In 1987 I became interested in the possibilities of creating a realistic reconstruction of some or all of the known structures in the Hellenistic city of Ephesus. At that time such an effort required computer resources beyond the scope of most colleges and universities; so I organized a small startup company, sold some shares, and began exploring how we would set up the operation. The objective was to create an accurate and realistic visualization that the visitor to the site could use as a reference to the actual ruins nearby. We planned to recover the costs of development and make a profit by selling admission tickets.
It was very important to me that the effort should be based solidly on the 100 years of excavation reports of the Austrian Archaeological Institute in Vienna, and to a lesser extent on the reports of others who have excavated there from time to time. I was convinced from the start that the effort could enlighten the research and study of the ruins at least as much as the reports could enlighten the visitors to the site.
That effort is in the past now, due to failed negotiations with the Turkish Minister of Culture early in 1993 and our decision not to continue further. It came to mind again as I read the discussions in the last issue of the CSA Newsletter, "AutoCAD As An Exploratory Device" and "Photorealistic Visualizations May Be Too Good." In the course of our work at Ephesus, we developed a 6-minute videotape of the interior and exterior of the Justinian church of St John, a structure well known to architecture students as a model of sixth-century Byzantine church architecture.
We built the structure in AutoCAD®, following the designs of the Austrian report on the building, Forschungen in Ephesos Band IV Heft 3: Die Johanneskirche, which was published in 1951. There turned out to be some minor but interesting inconsistencies between drawings in that publication, chiefly between the sections and the perspectives. We chose to follow the more detailed sections, taking the perspectives as impressionistic.
The exterior of the structure was believed to have been very plain, and we easily portrayed it without much texture on the walls, with a tile roof, and with a black lead dome. Once we had chosen a date for the portrayal (ninth century), the issues of renovation and alteration evaporated. Aside from minor problems of portrayal of the tile roof in moving images, the exterior was a very straightforward presentation.
The interior, however, was a problem of a very different scale. The interior walls, with rich mosaics and frescoes, were lost when the structure was destroyed by Tamerlane or earthquake in 1402. The presence of millions of tesserae in the rubble, samples of gray-and-white streaked marble on the piers, and large expanses of geometric mosaic patterns in the floor were recorded by the excavators.
|Computer Reconstruction of the interior of the Justinian church of St John at Ephesus. (Copyright Computer Visits, Inc.)|
There is also a written record of the interiors of churches throughout Asia Minor, dating from somewhere between the seventh and the late ninth century. A pilgrim, Agathias Scholasticus, traveled through the region, noting what he saw as he went from building to building. Though there are no chapter headings or other organizational identification aids, the record contains a list of 56 observations which begins with the dedication of the church and continues to the dedication of another church. Scholars have generally concluded that the comments represent either legends on the images in the church or the pilgrim's thoughts concerning the images. We used this list as the only available evidence about the interior imagery, and we assumed that all of the pictures were mosaics.
In the task of reconstruction, we then had two major problems. First, it was a long leap to go from the information that a certain image was about Hagar, for example, to having an actual image. Second, once we had a suitable mosaic of the sixth century, exactly where in the building was it placed? The list's order seemed to be random: it began with Nativity imagery, then moved erratically through the Gospel stories, and ended with a list of twelve church leaders. How should we decide about placement of the pictures?
I obtained representative images by scanning the Index of Christian Art at Princeton for sixth-century mosaics and then by relaxing the time period to extend to the eighth century. As it happens, over two-thirds of the themes in the church had been treated elsewhere, generally in Ravenna and Syracuse. For our purposes, developing a proof-of-concept, this was close enough. In an actual full-blown reconstruction, we would have created original art in the style and iconography of the time, and we planned to process the images electronically so they would appear to be mosaics.
