The SAA’s newest journal, Advances in Archaeological Practice: A Journal of the Society for American Archaeology is made available for free to SAA members. The most recent issue has an article by Michael Heilen and Jeffrey H. Altschul titled “The Accuracy and Adequacy of In-Field Artifact Analysis.” It attempts to evaluate the accuracy of the in-field analysis of artifacts collected over the course of survey survey. Heilen and Altschul examine the results of in-field artifact identification, identification based on digital photographs, and hands on study of the artifact in a lab. The study looked at two sites in the U.S. with both ceramic and lithic artifacts and the results are not encouraging.
Dimitri Nakassis pointed this article out to me because we analyzed our artifacts in the field over the course of the Eastern Korinthia Archaeological Survey. The decision to do this was largely dictated by the requirements imposed on the project by our Greek permit. Despite the official character of this limitation, the project has been consistently critiqued for this practice since its conclusion nearly a decade ago. It was vaguely heartening to realize that increasingly U.S. agencies responsible for archaeological sites have begun to question the value of collection outside of the exceptional circumstances. The reasons why are complex, but involve a desire to preserve artifacts in their archaeological context as well as the more practical concern for managing long-term storage and curation costs of survey assemblages. As a result, folks in the U.S. have begun to approach the practice of field documentation in a more critical way.
The article includes a good bit of reasonably understandable, social-science statistics and it well worth some critical scrutiny. Interestingly, it would appear that identification based on digital photography significantly outperformed in-field artifact analysis and best approximate the accuracy of artifact study in laboratory settings. The most concerning drawback of digital photograph analysis was its tendency to identify types of both ceramic and lithic artifacts that lab analysis did not identify in the assemblage. The authors do note that their sample is rather small, the number of potential variables influencing these patters significant, and that their research is just preliminary. They add the tantalizing remark toward the end that improved 3D imaging techniques might make it possible to produce higher quality digital replicas of artifacts in the field. This study captures digital recording techniques at a particularly exciting point in its development, but recognizes that the next generation of in-field digital recording could lead to substantial improvements in how we can analyze artifacts remotely.
While false positives are disturbing, perhaps the most important observations from the article note that more common artifacts are more easily recognizable through in-field analysis or digital photography. More unusual objects, however, were identified less consistently and successfully without careful lab scrutiny.
These conclusions have an impact on the results from our survey in the Korinthia. David Pettegrew has influentially argued that certain very common and highly diagnostic artifact types – namely combed and wheel-ridged amphora sherds typically date the Late Roman period – have shaped how we understand the Late Roman landscape in Greece. These common artifacts have produced, in Pettegrew’s words, a landscape with the “lights turned on”. In other words, a highly visible landscape as opposed to period characterized by less diagnostic material with the lights turned out or, to use John Bintliff’s terminology “hidden landscapes” shaped by material that is either less visible, less diagnostic, or less common in the Greek countryside.
As the authors point out, the circumstances that encouraged the increasingly prevalence of in-field artifact analysis are unlikely to change soon. The regenerative character of the surface assemblages is not enough for archaeologists to treat them as renewable resources. More importantly, perhaps, while digital storage continues to become less expensive and more readily available, long-term storage and curation of artifacts remains expensive. Moreover, the expense of storing and managing large assemblages of artifacts produced by survey projects often fall to institutions and communities that will struggle with the financial burden. Developing accurate and repeatable methods for documenting artifacts in the field should be an ethical priority for the discipline.