]
In its 2024 account, HMML reported about 4.25 million image files representing 302,000 manuscripts across its Mali projects. Images are not manuscripts: one object can require many photographs. The totals also cover work in more than one city. They describe the scale of a completed photography programme, while cataloguing for access went on.[16]
In August 2025, the return of manuscripts to Timbuktu began. Associated Press reported a first shipment flown from Bamako, with more to follow. The report concerned the start of a return, rather than confirming that every rescued collection had already reached home. It changed the geography of preservation again, without bringing its work to an end.[24]
During a visit to a Bamako safe house in 2014, Hammer saw Haidara bring out a sixteenth-century work from the Sankoré library. It had survived occupation and evacuation, and its writing could still be read. The chests had done their urgent job. The page could be opened.[13]
Share what you know
Been there, grew up there, or know something we missed? Add a comment; corrections with a source are especially welcome.