This left me with the mapping problem. Where did the images belong? I consulted art historians here in Cambridge and Boston as well as theologians in New Haven and Cambridge. There was no general pattern across churches at that time to be used for standardized treatment. There was no theological rationale that we could count on, except that certain images in the list could be associated with the exit to the baptistry and the stories in which women were prominent could be expected to have prominent positions near the altar. I had a preference for placing the twelve church leaders, including three of John the Apostle, in the apse at the east end of the building. (The commentary of Agathias Scholasticus indicates images of John as author of the Gospel, as author of Revelation, and as author of the Gospel again.)
It was interesting to me that none of the Austrians, none of the art historians, and none of the theologians I consulted had given much thought to this matter of image placement. After a while, it occurred to me that the only organizational information I had was in the original order of the listing in the pilgrim's account. Could the listing be in the order he came upon them as he walked through the building?
I tried it out. Enter the nave from the western end, and walk down one side. If the twelve leaders were going to be at the east, this meant either no images on one side, or random crossing from the north to the south sides of the nave. Nothing worked very well. It all seemed very random.
Several months later, as the project was moving to the point of actually placing photo textures in the interior and I HAD to find a solution, I looked again at the plans of the building. A fact appeared that had escaped me before. We were drawing the sixth-century building, but the pilgrim had come centuries later, after some architectural changes at the west end of the building. When he arrived, the only entrance from the town into the building complex was through one of two doors in the south transept. The western entrances opened onto a balcony by that time and did not provide either a ceremonial or a working entrance. The pilgrim could only have entered through the south transept!
This led to a new approach to the mapping. Assume he entered, looked around the south transept (Nativity scenes), then turned left down the nave to the end, back the other side, and then jumped across to the apse. There were just enough images to fill the spaces between arches, plus a few for columns and for second-story and ceiling locations he might have noticed. The scene over the door to the baptistry was likely the Ark passing over the Jordan, and the other scenes fitted in well enough.
The new placement, however, did not put the images of John, John, and John in the center of the apse. If we followed the strict order of the pilgrim's account, they were on the north side, following the Apostles Peter and Paul. This resulted in the placement of Matthew and Luke at the center of the apse. I had wanted the John images to be in the center and considered this discovery a flaw in the theory.
A Byzantine art scholar pointed out that it was common at that time to set mosaic tesserae to catch the light from the sun, passing through a high window. Perhaps they would be gold tesserae, set to reflect the sun's light from a certain narrow window onto the altar at noon on a certain day of the year, a festival day perhaps. This could not be accomplished if the images were located on the east or south walls, but only on the north wall. Perhaps there was a reason for the position of the three Johns after all.
Despite the intriguing characteristics of the mapping, it is not clear that we found the true location of the images in the basilica. I believe they should be portrayed in reconstructions in a way that indicates uncertainty and have suggested that showing transparent images will accomplish that. If the viewer chooses to click on them, some version of this reconstruction story could be told. Alternative mappings could be shown.
The use of transparency or of pastels is only one way to indicate uncertainty in a reconstruction. In cases where more than one well-formed model of the original structure has been developed, more than one can be available to plug into a larger setting. Scholars or others interested in the controversy should be able to access a discussion of the alternatives and of the assumptions that went into the presented versions. This can be accomplished in a scholar's layer of the presentation, separate from the tourist or high school layer.
This concept of presentation layers is exemplified in a test web site that I have prepared on the Sheeps Pool/Bethesda/St Anne complex in the northeastern quadrant of the Old City of Jerusalem. The site is not available to search engines because I have not sought copyright clearances for the work yet, but it is available to readers at www.ultranet.com/~grovesa. In the site we present a tourist summary layer, eventually with a photo drum to introduce the present-day site and orient the web visitor. Text references lead to more complex explanations, maps and photos of features of the site at different periods and eventually to a history of the excavation findings and interpretations. Some pieces I intended to construct have not been completed, including an orientation map of the Old City and a time line of dynasties and local rulers.
- Allan Groves
For other Newsletter articles concerning the use of electronic media in the humanities, consult the Subject index.
Next Article: Documenting the Older Propylon File
Table of Contents for the Winter, 1999 issue of the CSA Newsletter (Vol. XI, no. 3)
Table of Contents for all CSA Newsletter issues on the Web
Return to CSA Home Page