Chapter 3
APPROACHING THE ROMANO-EGYPTIAN HOUSE
Research Methodologies
Anna Lucille Boozer
In this chapter I review the excavation strategy and methodology, as well as the post-excavation approach to analyzing finds from House B2. I begin with an overview of B2’s location. Then I discuss the excavation strategies employed in order to achieve the project objectives and our methodological approach to the excavations, including an overview of how excavation proceeded on a daily basis. Finally, I consider the post-excavation analysis of objects, contexts, and plans leading up to the publication of this volume.↩#p119
The Excavated Area: Location
From the surface, the northeastern portion of the site (Area 1) appears to be a distinct occupational area. The main road in Area 1 is off axis from alignment with the rest of Roman Amheida, and the surface ceramics are slightly earlier in date than in Area 2.1, reflecting a predominantly late third century date rather than a fourth century date.↩#p120
When we commenced the excavations for House B2 in 2005, we had already begun excavating House B1 in 2004. B1 is a large, lavishly decorated house close in proximity to what appears to be the urban core of Amheida. The excavations at B1 revealed a great deal about the identity of the inhabitants and their preference for Greek and Roman material and visual culture. The surface remains in Area 1 suggested to us that we could find less wealthy families in this area of the site and would be able to determine material differences between these people and those who lived in Area 2. Upon excavation, it became clear that there were very different states of preservation between B1 and B2. The walls of B1 were preserved to a greater height than those of B2, which makes it easier to reconstruct the architecture of B1. On the other hand, we found that soft organics were preserved better in B2 than in B1 and that there appeared to be a higher density of occupational debris in B2 than in B1.↩#p121
Synopsis of the Excavations in Area 1
The DOP conducted the first archaeological work at Amheida in 1979 during a brief exploratory season, concentrating on two areas of the site, Area 1 and Area 2. The Area 1 work focused on ceramics and kilns, as mentioned in the previous chapter.↩#p122
The Columbia University (now New York University) Amheida Project conducted brief exploratory work, consisting of test pits and survey in Area 1 in 2000 and 2001 as well as survey in 2002. In 2004 some preliminary surface clearing took place in the eastern portion of the area in a building (Area 1.1) as well as a small portion of the street (S1) in order to assess the stratigraphy of Area 1. Area 1.1 had eroded to such a degree that only the foundations were visible. Survey continues to define this area of the site and all of the survey data is collected in CAD. As of 2013, the entire visible habitation area of Amheida had been mapped, mainly by Fabrizio Pavia.↩#p123
Systematic excavation of House B2 began in 2005, was expanded in 2006, and completed in 2007. This excavation involved the total excavation of a single house (2005–2006), a trench in Street S1 (Area 1.2) in front of the house (2007), and a trench (C2A) in Courtyard C2 (Area 1.4) behind the house (2007). We expanded our excavation to areas external to House B2 in 2007 in order to compare the stratigraphies of these areas with those inside the structure. We also wanted to determine activities that took place around the house in order to inform interpretations of interior activities. Our first exterior excavation took place in the street just north of the house. We laid out a 6.10 x 2 m east–west oriented trench in this area, aligning it against the north wall of B2. In this manner we also exposed the exterior face of the structure’s north wall. Likewise, we explored the exterior courtyard behind B2. This exterior courtyard initially appeared to be a large L-shaped open space (20 x 21 m) south and east of B2. We laid out a north–south oriented 10 x 2 meter trench against the south wall of House B2 and the west wall of Couryard C2. This trench was later expanded by an additional 1.5 x 1.5 meters to the east, against the south wall of Courtyard C2, in order to allow us to expose some features more fully. The trench covered the entire length of the wall that borders the courtyard to the west and its southwest corner.↩#p124
Excavation Strategy and Theoretical Perspectives
The objective for the excavation of structure B2 was to examine the daily life practices of a single household from Roman Egypt. I pursued this goal, as chapter 1 explains, because past approaches to domestic contexts in Roman Egypt have revealed very little about how individual households functioned in different socio-economic groups, ethnic groups, or regions. A contextual approach to material culture from domestic contexts will help fill this lacuna and will enable scholars of Roman Egypt to engage with other Romanists, Egyptologists, and other social theorists through new research questions that extend beyond traditional disciplinary boundaries.↩#p125
A contextual approach affected all stages of our research equally because attention to the context permeated the excavation process, not merely our study of the material in view of publication. Although this approach dominated my own research initiatives, I recognize that it is important to remain cognizant of the diverse range of methods that archaeologists have used in order to comprehend different categories of data. For example, archaeologists have typically analyzed categories of data on the basis of material types (e.g., ceramics, texts, or depositional information) and these close analyses are of great benefit to specialists as well as individuals who wish to delve deeper into specific research questions. This volume uses both thematic and material category modes of analyzing archaeological data in order to view objects from a variety of perspectives and to capture a broad range of relational significance. In this sense, I attempt to analyze the material not only in terms of its social value but also its typological significance.↩#p126
