Introduction
As previously outlined in our proposal for a Dumbarton Oaks project grant for 2004/05, the former church of Hagia Sophia in Vize (Turkish Thrace) has suffered dramatically from decades of neglect, vandalism, and a recent heavy-handed restoration conducted under the supervision of the Istanbul Vakιflar Bölge Müdürlüğü between 1979 and 1983. Given the importance of the church of Hagia Sophia as a pivotal monument in the history of Byzantine ecclesiastical architecture and its rather precarious state of preservation, a joint archaeological project has been established by the authors of this report, which aimed at a thorough examination of the building's fabric as well as a comprehensive documentation of the various architectural spolia scattered inside and outside the church. The first fieldwork campaign at Vize, conducted during the month of July 2003, resulted in accurate plans for the ground and gallery levels as well as longitudinal and transverse sections of the church. In addition, a catalogue of architectural spolia found in situ and scattered around the building was compiled and the fragments photographed and drawn in preparation for publication.
For the second season of fieldwork at Vize, we applied to the Turkish Department of Antiquities in Ankara for permission (1) to clean the u-shaped corridor around the church from brush and dirt that had accumulated there since the early 1980s, when the corridor was created; (2) to continue our survey work at the church with a photogrammetric evaluation of the building's exterior; and (3) to conduct limited archaeological soundings in front of and behind the present structure. While the proposed soundings east of the apse were aimed at exploring the relationship between the middle Byzantine church and its early Byzantine predecessor, soundings west of the narthex were aimed at clarifying the archaeological situation in front of the church in preparation for the insertion of drainage pipes—a measure that has become necessary due to severe water problems that have arisen since the Vakιflar "restoration."
Progress report
In contrast to our first campaign in 2003, fieldwork in 2004 started with a plethora of bureaucratic and administrative difficulties with the Turkish authorities. As a result, our first week in Vize was spent reviewing the secondary literature on the church, introducing students to archaeological survey methods and measuring techniques, and conducting excursions to related sites and monuments in Turkish Thrace. One of the positive effects of this week of intense preparation and study was the accidental discovery of several architectural spolia from our site at the museum of Tekirdağ, located ca. 80 miles south of Vize, on the shore of the Marmara Sea. These were documented, photographed, and drawn by members of our team after obtaining the necessary work permit and visas at the beginning of our second week in Vize. Unfortunately, bureaucratic problems continued even after the receipt of our work permit and visas from Ankara. While the work permit eventually granted by the Turkish authorities allowed us to continue the survey work begun in 2003, neither our request to clean the u-shaped corridors of brush, dirt, and garbage nor our request for limited soundings in front of and behind the church had even been addressed. Only during the last week of fieldwork in Vize were we able to obtain official permission to clean the corridors. Permission to conduct archaeological soundings was never granted.
Due to the requirements and limitations of our work permits, fieldwork in Vize during the summer of 2004 thus concentrated on four distinct areas: the photogrammetric survey of the church façades, the cleaning of the u-shaped corridors around the church, the completion of the catalogue of architectural spolia, and the rendering of a digital landscape model of the area surrounding the church. The results of the 2004 season of fieldwork can be summarized as follows:
Photogrammetric Survey
Using a combination of traditional and innovative architectural survey techniques (i.e., the tape measure system in combination with reflector and reflectorless tachymetry and digital photography), it was possible to render undistorted planimetric views of the building's main façades and to create an axionometric model of the church from the northeast. Instrumental in our effort to survey a large—and partly inaccessible—structure such as Hagia Sophia in a timely manner was the use of a reflectorless theodelite (Leica TCR 1 105 XR) in combination with a common digital camera (Nikon D 70) and the software PhotoPlan, an AutoCAD plug-in developed by Kubit GmbH, Dresden. The computer-generated views thus created of the four façades were printed out, checked against the fabric of the building and hand-drawn stone by stone. In a final step, the stone-by-stone drawings served as a tool to determine and record the various phases of construction, reconstruction, and restoration of the building. Once the stone-by-stone drawings are inked and digitized, the building's construction and restoration phases will be included in the final set of elevation drawings.
Cleaning of the U-shaped Corridors
A thorough cleaning of the brush, dirt, and garbage that had accumulated in the u-shaped corridors over the last twenty-five years was not only a prerequisite for the photogrammetry of the building's façades, but was also expected to yield valuable information about the early Byzantine church discovered and partially excavated during the Vakιflar "restoration" of Hagia Sophia in the early 1980s. Since this earlier structure had been left entirely undocumented by the Vakιflar architect/restorer, a cleaning of the corridors down to the 1983 cement floor was considered the first step in documenting the archaeological situation as it presented itself immediately after the end of the Vakιflar restoration campaign. As can be seen from the photograph taken at the end of the 2004 season, one of the most important results of the cleaning was the discovery of two sets of walls running parallel to the north and south aisles of the present church. During the Vakιflar restoration these walls were used as foundations for two large retaining walls flanking the church on to the north and south. As these walls closely correspond in their orientation with the round apse east of the present structure, it is assumed that they form remains of the outer side aisle walls of an early Byzantine basilica that preceded the present structure on the site. The cleaning of the main apse of this building, discovered during the Vakιflar restoration and first recorded in a sketchy drawing by Yιldιz Ötüken and Robert Ousterhout, allowed us to provide, for the first time, exact measurements and a ground plan for this structure.
Completion of the Catalogue of Architectural Spolia
Apart from the cleaning of the u-shaped corridors, limited cleaning in the church's upper storey allowed for the drawing of two transenna panels set between the tribelon arch in the west wall of the church. Other transenna panels from the church of Hagia Sophia, which were found at the museum of Tekirdağ during an excursion in early June, were also catalogued, photographed, and drawn for publication (these drawings are currently in the process of being inked). It is hoped that the completed catalogue of architectural spolia found on the site will not only enable us to compile an inventory of liturgical furnishings associated with the present and previous churches, but also to determine—through an analysis of spolia reused in the fabric of the current building—approximate dates of construction of the two churches.
Digital Landscape Model
In addition to the creation of an axionometric model of the present church of Hagia Sophia, a topographical survey has been started to render the acropolis of Vize with its Byzantine city walls and surviving ecclesiastical structures—the apse of a second, much smaller church or chapel has been discovered by Semavi Eyice southeast of the church of Hagia Sophia—in a three-dimensional digital model. The first step in this project was to explore the plateau on which the church of Hagia Sophia and its predecessor were built. As the remains of the smaller basilica discovered by Eyice are located on a higher level, an extension of the landscape model will not only allow us to determine the relationship between these two buildings, but also to reconstruct their placement on the acropolis in relation to the surviving system of walls and fortifications of the Byzantine and earlier periods.
Conclusion
As this brief report may indicate, the Dumbarton Oaks project grant 2004/05 greatly helped to advance the archaeological exploration, documentation, and analysis of the church of Hagia Sophia in Vize. Due to work permit restrictions, an archaeological exploration of the Early Byzantine basilica at Vize (as proposed in the 2004 grant application) had to be delayed in favor of a thorough photogrammetic evaluation of the façades of the Middle Byzantine church. The photogrammetric survey as well as the catalogue of architectural spolia associated with the church have been completed in this year's season of fieldwork. The results of the 2004 campaign will be published in Dumbarton Oaks Papers 60 (2006) (the fieldwork report for 2003 has just appeared in print as Die Hagia Sophia in Vize. Forschungsgeschichte—Restaurierungen—Neue Ergebnisse, in Millennium 1 [2004], 407–37).