History of the Project
1: Early Phases

Marilyn Kelly-Buccellati – October 2016

Introduction

Work on a comprehensive analysis of all the ceramics excavated at the site started with the excavations at Terqa. This effort was later expanded in the analysis and recording of the ceramics excavated in Mozan/Urkesh. The stages through which the analysis and recording evolved were conditioned by technical innovations that enhanced the results on the one hand and the in-house development of computer programs geared toward the recording of the exponentially increasing ceramic database on the other.

Developmental stages

Since no serious full scale excavations had ever been carried out at the site of Mozan, there was little ceramic data to start with from the site. Therefore at the beginning of our excavations a classification system based on the ceramics we were excavating was established using constantly developing ceramic shape catalogs organized by time period. Rosters and lexica date back to the very beginning, and received a stable formulation in a series of Field Encoding Manuals, of which the 1996 edition served as a standard for subsequent years. At the same time descriptions were made of the various ceramic ware types we were discovering in the excavations. Each ware type was given a letter code and the ceramic shapes were categorized by attributes including general shape, family, sub-family and type and given an alphanumeric code. The analysis using these codes was entered into excel files which were then integrated into the Urkesh Global Record using in-house computer programs written by Giorgio Buccellati.

At the beginning codes were developed for the ceramic vessel shapes only using the the general description and the family codes, but with further excavations gradually the system became populated with more variables and thus became gradually more refined and nuanced. This resulted in changes and additions to the system over time. While these additions and changes were recorded in a master list, the end result was that excavation units that had been excavated and completed in earlier seasons had some, but not the majority, of different codes for the ceramic shapes. This led us to begin a project to harmonize all the codes for all the seasons. From 2013 to 2016 one of our aims was to harmonize these codes for the units being published in the UGR in 2016-7.

From the beginning the methodology encompassed the entire corpus of ceramic vessel shapes from the excavations. This meant that all sherds, whether rims, bases, spouts, handles or body sherds were analyzed so that the sample analyzed is practically 100 percent of the total stratified contexts.

Within the large corpus of excavated sherds there are different levels of classification. There are essentially two levels: that used for body sherds and that used for shape and decorated sherds. The body sherds were analyzed for ware only but if a particular body sherd had an unusual decoration this was also coded with the classification system codes for decorative technique and design. The shape sherds not only had their shape analyzed and encoded but the rim and base diameters were also given. Additionally the percentage of the preserved rim diameter was given so that the accuracy of the rim diameters can be calibrated.

In 2008 we introduced assemblage numbers to link with the UGR ceramic assemblages collected by features, or for various typological or other contextual reasons.

Establishment of ceramic shape categories

The description of the shape categories was the first task in the field. Already in Mozan 1 an early selection of 163 types was published.

Extensive catalogs were prepared in subsequent years. One example is the PDF in house catalog of the AK ceramics, with full description of the wares and the drawings of 53 types, and it was produced in 1997 This material is just from the service wing of the Palace, but it serves as a good example of the work done in all areas of the site as a preparation to their full integration in the current digital book.

Another example is the PDF in house shape catalog of Phases 2-3 . It was compiled in 2006 and contains 53 pages with a large number of shape types from all excavation units with material from the Early Dynastic period.

All of these data are now fully integrated in the description of the shapes.

Establishment of ceramic ware categories

The description of individual ceramic ware categories was another important focus from the beginning of excavations in Mozan. The initial descriptions were made by Marilyn Kelly-Buccellati and published in Mozan 1. For example, the 1997 PDF in house catalog of the AK ceramics (just quoted for the shapes) includes a detailed ware description, which is now integrated in wares.

The initial descriptions were verified and amplified by Dr. Marilyn Beaudry Corbet from the Cotsen Institute of Archaeology. Subsequent studies on specific wares were made by Dr. Yoko Taniguchi of the Getty Conservation Institute, Dr. Ellery Frahm of Minnesota University and Dr Marianna Nikolaidou of the Cotsen Institute of Archaeology.

Research on the wares conducted by Marilyn Kelly-Buccellati and Marianna Nikolaidou is ongoing with a large database (more than 500 individual examples) of high resolution images of scanned sherd sections analyzed by a computer program we were given access to by Dr. Giacomo Chiari, Chief Scientist of the Getty Conservation Institute.

Assemblages

Another significant pre-digital endeavor was the creation of charts with the drawing of the most significant shapes from specific contexts. The context may have been one or more features from a certain excavation unit, or a given stratum or phase, or the like, see for example the list of 77 assemblages for J1.

The original assemblage charts were produced manually, and one may look as an example of this format at the .PDF file of the assemblages for excavation unit A16, as finalized in 2011. A slightly different format was used for the early version of the A16 digital book, see for example A6b8.

These assemblages were then digitized so that each individual sherd or vessel would link to the original file relative to that item. As an example one may look at J1b59, both in its original .PDF file (from 2008) and as it was reformatted manually in a UGR format. The new programs create the assemblage file directly from the idividual sherd or vessel files, see for example A15b30.

Digital version

These catalogs were all in a pre-digital format: they were of course available as digital files, but only as analogs of a printed format, i. e., in a .PDF format. They were then transferred to a proper digital format: this meant that each item was entered and coded individually so as to allow for the variety of sorting that is now possible in this digital book.

Ceramic Code Revision (2013-2016)

Beginning in March 2013 a major revision project of the ceramics code was begun in conjunction with the preparation of the Ceramics Book publication. As part of the process for preparing the ceramic data for publication we updated coding system. This revised system is currently in use throughout this book. A chart explaining the process for converting the old system to the current publication version can be found here.

Whole vessels

An important contribution to the book was the section Laerke Recht wrote on the whole vessels and nearly whole vessels found in the excavations, 405 examples. Emphasis is placed on the types of complete vessels found giving a sense of the combination of shape, ware, decoration and size. The catalog is organized by shape and details are given by clicking on the item number. Photographs and drawings of the vessel are given for most of the examples. Additionally she has provided an index organized by the excavation unit and a second one by field number, making it easy to find any example needed.

The CerPhaS program (2016)

A major result of the collaboration with Dr. Recht’s was the design of a tabulation system that would provide full sorts of the ceramic shapes by phase, stratum, feature, ware and shape, with inclusion of drawings and links to the unit books. A total of 2,429 items was included, and a highly differentiated set of tabulations was created. They can be accessed from the right-hand side of this digital book.

The program CerPhaS was written by Bernardo Forni.

Urkesh Ceramic Analysis

Introduction: The data

History of the Project1: Early Phases