Excavation Methods
With terminological modifications made by Paola Davoli, the archaeological director of the project, the excavations employed two standard means of capturing both the horizontal and vertical components of archaeological material: the Museum of London Archaeology (MoLA) system of excavation and the Harris Matrix. These methods were employed from 2004 at Amheida and also have been employed at the site of Ain el-Gedida. MoLA and the Harris Matrix offer valuable methods for understanding a complex archaeological site because they view the site as a palimpsest of discrete events that come about through either deposition or removal. These methods enable us to understand the sequential development of a site with maximum control over stratigraphy while simultaneously enabling us to view the archaeological record as a dynamic, creative process enacted by individuals.↩#p127
The MoLA system describes methods for recording single contexts (stratigraphic layers), which capture the horizontal component of archaeological sites in detail. In other words, the MoLA system helps us to understand activities and patterns that theoretically occurred within a single moment in time. Single context recording was developed as a means to avoid reliance on multiple-feature or composite plans (i.e., plans that record all or phased features respectively, on the same drawing). Single context recording recognizes that physical and stratigraphic relationships are not conflated but distinct. This method considers the stratigraphic sequence to be the primary consideration in the excavation and recording of sites. The basis of this approach is to divide up the site into discrete units or contexts that are excavated and recorded according to their stratigraphic position. In the words of Westman: “Any single action, whether it leaves a positive or negative record within the sequence, is known as a ‘context’.” One of the crucial assumptions of single-context recording is that each unit represents a single action or event such that its spatial limits also define temporal limits. An action, in this sense, denotes activities such as repaving a floor, repairing a wall, or even more ephemeral practices such as baking bread in an oven.↩#p128
The Harris Matrix enables us to understand the vertical, or temporal, component of the excavation area as well as how disparate locations within the excavation area relate to one another. Following the Harris Matrix, we used deposition stratigraphic units (DSUs) to build up the stratigraphic history of the house through comparing the stratigraphy of rooms. We did the same for the street (S1) and the courtyard (C2) and then compared these stratigraphies with those in House B2 in order to understand phasing. The conceptual foundation of the Harris Matrix is that any stratigraphic unit can have only three possible kinds of direct relationship to another: none, earlier/later than, or equivalent to. In practice, contexts may have indirect relationships with other contexts, as well as more direct relationships. The nature of this indirect relationship cannot be determined stratigraphically but must be discovered by other means, which involve grouping (or phasing) the matrix into broad divisions, which are approximately contemporary. Ideally, we can conduct grouping through major horizon breaks in the sequence and/or independent dating through finds association, such as datable objects (e.g. texts and coins). Essentially, the notion of grouping relies on permutations in the matrix, which occur in all multi-linear sequences.↩#p129
In terms of daily practice, we organized our excavations first by room and then by visible context matrixes. A supervisor controlled all work in the area, while an assistant supervisor and several local workmen were allotted a room. The area supervisor, with help from assistants, was in charge of the area documentation. Whenever it was advantageous to attain greater control over the vertical component of the excavation, we subdivided DSUs into arbitrary DSUs. For example, in situations in which we found windblown sand to a significant depth, we wanted to distinguish surface sand from the lower layers, although the matrix itself was identical. Whenever it was critical to attain greater control over the horizontal component of the excavation, we gridded the space and applied different DSU numbers to different portions of the grid. For example, some of the floor layers that we reached contained high densities of ceramics that were easier for the ceramicist to analyze if we gridded the floor levels, although from a depositional standpoint the units appeared to be identical.↩#p130
The recording techniques that we employed for this project included standardized forms, photography, and drawn plans. We made use of each of these recording techniques for each context. We took starting and ending elevations of each context and object with an automatic level, while we made use of an EDM theodolite (Total Station) for some of the small finds and all of the architectural features. We drew sections in situations that would clarify the depositional history of the structure along two different axes across the house (north–south and east–west). We drew plan views of all of the walls in structure B2 in order to capture details of brick laying patterns. These details and other exposed features were added to the digital CAD plan of the site.↩#p131
