Fire doesn’t heed history. It doesn’t care about posterity or culture or memory. Fire consumes everything and anything, even if that thing is the last of its kind. On Sunday night, it came for the National Museum of Brazil, burning for six hours and leaving behind ashes where there had been dinosaur fossils, the oldest human remains ever found in Brazil, and audio recordings and documents of indigenous languages. Many of those languages, already extinct, may now be lost forever.
It’s the kind of loss that’s almost impossible to quantify. For the researchers who worked in the museum, the conflagration sent their life’s work up in smoke.
“It is very difficult to react to reality and try to return to life,” linguist Bruna Franchetta, whose office burned down in the fire, told WIRED in an email. “At the moment we do not know the extent of the destruction of the Documentation Center of Indigenous Languages in the National Museum. We have to wait a long time for a survey of what is left in the middle of the rubble. At the moment I can say nothing about what has not turned to ashes, but I hear colleagues saying that it was all lost.”
It didn’t have to be this way. All of these artifacts could have been systematically backed up over the years with photographs, scans, audio files. The failure to do so speaks to a vital truth about the limits of technology: Just because the means to do something exists technologically doesn’t mean it will be done. And it underscores that the academic community has not yet fully embraced the importance of archiving importance of archiving—not just in Brazil, but around the world.
Though Franchetta says work had begun recently to digitalize the CELIN archive, she has no idea how far it had gotten, and it focused on only a part of the collection. “The loss is immense, and much of what has been destroyed by the flames can never be recovered,” she says.
In 2018, when an iPhone automatically backs up every photo you take, you might think knowledge is safer today than it was in the days of the Library at Alexandria. The fire in Brazil puts the lie to that assumption. To undertake the archiving of so vast a collection—the National Museum of Brazil reportedly lost 20 million artifacts in all—requires time, money, and a sense of urgency.
As museum staff and researchers try to pick up their lives, find a new offices to work in, and figure out how to continue their work, there’s plenty of blame to go around. Much of it belongs at the feet of the Brazilian government, which had slashed the budget for the National Museum and the University of Rio De Janeiro, which runs it. The museum was so strapped for money that last year, after termites destroyed a wooden base holding a 42-foot dinosaur skeleton, it started a crowdfunding campaign to raise $15,000 to replace it. The building had no sprinkler system. Government cuts are also why, when firefighters arrived to fight the flames Sunday night, they reportedly found no water in the hydrants, having instead to get water from a nearby lake.
All this austerity both made the fire more likely and made it burn more fiercely and longer than it needed to. Brazil’s cultural minister said that before the fire struck the museum, it had been poised to receive $5 million from the government for upgrades, including adding a fire suppression system.
But the lack of a backup archive goes beyond governments. Certainly funding played a huge part, but even scholars who spend their lives studying history and loss, researching how cultures end, can fall for the notion that there will always be more time.
“I think people just had the idea that, well it can be done someday, what’s the urgency?” says Andrew Nevins, a linguist affiliated with the National Museum. “The idea of digitizing as an urgent priority wasn’t in the air…Instead there was lots of funding and sources for going into the field and finding the last speakers right now of [a given language].” That’s obviously important work, but without a plan for how to safely back up and keep those records, much of it is now lost.
That loss is not merely to science, or to future museum visitors, but to those cultures who entrusted their histories to the museum. An estimated 500 indigenous tribes currently live in the Amazon, speaking around 330 languages, about 50 of which are estimated to be endangered—but before colonization there were likely as many as 2,000 tribes. The CELIN archive contained research into roughly 160 of these languages, estimates Franchetta.
Linguist Colleen Fitzgerald, who heads the United States’ National Science Foundation’s project on protecting endangered languages, notes that field work of the type that created the collection in Brazil involves deep collaboration with the communities being studied. Often over the course of many years, researchers gain access to people’s lives, stories, and customs. The responsibility to protect what they share is a solemn one.
“Brazil [does not have] a diffuse culture of file safeguarding, especially through scanning and storing in various backups and in different secure locations,” says Franchetta. She notes that their academic community rarely discusses best practices in how to create digital archives. Nevins agrees. Though students and professors labor to collect everything they can about endangered languages, he sees far less emphasis on protecting what they gather.
“The first reaction that many of us had was indignation: How could this happen, how could there be no sprinkler system? But I think as the dust started to settle there’s also some indignation at the state of library science in Brazil,” says Nevins. “Why isn’t library science in Brazil at a place where digitizing existing materials is as important as going out and collecting stuff?”
