Mapping Petra

By: Paul Zimmerman

Originally Published in 2000

View PDF

In 1993 Martha Joukowsky opened a new exca­vation at the ancient Nabatean city of Petra in southwest Jordan (see box on Petra, p. 30. Jou­kowsky, Professor of the Center for Old World Archaeology and Art and the Department of An­thropology at Brown University, is the author of the definitive guide to field archaeology, and her goal at Petra was not simply to excavate but to showcase cutting-edge archaeological practices. To help meet this goal, the University of Penn­sylvania Museum’s Applied Science Center for Archaeology (MASCA) was contracted to provide technical services to the project. Thus three Penn anthropology graduate students—Leigh-Ann Be-dal, Loa Traxier, and I—began many years of work at one of the world’s great archaeological sites (see also article by Bedal in this issue).

I was hired in 1994 as a trained user of MASCA’s surveying system and charged with mapping the site that Joukowsky was investigating at Petra. SiteMap, as MASCA’s software is now known, was originally developed as a tool for easily creat­ing archaeological maps. It consists of a series of custom-built programs that link total stations to handheld data collectors to field computers. A novel way of streamlining the process of ar­chaeological fieldwork, this system nevertheless was primarily intended for in-house use. Over the years, however, it was adopted by a number of projects—at sites as varied as Abydos, Tell es-Sweyhat, Copan, Cochiti, Petra, and Poggio Colla—and through many seasons of experimen­tation and application the software tools have been greatly refined. At Petra, the ongoing development of the software and its practical application have greatly improved my own survey­ing expertise, and as the tools and my skills have developed, so has the level of interaction and interdependence. In excavating the Great Temple at Petra, the interplay between planning, excava­tion, and interpretation in many ways relies on these surveying and mapping tools (Fig. I).


Excavation of the Great Temple is geared principally toward understanding the structure as an architectural edifice. Placement of excava­tion units, therefore, typically follows architec­tural features (or their expected location). And though the trenches are oriented to the site grid. they do not necessarily correlate with grid squares. In order, then, to manage this excavation strategy and keep it from degenerating into chaos, our total station surveying has imposed a conceptual grid upon the site—thus recording the excavation in three dimensions, with great accuracy and year-to-year consistency. Within this conceptual grid, the location of each unearthed block is known, having been recorded upon excavation. Though tedious to acquire, the detail afforded by this block-by-block mapping strategy has paid off, since each year’s plans are automatically correlated with previous years’ and all data can be updated daily and printed and checked in the field.

Beyond simply recording our work, though. this strategy provides us with the wherewithal to explore possible architectural configurations and reconstructions. For example, the theotron (the­ater) built into the middle of the Great Temple posed an interesting problem: how does one esti­mate the size of the seating area where only a few intact rows of seats remain? Using our Site Map ­generated plans, the problem was solved. Given the width of the lowermost walkway, the position of the five partially preserved rows of seats, and the presence of sloping arched spaces behind the preserved seats (i.e., underneath the proposed location of missing seats), the exercise became a graphical one. By projecting a series of concen­tric circles to indicate the edges of the rows of seating, and by then elevating these circles (in the conceptual three-dimensional space afforded by the computer) so that each row was a set height above the row below, we could estimate the slope, spacing, and number of rows of seats (Fig. 2). That this projected layout works structurally with the other known architectural features—provid­ing clearance for interior staircases and rooms, providing a landing for a short and otherwise inexplicable staircase, and meeting the back wall of the structure strongly suggests that it is cor­rect. In this reconstruction, up to 600 persons could reasonably sit in the theotron. The type of event that they would have been viewing, however, is open to speculation.

Application of the surveying tools also yielded insight into the layout of another feature of the Great Temple, its so-called Lower Temenos. During the 1995 and 1996 seasons, excavations revealed bronze drains set into some of the hex­agonal limestone pavers of the Lower Temenos platform. One of the drains was found at the southwestern corner of the platform, and an­other was found along its eastern edge. No pat­tern was initially evident in their placement, but upon considering their location, we noticed that they were both equidistant from the expected lo­cation of the southeast corner of the pavement. Though one might expect a square platform based on this evidence alone, we knew at the time that the platform was longer than it was wide, because additional pavers had been found to the north, beyond the drain on the eastern edge. As the platform’s northern end was still buried, the actual length was unknown. Turning again to the computer map, a square was drawn whose diago­nal connected the two known drains. This square was then reduced 50 percent in each direction, over the plan. anchored at the southwestern cor­ner of the platform (Fig. 3). This grid revealed that the Lower Ternenos probably had propor­tions of 2:3, and would have originally spanned the area between the lateral triple colonnades on its east and west borders, and extended from the facing wall of the Upper Temenos platform at its south to a badly damaged wall on the north. The subsequent discovery of four additional drains at the locations predicted by the model lends sup­port to this reconstruction (Fig. 3). Unfortu­nately, although the reconstruction is quite plau­sible and is supported by all available evidence, the actual northward extent of the pavement has long since been robbed out and will never be known with absolute certainty.


