Moving slowly and breaking things

The Day of DH is over in Madrid as I write this, but it’s still got another 35 minutes or so in Austin, and I’m still working (as it happens, on DH-y things). Over the course of the day, I was thinking about how the various different digital-humanities fragments that I’ve spent time on might come together as a coherent whole. This is tricky, because here’s the basic list:

  • I spent 20 minutes this morning alpha-testing a new reconciliation tool that allows you to match named ancient places with URIs for places in the Pleiades spatial gazetteer (handy if you have a spreadsheet full of ancient place-names that you’d like to geoparse)
  • I spent another half an hour or so cleaning up entries for events and objects in a Google spreadsheet that my undergraduate class on Archaic Greece created collectively in order to visualize important aspects of the period in space and time, using the Open Knowledge Foundation’s TimeMapper platform
  • Yet another small chunk of time was dedicated to manually cleaning a json file with period definitions extracted from Library of Congress Subject Headings, bound for incorporation into the PeriodO gazetteer of period definitions; the grad student who was working on this encountered a situation we hadn’t addressed in the interface design (reduplication of records that happens when you push an updated set of definitions that still contains the un-merged versions of periods that have already been merged with the canonical dataset)
  • And right before I sat down to write this post, I was putting together slides for a lightning talk I’m giving tomorrow in an annual Digital Preservation Symposium at UT Austin — the theme of the symposium this year is “Breakdown: the space between action and expectation”, and the talk is about the places where my dependency-heavy digital humanities/digital pedagogy projects show cracks or gaps

That’s a pretty disparate set of activities. But thinking through the lightning presentation gave me the thread that ties them all together — they’re all about the space between the potential of linked information systems, and the fragile, evolving, imperfect and easily broken content and tools that we hope will allow us to reach that potential. It seemed apt, therefore, to borrow the famous start-up mantra popularized by Mark Zuckerberg, “move fast and break things“, with a minor modification: on the DH side, we’re breaking things, all right, but the structure of academic funding and labor means that we’re usually moving pretty slowly while we do it. One of the projects I’m discussing in my lightning talk, GeoDia, started almost ten years ago, and more than five years ago it provided the foundations for what became PeriodO, which is only now starting to bear fruit. That’s a geological age in start-up years.

On the other hand, we’re moving slowly because, as humanists, we tend to take the long view — and we move forward by building on what came before, rather than by disrupting or creatively destroying it. If you have some patience, the long game starts to pay off. And that was the other side of my Day of DH: in addition to the broken things I was trying to fix, I began to realize that there were also a lot of seeds that had been produced through the cross-pollination of these various projects, and some of them were already beginning to put up shoots. For example, I was eager to test the Pleiades reconciliation tool because I’d just been talking to a graduate student about how to format a spreadsheet documenting particular spatial interactions in Herodotus, with an eye toward eventual mapping. I thought the tool might make her job easier — and I knew what that job would be, because I’d done it with undergraduates a couple of years ago in a class I taught on Herodotus and networks, as part of the Hestia2 project team. I worked on that with Elton Barker and Leif Isaksen, who are also the codirectors of the Pelagios project, which, as Pelagios Commons, is now focused on building a community centered around Linked Data representations of places (and eventually other entities, like people and periods) in data related to the ancient Mediterranean world.

Pelagios Commons is also supporting the construction of infrastructure through small grants to projects working with Pelagios-friendly Linked Data as either producers or consumers. The Commons encourages these projects to draw on work that has already been done, and as a result both Pleiades and PeriodO are benefiting from (and having to accommodate) an expanded user and contributor base. I’ve been working with students in my Greek Archaeology class on contributions to Pleiades since 2013, and as I corrected the spreadsheet entries for my current students’ TimeMapper project (they’re required to include Pleiades URIs for objects and monuments), I realized that several of them were using URIs attached to places my former students had created or co-authored (like the Temple of Apollo at Delphi or the Temple of Hera at Olympia). Some of the work for the TimeMapper project will be fed back into GeoDia, while the other project I’ll discuss at the Digital Preservation Symposium is a student crowdsourcing platform to generate Linked Data image tags for periods and places to make it easier to pull visual resources into platforms like GeoDia or Peripleo (the Pelagios browser).

That, for me, is Digital Humanities in a nutshell. Platforms and approaches are constantly breaking and needing to be fixed, and that can be frustrating — and exhausting. You can see why some scholars think this is all a waste of time. At the same time, even as the corners of our constructions crumble and the paint peels, if we wait long enough we can see both content and infrastructure begin to be used and reused, to become the foundations of a bigger, broader, more integrated body of knowledge. And this is immensely gratifying. We just have to remember to move slowly enough to ensure that our content persists in reusable forms, even as we break our websites and platforms.