How digital art decays—and how to save it
A team of curators takes up the challenge of preserving computer art
L ast summer, Matthew Epler was combing through the archives of the Interactive Telecommunications Program at New York University when he found something remarkable. ITP, a master’s program that focuses on creative uses of communication technologies, has been archiving its students’ thesis papers since 1979. Among them, Epler came upon works from Computer Graphics Art, an obscure quarterly magazine that between 1976 and 1978 published dozens of examples of early “code art.” These images, generated directly from software developed by programmers, looked shockingly contemporary.
“For the past six months, this is all I’ve seen on Tumblr,” Epler, who is a graduate student in ITP, says. Unlike code artists of the 1970s, today’s artists often work in color, but the basic look—dominated by simple curves and repeating geometries—had changed little since the 1970s.
Surprised that contemporary code artists seemed to be unwittingly mimicking the efforts of their predecessors, Epler became enthusiastic about preserving the early material. There was just one problem: When it comes to code art, the art isn’t necessarily just the final result. Some code artists, including Epler, argue that it’s the whole process. And the process was now irreproducible. “Some of these things were programmed on punch cards,” Epler says. There was simply no way to recover the original programs.
For help he turned to Rhizome, a New York organization dedicated to fostering and curating digital works. Rhizome had posted scans of the full run of Computer Graphics Art on its website, available to anyone. In November, Epler used these scans to launch a crowd-sourced effort, the Recode Project, which produces new code to generate the old images.
For years Rhizome has been wrestling with art-preservation problems just like the one Epler encountered. Rhizome began in 1996 as an e-mail list for discussion of digital art; two years later, it had become a nonprofit organization, and in 1999 launched an online gallery of important works known as the ArtBase. Rhizome has since grown further, becoming a patron as well as curator and exhibitor, commissioning and displaying new works in an attempt to build a canon of digital art. Today the organization and its ArtBase operate out of the New Museum in Manhattan.
The concern with preservation at Rhizome and like-minded institutions may seem counterintuitive. After all, digital media seems to avoid most of the problems of decay and degradation that affect traditional, tangible works. But as digital creators have carved out a place in the art world, digital-art preservation has become a huge technical challenge. Conservators face a range of problems unique to this media, problems that speak not only to the complexity of this relatively new kind of art, but also to pesky and fundamental questions about what makes something art at all.
Digital art—which may incorporate video, photography, animation, found and scripted stories, video games, physical installations, interactive narratives, and many media besides—does, in a sense, degrade. It needs to be displayable and sometimes usable decades after its creation. But the technology it depends on changes quickly. The original hardware might no longer exist, or the software might run on an operating system that no one has touched since the first Bush administration. A piece can become what Ben Fino-Radin, digital conservator at Rhizome, calls “a work that has obsolesced.”
Lynn Herrmann Traub, who catalogs and manages the permanent collection at the deCordova Sculpture Park and Museum, likens keeping up with changes in technology to “trying to pin Jell-O to a wall,” but says the more digital art “comes to the forefront in galleries and exhibitions, the more that it needs to be preserved.”
Conservators go to great lengths to prevent, or work around, its tendency to slide into rapid obsolescence. In one instance legendary among conservators for its difficulty, curators and programmers worked with artists Roberta Friedman and Grahame Weinbren to salvage “The Erl King,” an interactive video mash-up of Freud’s “The Burning Child” and Goethe’s “Erlkönig.” Friedman and Weinbren completed the project in 1985 using now-obsolete hardware and software, some of which they had custom-built. Nearly 20 years later, the Solomon R. Guggenheim museum took on the preservation of “The Erl King,” opening the door to trade-offs and compromises.
How would the program run without its original operating system, which couldn’t easily be emulated on modern hardware? Could the code be preserved while eliminating bugs that caused the original system to crash? How much original equipment needed to be used? The artists felt that the new exhibit ran too fast to accurately re-create the original experience, so it had to be slowed down. On the other hand, they were willing to replace the original touchscreen with an up-to-date equivalent. In short, the dilemmas involved in re-creating “The Erl King” forced conservators to decide what was art and what was expendable. Fino-Radin sums up the metaphysical predicament: “When you have a sculpture or painting, it’s very clear what the work actually is,” he says. But when you’re dealing with digital media, “separating what is the actual artwork from the technology that supports it can be a challenging thing.”
Fino-Radin similarly worked against technological advance when he recovered “Digital Diaries,” by Cory Arcangel, one of the leading figures of today’s digital art. “Digital Diaries” was made by scaling up very small videos in order to achieve a pixelated look. When browsers displayed the videos, they showed sharp, colorful blocks, as intended. But since the videos were created in 2002, browser technology has improved, and today’s software smooths off the hard edges in “Digital Diaries.” This smoothing is great for scaling up photos, but it undermines a work like “Digital Diaries,” which deliberately exploited limitations in earlier browsers. Fortunately, for Fino-Radin, the fix was relatively simple, because the old up-scaling algorithms are still available.
Arcangel’s creations are well documented and easy to find online, but with less celebrated works, even the materials necessary to view or run them may have disappeared. Sometimes preservation requires what Fino-Radin describes as “forensic” techniques, akin to what a criminal investigator might do—“recovering data from old machines that we’ve dug out of basements.” These days he is excavating floppy discs and vintage hard drives for material that will be displayed as part of an upcoming New Museum show, NYC 1993: Experimental Jet Set, Trash and No Star. In an indication of just how quickly media can die, Fino-Radin points out that “there is currently a great deal of interest in the preservation of CD-ROMs.”
As artists continue to explore the possibilities that new technologies offer, preservation is accruing new layers of complexity. One new hurdle, for example, is the use of social media to facilitate mass participation in artworks. Consider “Maybe New Friends,” a new piece by artists Lucy Chinen and Ed Fornieles, which involves actors who will play out a story using Twitter accounts. But the program won’t be
limited to the hired cast. Any Twitter follower will be able to join in, as will photographers uploading to Instagram. Though the project will spawn a video that includes much of this material, the interactive experience cannot be retained; it’s effectively a live, ephemeral event. With the public becoming involved in art creation in real time, the challenges of preservation will likely grow.
Fino-Radin is confident that conservators are winning the race against time and technological change, but given the pace of that change—in the digital world, platforms emerge, rise to prominence, and decay into obscurity in the span of a decade—they have a difficult job ahead. But it’s an important one. While art is not exactly progressive in the way science is, with each new object or discovery built on the shoulders of giant predecessors, it is nonetheless highly responsive to its own history. As Epler found, recovering the history of a form like code art can do much to inform its present—and perhaps help guide its future as well.
Simon Waxman is managing editor of Boston Review and has written for WBUR, Alternet, and McSweeney’s.