Urkesh Ceramic Analysis

General

Methodology and Process

Marilyn Kelly-Buccellati – October 2016

Back to top: Methodology and Process

Introduction

The philosophy behind the excavation methods and recording system is that everything counts, whether it is a single body sherd or a small lump of clay which might contain, when cleaned properly, a seal impression. These are all collected and recorded within an interlinked comprehensive system, the Urkesh Global Record (UGR). This system forms the background of the comprehensive analysis and recording system used for all the ceramics excavated in all the various excavation units. Over the many years of the Mozan/Urkesh excavations the methods and procedures were constantly refined so that they were more efficient. The description here gives a summary of the standard procedures that were in effect for most of the field seasons.

This section explains in detail the procedures both for collecting the sherds in the field and the steps for their analysis in the excavation house, including their final disposition. At the very beginning of the excavations in Mozan/ancient Urkesh we decided to analyze all excavated ceramic sherds, including body sherds and shape sherds. The whole and reconstructable vessels were analyzed within the procedures established for the analysis of all objects from the excavation, while the sherds were processed and analyzed according to the explanation below.

Back to top: Methodology and Process

Field procedures

As a principal all sherds were collected from every excavation unit and analyzed. The sherds were collected from by unit and feature in q-lots which were small contexts within the feature; coordinates of the context were entered in the field records. No q-lot is allowed to be more than is contained in a medium size plastic bag (approximately 40X25 cm although this varied according to what size bags were available in the local market). This restriction resulted in a fairly limited number of sherds in any q-lot. The tags closing the bags are written in indelible black ink on surveyors tape; the tags specify the excavation unit, feature number, q-lot number, initials of the person who wrote the tag and the date. These ribbon-type tags are used to close the bags. The fact that the bags were small had also the practical advantage that the sherd bags would not tear easily. No q-lot is kept over from one day to the next, no matter how few sherds it contains. The bags were taken back to the in-coming area of the dig house at the end of the field day. If the excavators saw an unusual sherd or cluster of sherds, perhaps indicating a whole vessel, they would put them into labeled a small bag within the sherd lot. However if the sherds appeared in the field to be all part of a complete vessel this would be considered an item and was processed separately from the rest of the sherds in the feature with a process used for items.

The sherd bags are returned to the dig house at the end of the day. Directly from the field the unwashed sherds were stored by unit and feature in a large storeroom filled with shelves which led from the field into the expedition house. Before the sherd lots were placed on the shelves, a separate inventory was taken so that no lots went missing and if a number was assigned and not used, that was noted. We had come to this procedure by trial and error because it was so time consuming to look for sherd lots that were missing in our record, sometimes not to be able to establish what had happened to them. It was from these storage shelves that the sherds were taken to be washed, always by unit and feature.

In the expedition house the ceramic processing took place in a large courtyard dedicated to this activity with the Ceramic Library room next to it. The courtyard is covered during the field season with roofing tarps so that the bright sunlight was filtered for the sherd analysis; the tarps also helped in keeping down the intense heat (on many days more than 40 degrees centigrade). The sherds are washed by unit and feature which means that there is a constant interaction between the field and the analysis teams.

The only exception to the rule of analyzing all sherds was the treatment sherds from topsoil contexts. These were collected but not within the q-lot system. They were inspected and any unusual types of shape, fabric or decoration were retained but the rest were discarded in the field.

The procedure methods developed over time and were not always entirely followed for a number of reasons. This resulted in gaps which can be seen in the record of the various books.

Back to top: Methodology and Process

Post-Excavation Methods -processing and analysis in the expedition house

Throughout the subsequent stages of processing and analysis the sherds were always kept together by excavation unit and feature. This meant that in general the sherds were washed by unit and feature. Only on a few occasions, because of the excavation strategy, a feature was continued to be excavated after a gap; in these cases they were analyzed over a period of time.

Back to top: Methodology and Process

Sherd Washing Stage (and Photos)

Because of the very large amount of sherds excavated in the various excavation units it was impossible to wash all excavated sherds on a daily basis. This stemmed from the fact that all excavated aherds were analyzed and was partly due to the fact that we hired two handicapped individuals to wash the sherds as this was the only activity they could successfully manage. Their work place was in the sherd yard, under a tree for shade and with access to water and a muddy water discard area as a part of their work area. The area was paved so that it was always clean and relatively dry (see image).

