I had the good fortune of participating today in a conference called The Big Ancient Mediterranean at the University of Iowa. The purpose of the conference is to discuss ways in which digital projects (including my own Inscriptions of Israel/Palestine) might better use linked open data to facilitate research. There is a nice cross-representation of projects that primarily provide data (like my own) and frameworks or services (like Pelagios) that bring together the data from different projects.
The idea behind linked open data is that different kinds of data (e.g., texts, coins, inscriptions, papyri) can be brought together based on one or more criteria. A simple example might be some connection to a place (as Pleiades does), but a date (or range) or person might be the criterion of selection. For example: Give me everything related to Jerusalem from the late Second Temple period. (This example is a little tricky because it involves defining the “late Second Temple” period. Fortunately, there’s a site, PeriodO, that will soon be able to do this.)
There is a technical issue at the heart of this kind of data gathering on which we have spent, and will continue to spend, a significant amount of time. For any service to gather data from another site according to a criterion it must know how to query the other site. Ideally for these purposes, then, all participating sites need to use a common, controlled vocabulary (or other stable identifier, or URI). Otherwise, if my “Jerusalem materials” are designated as belonging to Aelia Capitolina (the name of the city given by the Roman emperor Hadrian), a search might well overlook them. Having all participating projects use some common linking vocabularies is not impossible, but given both the scattered and often under-funded nature of such projects as well as preferences of individual scholars (never to be underestimated) it is challenging.
The more interesting issue, though, is why bother? Undoubtedly, the gathering of multiple kinds of materials with one search is potentially efficient, especially if I can more or less trust the results. If I want to write about the city of Sepphoris in Late Antiquity, a single search might ultimately bring me the bulk of the materials I need.
For most of the participants (and I include myself) at the conference, though, there is a more exciting if still hard-to-articulate promise to this kind of data selection. It is not just the collecting, but the digital assembling and visualizing of such data that will either answer research questions or pose new ones. I’ve been thinking a lot lately about this issue in general terms. How can the use of digital tools not just help scholars to do what they’ve always done as scholars quicker and better but also do something entirely new?
I do not have an answer to that question, but spending the day thinking about linked data has helped me to see better one possible direction. Digital visualization helps to shift the vantage point. Usually, for example, when I create my narratives I start with texts. I branch out from the texts, using geography, archaeology, etc. to enhance or challenge my texts, but the base is usually textual. Think of this as standing in a point (the textual point) at the lines of a web that connect to my other kinds of data. But if I move to another point on the web – say, geographical -and look back out at my web, it will look different. With the click of a button I can make my starting point a map, an inscription, the visual image of an archaeological site, or graph. More or less the same material, but a different view and context.
Transformative? I’m not sure. But I’m intrigued by the possibility of easily changing my vantage point, having my data shuffled, and seeing what new insights or questions emerge.
More to come. In the meantime, there has been an active twitter feed of the conference at #BAM2016. I look forward to tomorrow’s sessions.