Over the past couple of weeks, I’ve been working with David Pettegrew on a short paper that considers the role of intensive pedestrian survey in documenting and creating Byzantine landscapes in the countryside of Corinth. One of the challenges of this analysis is our scatters of Byzantine pottery tend to be rather small and sometimes amount to only four or five sherds.
The small quantity of Byzantine material present at any one place in the landscape makes it difficult to discuss the function of places in the countryside, to determine the relationship between survey assemblages and more robust samples of material from excavated settings, and to understand the extent, duration and intensity of activities in the landscape. As a result, survey projects have had to consider ways to evaluate periods that manifest in small assemblages of pottery.
A whole series of issues likely contribute to certain periods appearing mainly as small, low-density assemblages. It is almost certain that we have failed to recognize certain types of diagnostic material on the surface or even during pottery study and as a result certain types of pottery are not associated with particular periods. Certain periods also enjoyed problematic natural and cultural site formation processes. For example, sites occupied for a short time or seasonally from particular period could produce less ceramic material. Later activities could obscure the presence of particular periods in the countryside as well. Periods where groups settled on the
In a 2006 Journal of Mediterranean Archaeology article, David Pettegrew, Dimitri Nakassis, and I argued that survey units that produced small, but highly diverse assemblages of pottery because of with low surface visibility might actually contain higher density, very diverse assemblages lurking beneath their obscured surfaces. We suggested in these situations that it might be wise to increase our sampling intensity from the typical 2-meter wide swaths through the unit spaced at 10 m intervals to compensate for the effect of the obscured surface on the overall sample size in the unit. In other words, as densities fell because of poor visibility, we just increase our intensity.
In 2005, David Pettegrew and I concocted a series of experiments at our survey site in Cyprus to determine whether increasing the intensity of our collection strategy actually produced more robust assemblages. In these experiments – documented in an article in the Report of the Department of Antiquity of Cyprus in 2007 – we determined that grubbing around on the ground and collecting all artifacts from a 5% sample of the units surface produced interesting results.
First, our hands-and-knees 5% samples produced far more pottery than our 20% sample (where were walked across the unit counting sherds) predicted.
Second, and more importantly, the assemblages produced by these 5% total collection areas were more diverse than those produced by our effort to sample the artifacts present in our 20% samples of the unit. On the one hand, we discovered that our smaller total collection areas did not produce significantly more chronological information. In other words, we were not seeing periods in our super intensive 5% sample that did not appear in our less intensive 20% sample. On the other hand, our 5% hands-and-knees collection strategy did produce more diversity than our typical survey and sampling strategy. Our samples of 20% of the surface produced 11.2 chronotypes (or distinct types of pottery recognized by our ceramicist) per unit, whereas our more intensive (if smaller) sample produced 15.6 chronotypes per unit.
Our sample sizes remains extremely small, but they are nevertheless suggestive. I looked at the least diagnostic types of pottery (coarse, medium coarse, and kitchen/cooking wares) in each of our experimental units and compared the total number of chronotypes present in each of these classes with the number of chronotypes present in the larger 20% sample. I discovered that for coarse ware, there was a 5% increase in the number of chronotypes, for medium coarse a 35% increase, and for kitchen/cooking wares a 33% increase. There was a 50% increase in the diversity of the fine ware assemblages produced by a more rigorous effort to collect pottery from the surface of the ground.
What this all suggests is that small quantities of pottery based on our typical sampling and collection strategies might represent the tip of an iceberg hidden by collection strategies that ill-suited to documenting hidden landscapes. Of course, one upshot of the need to increase the intensity of surface collection is that it makes it difficult to conduct data collection on the regional level from problematic or less visible periods. This contributes to what Blanton has called “Mediterranean Myopia” or a tendency for Mediterranean survey archaeologists to focus on smaller and smaller areas while still attempting to address regional level survey questions.