One thing that not many digital humanists write about directly, but has become increasingly clear to practitioners in the field, is how ephemeral so much of our thought and work is, especially in comparison to traditional humanities products likes articles and books. What if, while still trying to make our projects more sustainable, we were also to accept ephemerality as central to digital humanities practice?
Not every project can last as long as The Valley of the Shadow, which, at around two decades online and still fully functional, is a marvel. It’s not just that things fall apart, although that can happen through link rot or changing standards that make projects increasingly inaccessible over time, but even that widely used and useful tools or can disappear entirely, as was recently pointed out about Jim Zwick’s seminal “Anti-Imperialism in the United States, 1898–1935”. Sometimes tools get replaced, very intentionally, by something better (like Scribe by Zotero), sometimes not. Sometimes some digital tool or platform that seems like a wonderful thing fizzles, like Dan Cohen’s marvelous Syllabus Finder, R.I.P., but at least eventually something more robust comes along. Even commercial tools get jettisoned. I’m looking at you, Google Reader. Sometimes they actually chug along in different ways that one might have expected; check out this article about Second Life, which is not the rage it was in academic circles a half-dozen years ago.
“It is only slightly facetious to say that digital information lasts forever—or five years, whichever comes first,” Jeff Rothenberg wrote in 1995, and I’ve heard it adapted by many people about projects in the digital humanities in general.¹
Compare that to traditional books and journal articles. By contrast, they are nearly permanent. I can go to my library, or OhioLink (OhioLink, how I love thee!), or interlibrary loan, and get scholarly books over a century old. Heck, most academic presses and even the big commercial houses have been using only acid-free paper for their prime hardback releases for a quarter-century. Journal articles are no less permanent, and that was true even before the databases to which many of us have access. During my grad student days at William and Mary in the mid-1990s, somehow an entire print run of the third series of the William and Mary Quarterly (the prime journal of early American history) became available, and one of my lucky peers was able to snag it. It was like he had inherited a treasure trove. But he was lucky only because of convenience: any of the rest of us could still go to the library and get articles from the beginning of the run. But I digress.
The point here is that ephemerality presents more than only a practical problem in terms of how projects are designed, administered, funded, and maintained. Rather, we must come to grips with the fact that most digital humanities work is inherently ephemeral. How does that make us think about DH differently?
I would suggest we could think and operate in these slightly different ways.
- Include a “snapshot” plan as a requirement for grants. The NEH, for example, laudably requires a data management plan for all DH grants. As Trevor Owens has noted, the issue of digital preservation is both crucial and quite problematic, and as Henry Gladney has written, long-term preservation is a tough nut to crack; Leslie Johnston provides a pretty good quick primer of the issues involved. At the very least, though, we should also include with projects what I’m calling here a “snapshot,” that is, when the project is completed or hits major stages or version thresholds (acknowledging how problematic determining those can be for ongoing projects), basic screenshots and a description of what the site or tool does, how it functions, and so forth, along with the source code, in the simplest and most uniform file formats possible ( such as .pdf and .txt), be deposited in a digital repository with a stable url. Maybe, as a set, they would be the equivalents of the kind of documentation used in other fields in which ephemerality is an accepted condition, such as the field reports of archaeological sites (which are, necessarily, destroyed in the process of investigation) or the curatorial books and pamphlets that document art or other museum exhibits.
- Embrace intentional project sunsets. Rather than having to plan for an unrealistic, unsustainable forever, we could conceive of the finite term as the norm for digital humanities projects, and plan accordingly for what happens when that term ends. How might the project be wound down, when the money runs out and people move on, when links break or formats become obsolete, when a domain is no longer maintained or a server is decommissioned? Maybe, instead of unrealistic efforts to keep things going, the usual expectation should be detailed plans to shut out the lights, so to speak, on a specific date. The project staff could take a final snapshot for the repository, maybe enter the url into the Internet Archive Wayback Machine. The last steps would be to mount a message on the project website and contact users indicating that it’s at the end of its useful life and won’t be maintained, and informing them where the project snapshots have been archived. Incidentally, except for the snapshots, which could be considered proprietary, that’s how businesses do it. My proposed rule of thumb for sunsets would match Rothenberg’s five years, and admittedly that’s arbitrary, but that might be a good general target, to be modified depending upon the needs, requirements, and resources of a given project.
- Live in the moment. Trust me, as a trained historian, this is very hard for me. I’m big on preserving stuff—that’s why, with Billy G. Smith, I started the Magazine of Early American Datasets—but maybe we should more intentionally think of some of our projects as ephemera, digital broadsides here today, gone tomorrow. Sure, some projects that have major institutional backing can continue. But most of us are engaged in smaller-scale efforts, depending upon only a few people, and without the kinds of infrastructure and resources needed to keep a digital project available for the long haul. So let’s enjoy the moment while it lasts, and put a little less energy into preservation of smaller projects in favor of more time spent on making them as wonderful as possible for users now.
This is cross-posted from my Intro to DH 2016 class blog.