On March 11th, 2011, northeastern Japan was struck by an unprecedented and cascading series of disasters: an offshore earthquake among the most powerful ever recorded, a tsunami that reached as far as six miles inland, and multiple nuclear meltdowns causing the evacuation of over 200,000 people. Thousands upon thousands of people lost their lives, their loved ones, their homes. Reports immediately spread of the remarkable individuals and organizations leaping to provide whatever support they could.

Few were aware—few still are aware—that as the catastrophe unfolded, its rapidly accumulating digital records became the direct concern of several organizations in and outside Japan. These organizations gathered tweets; saved documents; downloaded geographic data; recorded broadcasts; linked to and even captured thousands of websites, YouTube videos, and audio recordings. These undertakings were, in their understanding, archival. Could they be called anything else? And yet key aspects of their projects defied conventional understanding: What kind of archive emerges during an event? What kind of archive seeks to gather items largely already indexed by web crawlers and already accessible—or theoretically accessible—to millions? And what kind of archive includes holdings that may suddenly vanish?

Questions like these ran through a recent conference at Harvard called “Opportunities and Challenges of Participatory Digital Archives: Lessons from the March 11, 2011 East Japan Disaster.” Represented were ten projects related to 3/11 as well two non-Japanese interactive documentary projects: 18 Days in Egypt and the Bengali Harlem Project. Presenters included an unusually cross-disciplinary set: several academics, two geographers, a  developer, a designer, an internet archivist, a “chief strategy architect,” and a librarian. Listening and posing questions were attendees ranging from local graduate students to a medical worker from Fukushima.

Conferences on disaster and participatory archiving are by no means common—one unconnected event called “Archiving Catastrophe” took place last year in New York. At their core lies a new fact of life: disasters like those of March 2011 now necessarily involve the automatic production of vast digital records. From the textual to the audiovisual to the quantitative, these records flow through recordable live feeds in near-real time; those accessible on the web are continually and automatically indexed; and newly created and digitized documents continue to appear well after the events have reached their apparent end. The impulse to make archives may not be ours alone; we build on the spectral prototypes machines already assemble.

I attended as part of the team involved in building Harvard’s Digital Archive of Japan’s 2011 Disasters (JDA), a project initiated by the Reischauer Institute of Japanese Studies—host of the conference through the support of the Japan Foundation Center for Global Partnership—in collaboration with Zeega, metaLAB (at) Harvard and the Center for Geographic Analysis. Since mid-2011 I have worked in association with metaLAB to coordinate among the teams, help conceive the design and functionality of its platform, and to develop and articulate the projects’ overall goals.

The Digital Archive of Japan’s 2011 Disasters

A collection within the Digital Archive of Japan’s 2011 Disasters (JDA)

Of all the 3/11 archives, the Japan Disaster Archive most expands, or even violates, the accepted bounds of the archive. For a start, it stores almost nothing. Instead, it is networked: nearly all of its over 1 million items—documents, websites, tweets, images, audio, video—are not actually held on Harvard’s servers. They come instead from a federation of “content partners” like Yahoo! Japan and the Internet Archive. The partners’ holdings appear in the JDA’s list and map interfaces through the Zeega platform’s harnessing of APIs. The archive is also “participatory.” It provides its users with the means to annotate and geo-locate items and enables them to build “collections.” These collections can either stand on their own or form the basis for multimedia interactive narratives.

Preservation and Participation

Preserving the enormous quantity of digital media connected to catastrophes is a daunting task—even when items are successfully copied, the zeroes and ones that constitute them are by no means indestructible. Clearly the effort must be shared: the Internet Archive has captured over 18,000 websites (110 million documents or 4 terabytes) related to the disasters through its Archive-It tool; Japan’s National Diet Library has begun storing millions of items of digitized and born-digital government media; the All311 project is indexing several terabytes of interviews and other collected media; HyperCities has gathered hundreds of thousands of tweets from 3/11 and subsequent days; and the Fukushima Archive Project backs up its contents—things like radiation data, radio announcements, lists of shelter residents—in a faraway location “with excellent broadband.”

As unresolved as these questions remain, the conference’s shift in emphasis away from preservation was striking. Everyone was talking about use cases and storytelling, public participation and content creation. Akihiro Shibayama introduced Tohoku University’s Michinoku Shinrokuden project, which combines hard data on earthquakes and tsunamis with fieldwork and sophisticated photographic projects, including three-dimensional street views of ravaged coastal areas. Shibayama described his team’s overriding ambition to build an archive “that grows and evolves in response to interactions with its users.” Some presenters, like those from the Fukushima project and a project on the social contribution of religion, even dismissed the label of archives: though they gather and store content, they simply seek to provide space for citizens to find helpful information and share their stories with the world. Others, like Kuniko Watanabe of the Center for Remembering 3/11, and teachers and researchers who had begun to use the digital records, continued to call the represented projects archival but dwelled on themes you’d never expect to hear from the old archons: the democratizing power of citizen storytelling, the tensions between local and global needs for information, or the danger of assuming the whole world is digitally represented. The latter topic came up often and deservedly so: as potentially democratizing as they may be, digital archives run the risk of disproportionately representing people, places, and stories already (or easily made) digitally registered and intelligible. And without creative intervention, “participation” in digital cultural memory may never extend to those without access to—or interest in—networked culture.