Brazil is no outlier. Troves of important linguistic and anthropological collections exist in museums small and large, in institutes and universities in every nation, each with different budgets and practices around digital archiving. In fact, so many collections are at risk of loss due to catastrophes like fire or floods that just last month the International Centre for the Study of the Preservation and Restoration of Cultural Property held a simulation to train people in how to save precious artifacts after a crisis. Though the ephemerality of material knowledge concerned researchers for years, only recently have international standards for digital archiving emerged.
Fitzgerald notes that the NSF only instituted archiving data management requirements for work it funds in 2011. The Max Planck Institute for Psycholinguistics, in Nijmegen, Germany, created a group that runs a central digital archive in 2000 into which researchers can upload their linguistic field work. The group also funds archival work around the world; Franchetta says the National Museum in Brazil had received funding from them for some of its digitization work. And in 2003, different linguistic groups concerned with endangered languages formed the Digital Endangered Languages and Musics Archives Network, a consortium dedicated to digitizing the diffuse linguistics archives around the world. Though it has a handful of member organizations across the world, none are in South America.
Even when the decision does get made to archive a language, it comes at a tremendous cost. Just this year, the Archive for Indigenous Languages of Latin America, a DELAMAN member run by the University of Texas at Austin, finally digitized a collection of proto-languages from Latin America—among them Mayan, Mixe-Zoquean, and Uto-Aztecan—based on more than 100,000 documents, 900 CDs of audio recordings, and hundreds of boxes of field notes taken by renowned Mesoamericanist Terrence Kaufman. The project took six years, with full time work from professors and graduate students, and specialized equipment. It was only possible through a $302,627.00 NSF grant awarded in 2012.
That figure is more than twice the reported annual maintenance budget of the entire National Museum, which was reportedly $128,000—though this year it only had received $13,000, total, according to National Geographic. The collection in the linguistics wing of the museum alone was far larger than 100,000 documents. To digitize it all properly would have required not just buy-in from the powers that be, but also expensive specialized tools, like noninvasive scanners than can salvage audio recordings from the wax cylinders used a century ago to gather interviews.
And that’s just the equipment. Someone has to watch the tape to make sure it doesn’t skip. Someone has to mark the metadata that makes it possible to search through a digital archive. “Someone’s got to sit there while it’s being digitized. There’s human labor just in that process.” says Fitzgerald. That can be a graduate student or undergrad, notes Nevins, though some specialized equipment requires technicians with specific skills. Fitzgerald recently awarded a grant to a team in Hawaii that will work to make more advanced automated archival tools that might make this process easier—and, crucially, cheaper.
Much of the work digitizing cultural artifacts has always been a labor of love undertaken by dedicated individuals in their free time. A group like this had worked for years scanning small parts of the most important collection that burned on Sunday, known as the Curt Nimuendajú collection. Nimuendaju was a German linguist at the turn of the 20th century who recorded hundred of hours of Amazonian languages that are now extinct. Two linguists in Brazil run the group Etnolinguistica as an homage to his work. Though their website contains some scans of his documents, it is far from a comprehensive archive of his primary sources. “They’re an impressive group that scans stuff all the time but it’s not institutional at all,” says Nevins. “It’s just a bunch of people, a bunch of web denizens who go scanning stuff.”
In the aftermath of the fire, many crowdsourced campaigns have sprung up. Franchetta says the CELIN department has put out a call to any researchers and students who ever photocopied anything from the collection to please send copies back to the National Museum. “But that’s a drop in the ocean,” she says.
Academics from all over the world have been amplifying calls to share any photographs or recordings taken inside the museum in an effort to rebuild. Wikipedia put out a similar call. The spirit of collaboration and a sense of the community coming together in a time of crisis is palpable. But it can’t replace what’s been lost.
“My will, with the anger that we are all feeling, is to leave that ruin as memento mori, as memory of the dead, of the dead things, of the dead people, of the archives, destroyed in that fire,” Brazil’s most famous anthropologist, Eduardo Viveiros de Castro, who was affiliated with the museum, told a newspaper in Portugal this week.
The global academic community, and the researchers in Brazil, hope that memento mori provokes an awakening about the urgent need to digitize the world’s knowledge. If fire comes for another historically important collection, maybe then it won’t take the world’s knowledge with it.