Each day we made 1:50 scale day drawings of the excavated area, along with the location of small finds and starting and ending elevations. We used 1:20 scale drawings for each context to show a greater level of detail than our day drawings. Likewise, all archaeologists completed day notes forms, which recorded in narrative form the events, finds and decisions of the day. We used pro-formas to describe each DSU and FSU, which we entered into the relational project database upon completion of these units.↩#p132
Almost every day of excavation at this structure yielded a high density of objects. We treated all objects that were discovered and recognized as objects during excavation in the same manner. When an object was found, we gave it a unique number to identify it along with a brief description, recorded the exact find spot with respect to the x, y, and z axes, photographed it, and included the object on a context plan. In other words, we precisely located and documented all of the small finds in three-dimensional space within each context. These data enable us to determine the immediate environment from which the objects derived. We took soil samples of each sealed floor context and other sealed remains rich in organic materials for archaeobotanical analysis. These samples were also located within three-dimensional space and noted on plans.↩#p133
When we reached floor levels with particularly dense artifact densities we used a ¼ inch mesh screen for sifting 100% of the dirt from the DSUs. Some of our objects were found in the screens, and therefore these objects can be identified by context but not by precise location in three-dimensional space. It was not feasible or meaningful to screen windblown sand above these levels.↩#p134
On-Site and Post-Excavation Architectural and Object Recording
It is important to underscore that fieldwork does not simply entail the recovery of objects, designation of depositional units, or description of features. Fieldwork is an interpretive exercise during all phases of research, and we were mindful that it was important to connect theoretical perspectives with data collection during excavation, rather than only as we reached the brink of this final publication. In order to more fully unite our theoretical and methodological approaches we engaged—as much as possible—in reflexive archaeology.↩#p135
A reflexive approach to fieldwork methodology enables an interactive approach to the relationship between finds analysis and the determination of archaeological contexts. Reflexive archaeology encourages constant feedback, rather than a linear progression from excavation to finds processing to analysis and finally synthesis. In order to engage in constant feedback during excavation we encouraged close personal interaction between the excavators and the specialists so that finds analysis could inform excavation while it was still in progress. Second, we employed a relational computer database, which enabled us to enter most of our data during excavation so that we were able to compare many different classes of archaeological data during the course of our investigations. These data included digital forms, scanned paper forms, photographs, day notes, day drawings, and digital plans. The ‘real-time’ access to data, even when only roughly processed, encouraged a more informed excavation methodology than a linear progression of analysis. As is to be expected, it was not always possible to fully enter data or produce exhaustive analyses simultaneous with the excavation, but we found that even partial records helped us understand our excavation better as it took place, rather than only in retrospect.↩#p136
Off-Site Processing and Analysis of Objects
The objects from each day’s excavations were transported to the registrar’s workroom in the excavation house, at which point the registrar and registrar assistants recorded them in more detail and assigned them unique inventory numbers. Each object was cleaned as much as possible for recording purposes. Some objects were restored or cleaned by a professional restorer, while others required only minimal cleaning by the registrar. Once cleaned sufficiently, the object was fully described on an object form, which was then added to the project database. Staff analyzed the objects for form, phasing, dating, associations, and other attributes. The objects were also photographed and, in some cases, drawn.↩#p137
All of the ceramics were analyzed on-site in order to categorize the fabric types and sort the diagnostic sherds from the undiagnostic sherds. The diagnostic sherds were taken back to the excavation house for additional study. These sherds were then catalogued and a classification for Amheida was slowly built up over the years to interpret phasing and dating associations.↩#p138
At the end of each season the Supreme Council of Antiquities (SCA) inspectors selected significant objects for registration in the SCA magazine. These objects were given an additional recording number by the SCA and are now kept in the magazine near Ismant el-Kharab, operated by the SCA.↩#p139
After the initial recording of the objects, specialists revisited them in subsequent field seasons (2008, 2009, 2010) for more thorough analyses, and many of the objects were then drawn. All original paperwork for the excavation and recording of B2 is stored at the project archive located at the Institute for the Study of the Ancient World, New York University, New York City, New York, USA.↩#p140