The sherds were washed in plastic tubs which was manageable for our sherd washers because the q-lots of sherds never contained many sherds. As the sherds were washed they were put into plastic sieves by q-lot and the ribbons written in the field with the stratigraphic and q-lot number information was tied to the sieve; the typical plastic household-type sieves were ideal for this purpose and not expensive as we needed hundreds of them. We had a special metal drying rack for the sieves so that a large number could be drying at the same time (see picture). The dry sherds in the sieves were then arranged by excavation unit and feature along a sidewalk in the sherd yard against a wall. On the wall signs indicated what area and feature the sherds came from. The sherd analysis team took the sherds to work on from this sidewalk by unit and feature as the analysis was done by feature.

Back to top: Methodology and Process

Analysis stages

In the analysis stages there were essentially two different paths followed. Initially after the sherds in the q-lot were inspected for joins or anything unusual, then the body sherds were separated from the shape sherds and the shape sherds had the stratigraphic information written on each one. At this point the body sherds were separated from the shape sherds which were kept in the sieve with the ribbon written in the field still attached. A dedicated person analyzed all the body sherds, separating them by ware type. In order to determine the ware type the analysis takes into account the color, firing, inclusions, and any unusual characteristics. In order to do this the section of the sherd was broken with a pliers so that a fresh surface was studied. The body sherds were then separated and counted by ware type and recorded onto our ceramic analysis forms. The body sherds that were decorated were only saved if they had an unusual decoration. The body sherds were either reused by us or discarded. The reuse took two main paths. One was that they were embedded in the top layer of a cement matrix made in forms ca 60X60 cm. These tiles were used in the palace on top of the backdirt covering all the floors so that the visitors waled only on the sidewalk of tiles, thus preventing damage to the floors and mudbrick walls. In the excavation house communal walking surfaces were paved with cement with the top layer being embedded body sherds. The entire sherd yard was also paved in this way. The second path was to fill in the topography of the mound near the main door of the excavation house so as to make a flat surface for parking a few cars.

The shape sherds (rims, bases, handles, spouts, body sherds with unusual decoration, and sherds that had traces of secondary use) were analyzed by a local team who had been especially trained for this work. Primary among them is Hammade Hamza who is still doing this work even during the period when we are not able to go back to the site. At this point the shape sherds were still in the plastic sieves tagged with the stratigraphic information written in the field. At this point he shape sherds are given individual p numbers, connected with the excavation unit and q-lot (eg. J5q184-p2). These numbers are written on the interior of the sherd along with the feature number and for many also the ware type. Each shape sherd is then analyzed by categories that include ware type and shape type. The sherds were individually analyzed and ware type encoded on the encoding form. The shape types were indicated on the encoding form for the categories of overall shape, shape family, shape sub-family, and type. These codes were found in a master catalog of sherd drawings based on the shapes previously excavated from that time period.

Back to top: Methodology and Process

Chronological catalogs

The basic shape types, from the beginning of the analysis, were organized by shape into catalogs by period. These catalogs naturally grew as we continued the excavations in those time periods. Each shape was given a code based on overall shape, family, sub-family and type number. Examples are best seen in the catalogs and cross-horizon sections of the right hand side of this book. Throughout the long period of the excavations these catalogs were continually being revised and expanded to accomodate our constantly growing perceptions of the shapes and wares. For the ware types over the years a number of specialists worked on them: Marilyn Beaudry, Yoko Taniguchi, Marianna Nikolaidou and Ellery Frahm (see footnotes in our articles of 2008 and 2015).

Each contributed their expertise and experience to the ware descriptions as they had already been established in the field.

Back to top: Methodology and Process

Colormod project

Giacomo Chieri of the Gettty Conservation Institute shared with us his software for the analysis of sherd sections in the field so that a much larger dataset could be analyzed. This involved cutting the sherd sections with a high-speed saw and then further sanding them smooth when necessary. The resulting sherd sections are very flat and then scanned, in the field. At this point they could be used by the Chieri program; the publication of this large project is forthcoming by Marianna Nikolaidou and Marilyn Kelly-Buccellati.

Back to top: Methodology and Process

Update processes

As a preparation for the opening of this digital book all the shape catalogs and ware types have been reviewed and adjusted to reflect our current thinking. All the changes of shape through the various stages of the review process are shown in the Shapes Conversion List with the final descripton of types given in the Shapes-Codes page.

