Long time readers of this blog will remember that some very long time ago, I was working on digitizing and ultimately publishing the data from the Ohio Boeotia Expedition. This was a very early "second generation" intensive, siteless pedestrian survey conducted around the ancient city of Thisve in Boeotia directed by Tim Gregory. Recently Archie Dunn at the University of Birmingham began a new project there designed to document the Late Roman and Byzantine remains from the town. He invited us to re-examine the survey data in the context of his new work with the hope of producing a more thorough publication of the O.B.E.'s work as well as to complement the more recent work both at Thisve and throughout Boeotia.
I started work on this project about two years ago in Athens, but quickly moved it to the back burner when I was confronted by very difficult mapping and digitization issues. In short, I had a terrible time mapping into my GIS the transects that the O.B.E. walked in the landscape. This was primarily because I was deferring to some published maps of the project's work. I simply could not get my maps (based on some newly collected GPS data, the Greek Army Mapping Service 1:5000s, and augmented by Google Earth images) to coincide in any recognizable way with the published drawings of the survey area and the location of the sites. As I returned to the data this fall -- in preparation for a paper at the Archaeological Institute of America's annual meeting -- I decided to scrap the published maps and attempt to build my own new map. That was a much easier approach and quickly demonstrated that the published maps were, for lack of a better word, wrong.
The other decision that I had to make was to accept that I would not be able to map some of the transects. The recorded data was simply too flawed (and the equally, but differently flawed published maps did not help) to allow me to reconstruct all the transects that the O.B.E. surveyed. This was disappointing, but also liberating as many of the project's transects digitize quite easily.
I also began to key the finds data from the project. The density data, as shown in the image above, produces some recognizable trends. Artifact densities appear to be higher in the immediately vicinity of local (and presumably ancient and Byzantine) villages, decline across the open agricultural plain, and pick up again along the southern boundary of the plain. The reasons for this might be geomorphological. There is evidence, in fact, for an ancient water management system in the plain which Pausanias described:
Paus. 9.32.3. Nothing would prevent the plain between the mountains becoming a lake owing to the volume of the water, had they not made a strong dyke right through it. So every other year they divert the water to the farther side of the dyke, and farm the other side. Thisbe, they say, was a nymph of the country, from whom the city has received its name.
So, it may be that the sediment from the flow of water through the plain was covered ceramic evidence there (although there is some evidence for ancient material in the Thisve plain). It could also mean that this was largely agricultural ground and did not see the kinds of activities that deposited large quantities of ceramics. The only problem with this is that the Cambridge/Bradford Boeotia Project which along with its various spin-offs has surveyed numerous cities in ancient Boeotia, has argued that fields around Boeotian cities were often filled with ancient ceramics deposited through the process of distributing manure in the fields nearest to the town. A more sophisticated and careful comparison of the distribution of material outside of Thisve and outside of the neighboring city of Thespiae may shed some light on this practice.
More importantly, the quality of the finds data is actually quite good. I am confident that digitizing the finds collected and analyzed by the O.B.E. will produce a nice corpus for comparison with other survey projects both in Boeotia and in the "suburbs" of other cities in Greece. Over the next few weeks, I'll update my progress and analysis of this data more or less regularly.