I managed to work my way through J. Bintliff, P. Howard, and A. Snodgrass, Testing the Hinterland: the work of the Boeotia Survey (1989-1991) in the southern approaches to the city of Thespiai (Cambridge 2007) over the past couple weeks. A single read-through only began to scratch the surface of this remarkably rich and dense report on the work of the Boeotia survey around the city of Thespiai. It goes without saying that this book should have a significant impact on how scholars present and analyze data produced through intensive pedestrian survey in the future.
Thespiai is the nearest ancient polis to the city of Thisvi where I am working (in fits and starts) with Timothy Gregory, Archie Dunn, and others to re-examine the data gathered over the course of the Ohio Boeotia Expedition. Gregory and his colleagues produced the data from the Thisvi survey about a decade earlier than the work by the Boeotia Survey around Thespiai and with a much smaller team. Consequently the data is of a lower resolution and reflects intensive survey practices that have now endured rather significant critiques. Nevertheless, the high quality data set available from Thespiai should help fill in some interpretive and evidentiary gaps from the Thisvi survey and hopefully enable us to make a more substantial contribution to the archaeology of Boeotia (and Greece) through our rehabilitation of this data.
From Testing the Hinterland, five things struck me as particularly significant about this volume:
1. Thespiai as Large Site. As I have noted several times previously in this blog (Some Thoughts on Future of Survey Archaeology in Greece (and the Eastern Mediterranean) and More Large Site Survey), large sites of over 50 ha but generally smaller than 700 ha are poised to become the basic unit for interpreting the archaeological landscape in Greece. This first volume from the Boeotia survey analyzed a mere 500 ha. While the total work of the Boeotia Survey encompassed a much large territory, the resolution at which they surveyed so much of western Boeotia made it practically impossible to produce a single volume of interpretation that did the data justice. In fact, of the 200 sites that the Boeotia survey documented, they included merely 18 in this volume (171). It is doubtful that future intensive survey in Greece will even cover as much ground as the entire Boeotia survey project, but the sophisticated and largely independent analysis of a 500 ha in a single volume should be cause for optimism (and a model) for the next generation of small scale intensive survey project in Greece and elsewhere. The vast quantities of fine resolution data produced by the most recent generation of intensive surveys can be brought to bear on problems of historical significance (settlement patterns, rural land use, et c.) despite arguments for Mediterranean Myopia (see: A Rambling about Survey from a Regional Prespective).
2. GIS and Survey. This volume has achieved the full integration of Geographic Information Systems (GIS) and intensive survey. Bintliff et al. brought together view-shed analysis, least cost path, and the integration of divergent data sets (soil types, different types of artifact density data, et c.) into a single analytical framework. While GIS has been a central aspect of Mediterranean survey for the last 25 years, it has often served as a convenient receptacle for data collected in the field, but not played as active a role in the analysis of that data. The Boeotia survey used both viewshed analysis and a relatively sophisticated least cost path model to understand the density values present in various parts of their survey area; in doing this, they successfully manage the limitations of sample size and the computational capacities available to them to produce plausible analyses. They were able to define the site of Leondari Southeast 7 (LSE7), for example, by recognizing that the artifact densities for this site far exceeded the off-site density values expected on their basis of their least-cost path model. LSE7 stood in a zone of "very high 'friction' where access is not easy" (233)
3. Density and Intensity. The Boeotia survey is among the first surveys in the Eastern Mediterranean to recognize and attempt to control for differences in density produced by different methods of artifact collection. More intensive collection techniques produce greater quantities of artifacts. This leads them to attempt to compensate for the differences between densities produced my more intensive "on site" data collection and less intensive transect field walking or off-site data collection practices. In general they found that on-site data collection methods produced between 2.5 and 3.5 times as many artifacts as transect walking. Using these figures to adjust their comparison between transect density and the densities produced by on-site collection, they were able to determine whether specific chronological components of a site (say the Late Roman material) actually represented particular Late Roman activity at the site or simply part of the larger, but lower density carpet of Late Roman artifacts recorded by transect field walking. This is a valuable contribution to survey archaeology and can be compared to our recent experiments at the Pyla-Koutsopetria Archaeological Project and published in the 2007 Report of the Department of Antiquities of Cyprus.
4. Questions of Procedure. One striking thing about this volume was the lack of sustained discussion of field procedure. While I recognize that innumerable smaller publications have proceeded this volume and it is likely that their field procedures were discussed at length in these papers, it is nevertheless disconcerting that a project that committed so much sophisticated thought to the analysis of their artifacts densities would not tie this analysis directly and clearly to field procedures. This is particularly significant when we consider the sample of artifacts gathered from individual survey units that forms the basis for their chronological and functional analysis. It seems to me that the day where we can simple claim to have collected "diagnostic artifacts" from a unit is over. The Eastern Korinthia Archaeological Survey and Pyla-Koutsopetria Archaeological Project employed chronotype sampling which ensured that we collected at least one example of each type of artifact present in the unit. As we (me, David Pettegrew, Dimitri Nakassis, Scott Moore) have argued, this does distort our sample, but in relatively predictable ways. Other projects have defaulted to areas of "total collection" to act as controls against less systematic grab samples of diagnostic sherds from survey units.
5. Manuring the Countryside. This volume represents the most sophisticated and sustained argument for the manuring hypothesis. It contends that Boeotian cities transported huge quantities of manure from the urban center to the countryside and with this manure came pottery and other forms of domestic waste that formed a visible halo of artifacts around both the urban center and more substantial sites in the countryside. I must admit that the simplicity of this argument is appealing (although I do have particular loyalty to David Pettegrew's challenge to it!). Unfortunately by separating the analysis of the urban fabric from the countryside as they have in this volume, it is difficult to understand the relationship between the material remaining in the urban center of Thespaia and the material found in the associated halo produced by manuring. The manuring hypothesis will be more persuasive when they can show that the material from the halo and the city center is fundamentally similar. At PKAP, for example, we discovered that the highest density areas of the site during the Late Roman are surrounded by lower density concentration of Late Roman pottery. To test for the manuring hypothesis we compared the types of material present in the highest density zone to the material present in the lower density halo. The material in the halo was different. For example, we found very little Late Roman African Red Slip in the halo (but plenty of other contemporary, imported and local fine wares) while it is remarkably common in the highest density units. Unless we assume that something about ARS led to different discard behavior, it is hard to understand how ARS did not appear in the halo if it was present at the center of the site. While this one type of artifact alone does not completely eliminate the manuring hypothesis as a possible explanation for the low density halo around our highest density units, it nevertheless produces a kind of challenge made possible by integrating the analysis of on-site data with that gathered from off-site distributions.
There is much more to this book than these 5 observations: the detailed documentation and interpretation of individual sites, the clever ternary analysis of site function, and the effort to deal with the post-antique survey data (albeit in a cursory way) among many other fine points. Needless to say, this work will emerge as a point of departure for many subsequent studies of intensive survey in the Greek countryside.