In 1998, under the Brown University permit, the adjoining so-called Lower Market area was also opened for study. Mapping duties included extending our topographic map eastward from the Great Temple complex, and laying a sampling grid across the “Lower Market” for locating sur­face collections. After these duties were met, the surveying team became interested in documenting the numerous doorways in the so-called Portico Wall that faces the street running through the center of the city. The original reason for un­dertaking this side project was to understand bet­ter how the Great Temple’s lateral triple colon­nades met the Portico Wall and, in particular. whether one could access the spaces between those colonnades from street level. Our success in the reconstructions described above emboldened us, and we were reasonably sure that the answer would be revealed through a similar application of our mapping techniques. Unfortunately, we gained very little insight into this question, but the exercise was not worthless, since it brought to light some previously overlooked features that may change our understanding of the city’s layout.

Specifically, we noticed that the colonnade for which the Colonnaded Street is named does not, in fact, extend the entire length of the street. In fairness, this is not an entirely new observa­tion, having been noted in passing first by Diana Kirkbride and subsequently by Judith McKenzie. However, it seems that the true import of this fact was never previously realized. Though Kirk-bride found column bases in situ only along the stretch between the so-called Upper Market and Lower Market, she attributes their absence along the segment in front of the “Lower Market” and the Great Temple to shoddier construction here (presumably due to a lack of funds).

It is my contention that the difference be­tween the eastern (colonnaded) segment of the city’s central corridor and its western (wicolon-naded) segment is deliberate and meaningful. Instead of a portico-lined street unifying the city center from “Upper Market” to Temenos Gate, there would actually have been two roughly equal-length zones: one spanning the “Upper Market” and “Middle Market,” and one (lacking a colon­nade) spanning the “Lower Market” and Great Temple (see Fig. 5). Moreover, recent studies of the latter two sites suggest that together they constitute a palatial or administrative group (see Bedal’s article in this issue).

In support of my argument are the following points: (I) Whereas the curbstones upon which the colonnade is built are large blocks of high-quality sandstone with a gray or light purple color, those to the west are smaller blocks of softer red sandstone of a kind that is never (to my knowledge) used for stylobates (the paving on which columns rest). (2) Although most of the column bases on the southern side of the eastern segment of the street are in situ, no column bases or the char­acteristic weathering marks they leave are found along the western segment. (3) Whereas a second curb, also built of large blocks of high-quality sandstone, is found on the north side of the street, exactly opposite the in situ column bases, the curb opposite the western half of the street is made of the smaller red sandstone blocks that are inap­propriate for supporting a colonnade. (4) The westernmost column base lies exactly on the corner block where the curb turns southward, indicating that this is indeed a corner. (5) This corner is aligned with the eastern edge of the “Lower Mar­ket.” (6) The eastern and western segments of the street are of nearly equal lengths. (7) In the Great Temple, multicolored sandstone curbstones are used for decorative effect, and it is likely that the red sandstone of the western segment of the street was chosen to heighten the difference between its two halves (Fig. 4).

Despite Kirkbride’s note about the difference in character between the curbs along the eastern and western sections of the street, and despite the decades in which they were in plain view, the name “Col­onnaded Street” has stuck and, it seems, colored subsequent interpretations of Petra’s urban center. Reconstructions such as lain Browning’s, published in his popular book Petra (Park Ridge, NJ: Noyes Press, 1973), that envisions a portico-lined street are compelling but probably erroneous.

In light of this evidence, discussion of Petra’s center should no longer split the city into two halves, the Qasr al-Bint Temple precinct to the west of the Temenos Gate and the “Market” and Great Temple complex to its east. Now, it seems, a three-zone system is more reasonable: an inner (western) zone of religious activity, a middle zone of administrative and civic activity, and an outer (eastern) zone of, perhaps, eco­nomic activity (Fig. 5). Such a reconstruction is also supported by the chronology of major build­ing projects in each zone: the construction of the Qasr al-Bint Temple by the early 1st century CE leads the pack, the erection of the Temenos Gate follows in the late 1st century CE, and modi­fications to the Great Temple and the “Lower Market” in the early 2nd century CE. If, as I sus­pect, the modifications to these latter two struc­tures coincide with the construction of the Por­tico Wall and shops and the paving of the street itself, then these renovations would have carried significant political and cultural weight. If indeed contemporary, such a massive civic renovation project was probably undertaken shortly after the Roman annexation of Petra in 106 CE—cement­ing Rome’s ties to the Nabataeans. While a newly paved street would have physically unified Petra’s urban center, the partial colonnade would have compartmentalized it visually and conceptually.

Mapmaking is an integral part of any archae­ological excavation. The nature of that integra­tion, however, may vary widely. On the Great Temple excavations and the adjoining “Lower Market” survey, a tight integration of the survey­ing and mapmaking with the broader archaeolog­ical project has yielded interesting insights into the site and the ancient city. Surely, it is the favorable blend of excavators, technicians, and technologies that has made this particular map­ping project so fruitful.

Cite This Article

Zimmerman, Paul. "Mapping Petra." Expedition Magazine 42, no. 2 (July, 2000): -. Accessed February 27, 2024.

This digitized article is presented here as a historical reference and may not reflect the current views of the Penn Museum.

Report problems and issues to