Doing Some Data

As I’m struggling to get some momentum in the new semester, my attention has turned to a few data projects. These projects are either at their very early stages where it’s pretty easy to get a false sense of progress or sufficiently mature and robust that I can begin to pose some testable hypotheses. All data projects, though, reveal something of how the data was made.

Here’s a little window into some of my ongoing data work.

1. Isthmia Data. This summer, I spent some time with Jon Frey and Tim Gregory at the Ohio State Excavations at Isthmia. We discussed how to get various datasets from various sub-projects at that site to communicate with each other. Like most archaeological sites, each archaeological context has a unique number and this number should be part of the data associated with each object or group of objects from that particular context.

One of the challenges with Isthmia data, however, is that inventoried objects – and these are finer examples or historically or archaeologically significant finds – receive a sequenced number that is based on the year that the object was inventoried and the type of object. So we have IPL for lamps, IPR for Roman pottery, IC for coins, et c., and for each inventoried object we have an inventoried card, which looks like this. These cards may or may not have a “lot number.” The lot number defines the archaeological context for the object. Isthmia used a version of the Corinth system in which each archaeological context is a “basket” (or sometimes “box”).

This is where it gets fun. Each lot number consists of three numbers. The first part of the number is usually a two-digit indicator of the year: 78-. The second part of the lot number refers to either the excavator or the area excavated.This is sometimes represented by the initials of the excavator (which correlate with the initials of a notebook). Each initial is made into three characters. The initial CV then becomes CVO. In some cases, however, instead of the initial of the excavator, a lot gets initials from an area. This often happens when several excavators dig in a particular area over the course of year. So the south area of the Roman bath gets the initial RBA associated with those lots even though the notebooks have the initials EM/KK associated with them. This is important because the initial EMK is used in lots excavated in another area. Finally, the basket gets a three-digit number starting with 001.

This is clearly an ad hoc system designed to eliminate confusion at the time of recording. Later, however, it adds to the confusion because to add lot numbers to inventoried artifacts, one must trace the initials of the notebook, excavator, and area to determine the name of the lot assigned in that particular situation. Most of the time, this is possible. Some of the time, it is not. It’s important though to be able to tie together inventoried finds with context pottery.

2. Modern Landscapes of WARP. I’ve been working on an article with my colleague Grace Erny that documents the Early Modern and Modern site of Chelmis in the Western Argolid. As part of that work, I’ve started to think about the Early Modern and Modern ceramics across the survey area. For our project – and in most of Greece – the Early Modern period dates from around 1821 and the Greek War of Independence (or more usefully for archaeological artifacts 1800) to the early decades of the 20th century.

I’ve become particularly interested in the relationship between Early Modern ceramics and settlement in our survey area. Our survey area included three villages: Schinochori, Sterna, and Lyrkeia. What initially struck me as odd is that only Lyrkeia (formerly known as Kato Belesi) had a clear halo of Early Modern fine and table wares. The little triangles are Early Modern fine and table wares on the map below and the cluster around the village of Lyrkeia is clear.

Basic GIS Final
 This pattern baffled me until I remembered a conversation with project director, Dimitri Nakassis, who noted that in the early 18th century Venetian maps and records the villages of Sterna and Schinochori were listed in Venetian documents as deserted. While we know that both villages existed by the early 19th century, they were smaller than the village of Lyrkeia/Kato Belesi and lacked prominent roles in the regional administration in the 20th century. In this historical context, perhaps the data makes more sense. The presence of the halo around Lyrkeia likely attests to its size and prominence in the region. Sterna and Schinochori, in contrast, remained smaller and almost certainly less wealthy during this period.

The even smaller area of Chelmis lacks much in the way of Early Modern fine ware as well. Not only is this settlement smaller still than the villages of Sterna and Schinochori, but it was also a seasonal settlement. By comparing its rather limited assemblage of Early Modern table ware with others in the region, we have a basis to at least understand the variations present in the density and character of Early Modern material across the survey area.

3. PKAP 2 Data. Over the last six months (and six years), David Pettegrew and I have been pulling together the data from our excavations at Pyla-Koutsopetria on Cyprus. Moving data from something that’s central to “in-house” analysis to being available to the public is always a revealing process. We have now submitted our dataset to Open Context where our the data from our survey was also published. During the data preparation and publication process, for example, we discovered, for example, some records that preserved some experimental recording practices using so-called “iPads” in the field. We also found some little incompatibilities between our catalogued artifacts, which receive much more scrutiny, and our “context pottery” or finds that are read as excavation is taking place.

Finally, we’ve thought a bit more carefully about how to prepare our data to support the kinds of analysis that we want to publish. For example, we want to group our stratigraphic units from across trenches at Vigla into phases, and to do this, we need to include a Vigla phase number to each SU entry. We also need to be attentive to other opportunities to aggregate our granular data into digestible assemblages. For example, last spring, we asked for a stable identifier to all Late Roman fine ware from the survey phase of PKAP.

This kind of fussiness is important because we’re publishing our data in advance of publishing our monograph on the work at Koutsopetria and Vigla. This should allow us to finish our manuscript with substantial and consistent links to our data.     

Leave a Reply

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

You are commenting using your 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 )

Connecting to %s