Home > Pyla-Koutsopetria Archaeological Project, Survey Archaeology > Mapping Roman Pottery at Pyla-Koutsopetria

Mapping Roman Pottery at Pyla-Koutsopetria

Over the last two weeks, I’ve been working to map the distribution of Roman and Late Roman pottery from the Pyla-Koutsopetria Archaeological Project.  Roman and Late Roman pottery was collected over the course of intensive survey.  The intensive survey mapped the overall density of pottery across the site on the basis of field-walkers spaced at 10 m intervals who counted all visible pottery 1 m to either side of the their swath through the unit.  While it is typically necessary to adjust the sample size (and the estimated densities) on account of surface visibility (e.g. density of vegetation covering the surface), it has nevertheless proven to be a consistent measure of archaeological material on the surface of the ground.  In fact, intensive surveys from across the Mediterranean have used this technique with only minor variation for the last 20 years of intensive survey. 

To sample the variation of pottery within this sample, our project asks each field-walker in a unit to collect one example of each unique artifact type.  We call this the chronotype sampling method and I have talked about it here on my blog.  We consider it an improvement over earlier methods of sampling variation in surface assemblages (that is all the material present on the surface of the ground).  Some techniques which relied upon the unsystematic collection of “diagnostic” pieces (often rims, handles, or sherds with some outstanding features) failed to account for our ability to analyze subtle differences in fabrics that might not appear to be diagnostic to a relatively untrained field-walker.  Other projects have used total collection, but this would not be feasible for a survey area with large, high-density concentrations of material like Pyla-Koutsopetria where artifact densities often exceeded 5,000 artifacts per hectare. 

While we are confident that the chronotype sampling system produces a representative sample of the variety of material present in the unit (and we have done experiments that have supported this assertion), we are less sure how to proceed in mapping the distribution of artifacts across the site.  The chronotype sampling strategy tends to under-represent very common artifact types and, in any case, makes no effort to produce a sample that has any correlation to the actual number of artifacts of a particular period visible on the ground.  In this regard, it is no better than collecting diagnostic material, although we have argued that the variation in the chronotype sample is often representative of more intensive use of a particular area in a particular period.  Even if a greater number of chronotypes (or unique types of artifacts) does correlate in some way with more intensive use of a particular unit, the opposite is probably not true: certain kinds of sites — for example a warehouse — might produce a substantial assemblage of relatively homogenous material which would be radically underrepresented by the chronotype sample.  Even total collection from each unit (or even just one swath of each unit) would not necessary produce a sample of chronological distribution of artifacts in the landscape in the same way that field-walking produces a sample of artifacts present on the surface.  As many scholars have pointed out, our ability to identify particular types of artifacts (say, coarse wares) varies widely across the periods possible across the landscape.  The “differential visibility” of particular types of pottery — particularly less well-known local ware and utility wares — means that certain types of pottery which are the present of particular economic conditions or functions of the landscape will be systematically underrepresented.  Thus, we are less able to map the distribution of material from periods represented heavily by local wares or by less diagnostic coarse.

Site based surveys often define their sites based on overall artifact densities and then sample for chronology within the limits of this elevated densities.  Thus the site definition and the distribution of material from a particular period across the landscape of individual sites are, in effect, separate functions.  Siteless survey, however, requires that we develop a method for mapping period distributions across the landscape that works more closely with more traditional maps representing the overall distribution of pottery.  Our current methods makes it challenging to identify concentrations of single periods in the landscape and this is a vital task if we seek to embrace the arguments central to siteless survey. 

As a case study for this difficulty, our site at Pyla-Koutsopetria has two well defined zones of elevated, overall artifact density on the coastal plain.  The material collected from these two areas are overwhelmingly Roman and Late Roman in date.  We have been able to argue that these two areas represent functionally different activities during the Late Roman period by comparing the artifacts from each zone.  Last week, I attempted to compare another area of the site to the distribution of Roman and Late Roman pottery in zones 1 and 2.  This third zone represented an area of overall artifact densities that was comparable to those found in zone 2, but the assemblage of material in zone three was far more chronologically diverse than that found in zones 1 or 2.  In particular, the Roman and Late Roman material from zone 3 represented a smaller percentage of the overall assemblage than the Roman and Late Roman material in zone 1 or 2.  So, since our sample of Roman material does not represent the overall density of material in a particular unit, how do I compare in a meaningful way the three areas? My goal here is to evaluate the distribution of Roman and Late Roman material in zone 3.  What is the relationship?

Advertisements
  1. No comments yet.
  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: