There are at least 7 dead emperors, 3 dead kings, 2 dead queens, some dead generals, and one wonderdog following me on Twitter.
This is an occupational hazard, when you put ‘digital archaeologist’ on your business card. Any day of archaeology for me involves standing up at my standing desk, pounding the keyboard in frustration as something that should work, at least in theory, does not. But what on earth do I do? Here’s the quick version:
Archaeology generates incredible volumes of information. A lot of that information is in fact information about the information. The description of the relationship between two contexts? The description of a subgroup of contexts? The spatial and chronological patterns? When the excavation is over, all we’ve got are our records. What I’m working on right now are ways of mining that data (and especially its metadata) for new insights. I’ve recently come back from the world digital humanities shindig, #DH2013 in Lincoln Nebraska. There was an entire session devoted to digital archaeology there; I presented a paper called ‘topic modeling space and time‘.
Topic modeling is a technique that uncovers patterns in unstructured text. It’s a kind of unsupervised clustering routine. Tell the computer, ‘go find me 15 topics in this data, and tell me the relative percentages of each topic in each context record’. This lets me see patterns in the ways the archaeologists themselves worked with the original data whilst excavating.
Very meta. I don’t do much field work any more; hard to do much Roman stuff when I’m here in the Ottawa Valley. But I think I’m finding ways of pulling out meaningful patterns from old excavation notes. You can see – and hear – for yourself where I’m going with all of this at the website for our session. One of the things that is coming out, though there is some disagreement amongst those with whom I’ve shared these thoughts, is the voice of the individual excavators again. The final monograph, the final site report, is always written in a kind of vanilla pablum-esque-ese (to coin a word) that glosses over all of the personalities and social factors that go into that great cauldron, the excavation project (especially the academic ones). I’d like to figure out ways of returning some of the subjective bits back into things.
That’s what this digital archaeologist has been working on, on this day of archaeology, when there was a down moment. Next week, I gear up for back-to-school in September. I’ve got two classes to write more-or-less from scratch. One, ‘Roman Archaeology for Historians’, will have a significant open access portion to it – feel free to drop in, or ping me for more details.