When the Malian city of Timbuktu fell to Islamist rebels last April, the destruction of its immense cultural treasures did not begin at once. It wasn’t even clear, at first, who exactly was in charge of the city, or what they might do.
But by July, when a coalition of new leaders consolidated control and began tearing down some of the city’s most important burial shrines, the heads of Timbuktu’s libraries had already been working quietly for months to move and hide their collections—one of the most significant troves of unique medieval manuscripts in the world. And among the first materials that they smuggled out were something surprising: hard drives.
These hard drives were packed with digital backups of thousands of precious library documents. Though today Timbuktu is a remote and dusty city of 54,000 at the edge of the Sahara, 500 years ago it was a major commercial crossroads and a great center for
scholarship. Copied onto those disks were scanned versions of some of the world’s most important surviving medieval manuscripts, texts on Islam, politics, math, and science that illuminate the city’s past as a center of learning.
In the last decade, Timbuktu’s libraries have been working with partners in the United States, South Africa, and France to create digital archives of its crumbling documents. Expensive scanners and digital cameras have been ferried up the slow Niger River or along the long road to the isolated city to capture these texts electronically and in some cases to post them online, making them widely available to scholars for the first time.
In the broad recent push to digitize libraries around the world, most of the rationale has been about such access. The team behind the Digital Public Library of America project, for example, which includes Harvard University and the Library of Congress, says its goal is to “make the cultural and scientific heritage of humanity available, free of charge, to all.” A similar project at Google Books aims to digitize all the world’s estimated 130 million books by 2020. But when Timbuktu became the target of jihadist rebels, the urgent efforts to document and save its manuscripts made it an important test case for a different role for digitization entirely: to make backup copies of unique, endangered collections that otherwise might be lost to us completely.
Turning to digital copies as a tool for cultural preservation carries risks, and raises its own questions: What do you prioritize? How do you handle the expense and potential damage to priceless old documents? What about the fact that these digital files, too, are unstable?
But what happened in Mali this year suggests exactly how high the stakes can be. Last month, as the rebels retreated from the city, they stormed into one of the city’s largest libraries, the state-sponsored Ahmed Baba Institute, and burned everything they could find. For a few days, before the institute and other libraries revealed just how many original manuscripts they had hidden away, it seemed possible that whatever digital copies had been spirited out or placed online would be the only versions we had left.
In its heyday in the 15th and 16th centuries, when trade in salt and gold was thriving and its great mosque made it a religious center, Timbuktu was populated by scholars from across the Muslim world, who collected and wrote books on the classical texts of Islam and on contemporary questions of medicine, astronomy, math, politics, and more. After a Moroccan force invaded in 1591, the city’s intellectual life faded. But the manuscripts survived, handed down through families for centuries and hidden whenever the political situation turned dark.
Over the past four decades, those documents have begun to be consolidated into bigger collections that have attracted the attention of the outside world. Since 2000, Timbuktu’s libraries have received funding from groups including the government of South Africa—whose then-president, Thabo Mbeki, called the documents “the most important cultural treasures in Africa”—and from Western institutions such as the Ford Foundation. Often, this outside funding for conservation included provisions for making digital copies.
It was a slow process, to say the least. The desert light is too bright for easy photography, and the sand can sneak into delicate equipment. Getting the high-resolution cameras or scanners to the city is its own challenge. These hassles meant that the cost of digitizing a single page might run upward of $25, according to one estimate.
Despite these difficulties, a handful of groups began scanning these manuscripts and putting them online, at first in small batches, then by the thousands. The Tombouctou Manuscript Project, based in South Africa, has a few dozen manuscripts digitized; a US-based project called Aluka (which later merged with the digital library JSTOR) has image files for 300; the Ahmed Baba Institute has a catalog of 9,000 available digitally. A project based at the Institut National des Sciences Appliquées in Lyon, France, which began in 2008, had scanned some thousands before it was forced to cease operations last March, when the rebels closed in. Altogether, more than 30,000 manuscripts from public and private libraries have been digitized, one researcher estimated.
