Thursday, November 17, 2011

A Mid-sized Site in Sicily

A Mid-sized Site in Sicily « The New Archaeology of the Mediterranean World

A Mid-sized Site in Sicily November 16th, 2011 §

This past week, I was all excited to get my annual copy of the Journal of Roman Archaeology. So far, the most exciting article in this year’s volume is the brief report by K. Bowes, M. Ghisleni, G. Francesco La Torre, and E. Vaccaro on the site of Sofiana/mansio Philosophiana in the hinterland of the Piazza Armerina. The report describes recent work by a combined American and Italian team on the site of Sofiana. The site is located some 10 km south of the famous 4th century Piazza Armerina, and scholars have generally understood it to be some kind of service village or statio for the larger villa and located on its estate.

Excavation at the site in the 1980s and 1990s produced a domus, a bath, some cemeteries, and a later Christian basilica dating from the 1st to 4th centuries or so. The site extends for around 15 – 25 ha (at least according to the maps provided), and this would place it between the small sites like isolated farms and larger urbanized sites in the Mediterranean basin. Work at the site, however, identified at least two grid orientations suggesting that the site did have some features – like a planned road system – typically associated with urban sites.

The Italian-American team sought, in part, to expand the scale of work at Sofiana to understand the exact nature of places which would appear to share features of both rural and urban settlements and attempt to determine its function in the Sicilian landscape. To do this they used both intensive pedestrian survey and magnetometry to document surface and subsurface remains.

The relatively compact area of the site allowed the team to employ a rather intense form of pedestrian survey. The gridded the site into 10 x 10 squares in their GIS and then surveyed one of every three grid squares for a 30% sample. The teams collected all artifacts from the surveyed units to avoid biases associated with collecting diagnostic sherds (or chronotypes!). They then plotted period specific artifact densities in their GIS using Kernel Density Estimates (KDE) to smooth their results across the entire site. We followed similar methods of surveying a slightly larger site at Pyla-Koutsopetria on Cyprus; we opted for 20% sample based on 40 x 40 squares and collected chronotype samples from each grid square. Both surveys functioned at a similar resolution but my guess is that the Sofiana project produced far more pottery. The Sofiana project’s use of KDE smoothing produced convincing and easy to understand maps for each period at their site.

For a 2 km range beyond the core site, the Sofiana project conducted a more traditional, regional level intensive pedestrian survey over 280 ha. Field walkers were spaced 8 m apart except when visibility was particular poor then they spaced their walkers at 5 m spacing. We gently hinted at this technique in a 2006 article in the Journal of Mediterranean Archaeology and while they don’t seem to know about our work, it was great to see this method implemented. They appeared to have sampled artifacts on the basis of “on-site” and “off-site” designations. As this is a preliminary report, they did not provide more information on their more regional survey, but I’d be curious to understand how they sampled for chronology and function.

They complemented their work at Sofiana with a magnetometry survey and some 32 test pits. The magnetometry and test pits demonstrated that some basic grid orientation seemed to exist at the site and some production of building material (particularly kilns for tiles). They also provide a smoothed chronological profile for the site using Individual Weighted Means method which is a form of aoristic analysis (or vice versa, I’m not entirely sure). There is an obvious peak in activity at the site in the 4th through 6th centuries covering an area of 21 ha. The site continues substantially even later.

This preliminary report provide only a hint as to what the authors think this site represents in the Late Roman countryside.

On the one hand, the site is smaller than most of the small town on Sicily. On the other hand, the orientation and presence of at least some feature expected of urban areas (which tended to have administrative functions in the Roman world) place this site in the realm of the rural vicus, which could have some urban features but lacked clearly defined administrative roles in the Late Roman state. Such vici are good examples of the kind of third spaces recently discussed by Myrto Veikou for Byzantine settlement in Greece. Large rural “agglomerations” like Sophiana with production, grid planning, and imported ceramics defy our understandings of the rural/urban distinction in the ancient world and interrupt any simple cultural, economic, or social model grounded in the simply in the administrative structure of the Late Roman state.

Categories like statio – or road side settlements – add even more complexity to how we imagine the ancient world. These sites can be rural (in fact, they are typically situated between urban sites), but they also can enjoy urban features. The importance of Sicily to grain production toward the end of the Late Antiquity (especially after the fall of Egypt)

might help us understand the continued prosperity of the site of Sophiana and its continued significance into the 7th century.

While I look forward to the final report on the work at this site, the preliminary report highlights so many of the crucial issues facing our understanding of the Late Roman countryside and settlement more broadly. As intensive survey allows us to document rigorously more and more rural space the old distinction between urban and rural breaks down offering new perspectives on the production of culture and society in the Late Antique world.