A still from The Motorcycle Heroes

A still from “The Motorcycle Heroes

Interaction designer and software developer Yasmin Elayat presented a poignant example of the emerging links between collaborative storytelling and enduring archive. Her project, 18 Days in Egypt, started out as a more traditional documentary of the Egyptian Revolution but early on in production she and her collaborator, Jigar Mehta, radically altered their mission. They saw the thousands of stories in people’s social media and cell phone recordings awaiting collection, expansion, and sharing. These stories—the smaller and overlooked stories—could counteract Egypt’s track record of mis-recorded history and lack of independent media, and convey the actual lived experience of the revolution.

The platform Elayat and Mehta created for the project, GroupStream, allows any visitor to share their media—photos, videos, tweets—and, should they wish, turn them into audiovisual narratives. The hundreds of narratives created since are intensely moving. Elayat walked us through “The Motorcycle Heroes,” about the motorcyclists who risked their lives to transport the injured to field hospitals outside the range of tear gas. “They carry the wounded behind them,” the author writes, “and another protester hops in behind the wounded, to make sure they don’t fall, and drive to the nearest hospital.” A couple of slides later, we see graffiti of a motorcycle labeled “Popular Ambulance.”

Like 18 Days, the JDA also underwent a critical evolution early in its making—one likewise propelled by the explosion of the participatory web. The project’s leader, Harvard professor Andrew Gordon, noted that the JDA was initially inspired by a 2005 collaboration between the Reischauer Institute and Harvard’s Web Archive Collections Service that sought to collect and preserve born-digital materials around Japan’s ongoing constitutional revisions process. This more traditional model quickly morphed, however, when Gordon encountered metaLAB and its then incubating Zeega project (now a spun-off startup). Zeega—whose vision is no less than to “remake the Internet”—was presented by its co-founder Jesse Shapins. Zeega furnishes users with a bookmarklet that lets them collect media libraries from sources across the web. Users can subsequently re-mix these libraries into interactive narratives and websites with non-traditional search environments, like the fantastic Austin Music Map. In the case of the JDA, a Zeega model adds value to archival efforts by collecting in one place otherwise unconnected material. Both Zeega and JDA rely on the same technical architecture and likewise store little to nothing. In a sense they realize a philosophy that, to quote a sutra, “emptiness is form.” Scattered materials congregate where users weave them.

Opportunities and Challenges

An equally important expansion of the concept of the archive is its becoming an active site of shared reflection and mourning. There are precedents in the United States like the Hurricane Digital Memory Bank and the September 11th Digital Archive. In the case of Japan, the JDA may well be used in such a way and includes media from the deeply reflective Center for Remembering 3/11, based in Sendai. The latter has a beautiful and inviting website filled with writing, audio, and video created and collected by people in affected areas—just as many of the stories and interviews document restoration and recovery as they do upheaval.

Aspects of the JDA project continue to puzzle me. Our participation in a Japanese archival project, highly unlikely before the Internet, got us involved in constructing cross-national histories of tragedy. Little lies at hand to help us think about the ethics and implications of this new practice. I find myself asking where I stand when, as at the conference, I watch a seemingly endless video of the tsunami wreaking havoc on a town in Sendai or as I listen to a presenter describe a scanned photograph of a beloved childhood beach in Fukushima, now in the government’s “no-go zone,” to which he may never return in his lifetime? What does it mean for outsiders to work to spread such records of upheaval? What are our motivations? How can we—how can archives—meaningfully serve the recovery effort?

One of the major takeaways from the conference was that digital archives stand ready for innovation by anyone. Putting theoretical language to the seemingly radical potential was metaLAB faculty director Jeffrey Schnapp. Archives in the digital age are, or should become, what he calls “arch(l)ives.” Arch(l)ives are places where people do things: submit, annotate, map, collect, connect, re-code, author. They value “activation as much as presentation”; they work to “launch the afterlife of historical data”; and they build multiplicities of use scenarios into their very construction. To Schnapp, the heterogeneity of participants and included records radically alters the traditional balance of power: a far wider spectrum of stories and events is found worthy of reflection and preservation.

Unavoidable questions remain: What limits do the horrific nature of these disasters place on participation? How might the JDA serve as a model for future archives? What will draw users to them? Who will they serve? Can scholarly argument coexist with public narrative? What dynamics of cash and culture may determine whose major events end up archived? How would people in the United States react if a Japanese university proposed to substantially collaborate on an archive of 9/11 or Hurricane Katrina? On the question of foreign and particularly Harvard’s involvement, Japanese presenters did not express reservations, and instead noted the value added by Harvard’s participatory model and the potential for global reach its resources and standing make possible. Further unknowns included the danger archives (or arch(l)ives) bearing real-time, high-granularity data may present to people in ongoing wars and revolutions: Yoh Kawano noted that, for their work on Egypt, HyperCities realized the danger of turning protesters into traceable targets—they systematically cropped the decimal points of tweets’ geo-tags. Toward the end of the conference, the Fukushima medical worker remarked that while it was good to share the tragedy of these events with the world, local people would like to see more about their “resilience and positive action.”

Front-row as I was, I still had trouble collecting this kaleidoscope of themes into any meaningful question—aside from briefly pointing to the challenges of audiovisual storytelling. Over the weekend that followed, however, some of the most striking themes crystallized. There is a remarkable desire among those involved in these projects to, as Kawano put it, “merge the past with the present and into the future.” It seems that, paradoxically, the ever forward-facing advances of digital and web technologies are compelling us to reinvent how we build and “use” the past. And yet vast electronic traces are expressing to us truths often relegated to academic theory but known by people on the ground: Events are never monolithic but multiple and polyvocal, contradictory and complex. No medium quite immerses us in the experiences of those who lived them—and yet these bridges deserve our crossing, even our efforts at preservation. As Elayat put it: “History isn’t linear—and it can’t be with these new technologies.”