Back to top: Methodology and Process

Sources of problems

In the field the methodology was clearly defined but some problems were identified during the work. Mostly this was a single mistake but others were more apparent. To name the most apparent: skipped or mislabeled q-lot numbers, duplicate p numbers, or skipped numbers within the p number sequence for the q-lot, mismatch between the rim diameters determined in the field and those determined by the draftsman who drew the sherd, mismatch between numbers on the drawing and ceramic analysis numbers. For the most part either we were able to resolve the problem,either directly in the field or afterward, or we discarded the data.

Back to top: Methodology and Process

Physical arrangement of the sherd yard

The sherd yard was next to the incoming area and near the main laboratories in the exavation house. It consisted in a paved patio-like area covered with a tarp to avoid glare and to cut down on the heat. In this area were a number of work tables where the sherd analysis was done. The washed sherds were placed on a metal drying rack near this paved work area. When the analysis of the shape sherds was completed the sherds were replaced into the sieves and taken to the drafting area inside the sherd library room. Here the sherds chosen to be drafted were drawn in pencil and then immediately a vector drawing was produced. If the sherds were to be photographed they then went to the photographer. At the end of the analysis phase the sherds were either placed in one of the in-house displays or were placed in white styrofoam cartons by area and feature and permanently stored on metal shelves organized by excavation unit in the large expedition storeroom attached to the southern side of the expedition house.

Back to top: Methodology and Process

Three display areas

In the exavation house there were three display areas having different puropses, audience, sizes, functions, and time periods during which they were available.

Back to top: Methodology and Process

Ceramic overview display

The expedition house has a meeting room which is furnished with a large dining room-type table, a couch and a number of chairs. It is in this room that staff meetings took place; guests were received in this space. A part of one wall contained five well constructed shelves. On each shelf a few typical ceramic sherds from a particular time period we exavated at the site were displayed. During each time period ceramics had specific characteristics that allowed the archaeologists to identify the periods they represent. This display was purposely kept small and only changed if some especially beautiful and typical sherd was found in the excavations. This new sherd would then replace another one in the display so that the overview aspect was always retained.

Back to top: Methodology and Process

Sherd walkway display

Just outside the kitchen/dining room was a long and wide outdoor corridor which led from the work rooms and laboratories to the dining room. Because the excavation season was very hot, this walkway had a canvas cover to protect it from the sun, thus reducing the heat along the walkway. On one side of this walkway was a wall about a meter in height that protected the walkway from the courtyard which was on a slightly higher level. The wall was topped by a wide shelf-type platform, also about a meter wide. On the opposite side of the walkway there was room for a series of tables. Displayed on both the tables and on the platform were the analyzed shape sherds from importtant features from units we were at that point excavating. The primary audience for this display was the archaeologists who would be able to readily examine the ceramics they were discovering; the display would also indicate to them the characteristics of the ceramics in the time periods of their excavations. The large sidewalk display was also for visiting archaeologists to be able to inspect in detail the ceramics from the periods we were excavating. This display had usually a large number of sherds as all the shape sherds from important features were put out in the display. The display sidewalk changed throughout the season.

Back to top: Methodology and Process

Sherd library

A large room in the excavation house was dedicated to the Ceramics Library. The library is a room approximately 5 meters long and three meters wide. Metal shelves were installed for the length of the room on both sides. The Library contained not only the best examples of a type, either shape or ware type, but also encompassed their changes through time. This was especially helpful in the case of the ware types for which there was a marked continuity but with gradual changes over the time periods we were excavating. The largest and most extensive display of sherds and partially reconstructed vessels is contained in this library. The first shelves opposite the door contained typical examples of all the ceramic ware types and their variations through time. This collection was kept in small white boxes that could be transported to the sherd yard when needed. Also on these shelves were various shape types, placed in small boxes for easy transport to the sherd yard. Other shelves were dedicated to a display of typical sherds for each phase excavated in Mozan/Urkesh (see here). A large number of shelves contained shape sherds from the most significant features excavated. This library was extensively used for the analysis of the Mozan/Urkesh ceramics but also used by other excavators in the region for comparative analysis especially to help them date their finds.

Back to top: Methodology and Process