In a world where we’re told to back up every bit of work to a hard drive or to the cloud, it may seem obvious that making electronic backups of fragile paper is smart. But if it took a while for archives like those in Timbuktu to start scanning, that’s because digitization actually isn’t an obvious conservation tactic. Archivists and scholars do like the idea that digital documents can potentially be studied anywhere, by anybody—opening up knowledge beyond the handful of people who can travel to see a manuscript in person. But when it comes to preservation, bits are an ephemeral medium for storing information, and digital documents require careful tending, if they’re not to become locked up in obsolete storage technologies. Many archivists still prefer microfilm: It’s more stable.
Furthermore, whatever the medium, there are inherent risks to scanning or photographing fragile documents. “The process of digitization could be the most aggressive handling that the manuscript has undergone in hundreds of years,” says Simon Tanner, deputy head of the digital humanities department at King’s College London, who helped Aluka with its work in Timbuktu. A small error can damage the original forever: The wrong lighting, for example, can heat up a document and shrink it. “Manuscripts tend to be very robust and live very well as long as they’re in a good environment,” says Tanner. The Dead Sea Scrolls, for instance, which he also helped to digitize, “look like my shoe leather, but they did very well for thousands of years in a cave,” he says. “They’ve deteriorated faster out of the cave than in the cave.”
Some curators, like William Noel, the director of the University of Pennsylvania’s Special Collections Center, point out that simply making digital files available serves an important conservation function: It gives many more people a stake in preserving that information and the documents that contain it. “It is only if they are really useful that they themselves are going to survive,” Noel wrote in an e-mail.
For most of Timbuktu’s manuscripts, the digital backups ended up being just that—backups. But for some documents, those digital copies are now the only versions remaining.
The city’s libraries were able to quietly move thousands of manuscripts out of the city or hide them in still secret spots, but the rebels who set fire to the Ahmed Baba Institute are believed to have destroyed about 2,000 manuscripts—a fraction of what could have been destroyed, but still a significant loss.
Some of those texts may be lost forever, but copies of many of them, researchers believe, are on the hard disks they moved out of the city. In interviews, several people who work with the Timbuktu libraries, both inside Mali and from abroad, confirmed that the hard drives had made it out of the city and to Bamako, Mali’s capital, early on, perhaps even before rebels arrived. (Because of continued instability in Mali, they were reluctant to give specifics on exactly when or how.)
The Timbuktu effort now has counterparts elsewhere, working to preserve endangered or politically volatile material. UCLA just began a project to capture ephemeral media, both paper and digital, that document today’s political life in the Middle East. In the past two years, Yale University Press digitized Stalin’s personal archive, which is among its holdings. The University of Texas’s human rights project is working in Burma to digitize threatened records of human rights struggles. Last year, the British Library’s Endangered Archives Programme funded projects to digitize monastic archives in Ethiopia and Mongolia, East India Company records in St. Helena, public records in Northern Ghana, and manuscripts in East Jerusalem’s Al-Aqsa mosque.
As this huge range of material suggests, it’s not practical to digitize everything, and what to prioritize is a significant choice unto itself. Michele Pickover, principal curator at South Africa’s University of the Witwatersrand, has written that archivists inevitably end up making decisions that determine how history will be understood. While South Africa provided financial support for digitizing Timbuktu’s manuscripts, for instance, only a select portion of its own African National Congress documents have been digitized.
There are questions, too, about who should own digital copies and where they should live: In some cases, digital versions of manuscripts have gone behind paywalls unaffordable to researchers in the country where the original manuscripts are kept. Even the decision to make documents available online and to whom can be problematic. While advocates like Noel and Tanner argue that digitization can increase access and interest in a collection, not all those who run archives are convinced of that relationship. Some researchers in Mali “don’t want to give these images away, because they want people to keep coming to Timbuktu,“ says Shamil Jeppie, the director of the Tombouctou Manuscripts Project.
For the moment, though, those who would willingly travel to Timbuktu to study at its archives must bide their time: The libraries are keeping their manuscripts safely stowed away. Until the political turmoil calms, the only way to see even a portion of those documents is online.
Sarah Laskow is a freelance writer and editor in New York City. She edits Smithsonian’s SmartNews blog and has contributed to Salon, Good, The American Prospect, Bloomberg News, and other publications.