Object Location, Storage, and Conservation
All objects recovered from the Amheida excavations, with the exception of those registered by the SCA, are stored in the Ain el Gindi Excavation House magazine (but not individually registered). All objects were individually packed, according to material and fragility and are separated by the excavation season and the structure excavated. The objects are wrapped and placed in rigid plastic boxes of varying size, according to the object size. These boxes are labeled with the inventory numbers assigned to them by the registrar.↩#p141
Object Description
We analyzed the objects for form, phasing, dating, associations, and other attributes. The objects are given specialized treatments in subsequent chapters so that they can be easily accessed for in-depth study. Each of these object categories contains its own methodological perspectives for analyzing the material, and therefore these methods will be presented along with the material. Although we conducted specialized object studies, it was imperative that we fully examine and consider each object in context and as part of society in Roman archaeology. With this objective in mind, Chapter 7 introduces the small finds and various themes found between object categories, giving a general overview of the finds as well as how these finds relate to other houses. Likewise, I discuss the small finds within a more holistic exploration of B2 in the final chapter in order to explore the contextual significance in more detail.↩#p142
The general layout of the catalogue used by specialists is as follows:↩#p143
- Catalogue Number: the relevant number of the object for this publication.
- Figure Number: the reference to the relevant illustration of the object.
Amheida Inventory Number: the inventory number assigned to the object by the Amheida project once the object was brought in from excavations.↩#p144
- Context: the find spot of the artifact. If the artifact was recognized as an object during excavations, a specific location has been isolated for this object. Otherwise, the context refers to a specific matrix in the building, the street, or external courtyard.
- Material: the material or materials used for the objects are described, with details of variations and numbers. Some specialists relied on particular systems, which they explain within their catalogue.
- Dimensions: the measurements of the object. The expressions for these measurements vary depending on the class of artifact (e.g., coins are given diameters and thicknesses rather than length, width, and depth).
- Technology: the visible remains of methods of manufacture are noted, often in detail when the method applies to all objects in any one class. In other classes, where greater variety occurs, only general observations are made and details are reserved for the individual catalogue descriptions.
- Condition: denotes the level of preservation of the object.
- Description: a detailed description of each object is provided here, including visible details and comments on some of the more noteworthy parallels.
- Date: a general date ascribed to the object based upon comparanda and/or its provenance.
- Parallels: published parallels for the object class as a whole are presented here, while individual parallels are included after the object descriptions in the catalogue at the end of the section. General remarks concerning the significance of the presence, concentration or scarcity of any one object class are made, but again are only of a preliminary nature.
Object Illustrations and Indexes
We have illustrated as many of the objects from House B2 as possible in both line drawings and photographs. For some classes of objects, notably the figurines and objects of adornment, nearly all objects have been illustrated by line drawings. For other artifact classes, however, we have avoided repetition by the illustration of types that recur with great frequency, as is notable in the case of the ceramics. In other instances, some of the artifacts were too poorly preserved to be identified properly or illustrated by line drawing, but we have made an attempt to describe and photograph each of these artifacts in our catalogue in hopes that they will be of use when compared to remains from other sites. The desire to include these objects reflects a need to publish artifacts that cannot be properly labeled, which have often been ignored in past studies. All records for objects in the online version of this book are linked to the Amheida project database, designed by Bruno Bazzani, where readers may find more extensive information and illustration. This database is open for public use. All of the original documentation of the excavation of B2 and adjoining areas is available fully to readers wishing to see any of the elements of the documentation described above.↩#p145
The project database, available to the public, provides a complete list of all of the objects included in this volume, listed by the Amheida inventory number. The SCA registration number is given when relevant. This index lists where the object is mentioned and illustrated in this volume, as well as the type of object, its material, the context in which it was found and the level to which that context has been attributed in the Harris Matrix.↩#p146
Conclusion
The excavation strategy and the methodologies designed for this project encompassed every phase of research and we hope that they have resulted in satisfying analyses and writings on the material produced from this excavation. Subsequent chapters reflect our attempt to explore archaeological material in an accessible and multifaceted manner.↩#p147
