2. Architecture

Giorgio Buccellati – February 2007

2.1 The articulation of the revetment wall

While originally understood as an oval, it is now clear that the revetment wall of the Terrace is polygonal in shape. The bend in the revetment wall at its southwestern end shows clearly (Fig. 3) as it takes a sharp bend northward. The top of the wall is at the same elevation as the rest of eastern portion, showing that the degree of preservation remains the same as excavations proceed westward. We consistently have the top of the wall as originally constructed.

The second main objective for J1 is to reach the third millennium floor of the Plaza. To accomplish this, we had to widen the total area of excavation in order to have sufficient space to open an area of about 5 x 10 meters to the base of the wall (Fig. 2).

This expanded the excavation area to such an extent that in the end we were not able to reach the desired depth. Another important reason for the delay was the fact that as we reached the earlier second millennium levels, the accumulations became more significant. While in the later periods (Mittani) we have primarily a series of sediments that have washed down from the built up areas surrounding the Plaza, in the earlier periods (Khabur or Old Babylonian) we have a more distinct cultural buildup of the strata, with better defined floor surfaces and heavier concentrations of sherds and animal bones. Fig. 4 shows details of such surfaces in the late Khabur period.

2.2 The main staircase

2.2.1 The level of the Plaza

2.2.1.1 The reduced monumental access

The first goal in J3 was to expose the entire secondary apron (Fig. 5). Or that is what we considered this component of the staircase to be: a stone frame for the upper part of the staircase, mirroring the primary apron which flanked the lower part of the staircase. We assumed, in other words, that the lower frame (the primary apron) and the higher frame (the “secondary apron”) had been conceived together as an integral part of the construction of the entire complex, i.e. the staircase and the revetment wall. In reality, the 2007 excavations have shown that the higher frame is a later addition that ignored the existence of the revetment wall and of the staircase. It rather imitated their function in a general way, i.e. as a boundary to the Terrace (we might call it a “hinge” between the slope and the flat area), and it served at the same time as a reduced frame for access to the Terrace. In other words, it was built after the staircase and the revetment wall were no longer visible, hence independent of them. The evidence in support of the new interpretation is as follows.

The wide band of stones which at first appears as a wing connected to the monumental staircase is in fact separated by a side gap both to the east and the south (Figs. 5-6). The triangular effect (with the acute angle to the west), which also seemed to support the interpretation as a wing or secondary apron, may be explained as a way of emphasizing the access where the slope was at its highest point, i.e., in the east. Here, the bricks uncovered in B6 were still visible in the latest phase of the Terrace use, and thus the triangular stone structure serves as if to point in the direction of that line of access to the Temple.

One additional reason for this interpretation is that it appears now that the old staircase was in fact blocked by a mudbrick wall (Fig. 7, in J2). Such blocking would have redirected to the west the access to the upper part of the Temple Terrace. This may have resulted from the building development to the east of the staircase.

2.2.1.1.1 The “memory stones”

When the sediments above the Plaza reached the top of the revetment wall, (in the latter part of the Mittani period), there was no attempt at raising the wall itself. This may be indicative either of a lack of resources to undertake a major renovation project, or of the fact that raising the wall would have lessened the visual impact of the slope leading up to the Temple, or of course both. Whatever the case, the flat surface of the Plaza came at one point to coincide with the base of the slope of the glacis. At this juncture, individual stones were placed in a loose row just within the perimeter of the revetment wall, to mark the boundary between the flat area and the slope, as if a symbolic hinge that retained the ideological, if not the structural, valence of the boundary between the two spheres. These we have called “memory stones” (Figs. 8-10), referring to the memory of the ancient revetment wall.

2.2.2 The top surface of the Temple Terrace

2.2.2.1 The final tell surface

In our search for a clearer definition of the glacis, we have identified a later (mid second millennium) mud surface that follows closely the slope of the original (third millennium) baqaya glacis, and which seems to be the latest moment of the sacral use of the Terrace (stratum 11). Immediately above it, there are outdoor floor levels with a tannur and pits that belong to the subsequent, non-sacral phase (strata 6-8, Figs. 9-10).

Eventually, even the memory stones were covered by the sedimentation generated by wind and rains, and the tell assumed the shape it has today. The hinge between the ancient Plaza and Terrace was now represented simply by the difference between the sloping and the flat area. The Terrace had simply become the topmost hill on the tell, with no sacred meaning attached to it any longer. But – for our good fortune – the situation that followed the main use as a sacral area was such that no intrusion ever took place. Abandonment did for the site after 1250 B.C. what sacrality had done for it before that date.

2.2.2.2 The glacis below the reduced monumental access

The reduced monumental access (previously understood as a secondary apron, see above 2.2.1) rests immediately above the slope of the (third millennium) glacis that leads up to the Temple. We have clear evidence of this in a few places where both the baqaya surface and the bricks that covered it are still visible. The few sherds recovered from this small area include a considerable number of phase 3, which is in perfect keeping with what else we know about the Terrace in general and the glacis in particular.

2.2.3 The rule of asymmetry to the east of the staircase

We opened 8 squares in J4, with the goal of reaching a level equivalent to the western primary apron. Our assumption seemed inescapable, that there should be to the east a specular version of this apron, and possibly even of the secondary apron.

The new interpretation of the secondary apron, resulting from the excavations in J3 (1.3.3 and 2.2.1), first suggested otherwise. Since the presumed upper apron was in fact a reduced monumental access, which did not flank the staircase, but rather replaced it; and since in J4 a mudbrick wall effectively blocked the original staircase – it seemed clear that access to the glacis had been, in the later periods, deflected to the west, thereby lessening the significance of the centrality of the staircase itself as it related to the Temple.

Even so, the existence of an eastern primary apron could not be immediately excluded. The main supporting argument was the oblique line that already had been exposed in part in an earlier C2 trench, an oblique line which continues strongly to the south (Fig. 11). This line is a clear mirror image of the one to the west, and the overall configuration is such as to imply that it had been conceived as a single structure from the beginning. But if so, it seemed equally “certain” that the oblique line should continue south to a point where it would match with full symmetry the oblique one already exposed to the west. But symmetry does not seem to have been a Hurrian aesthetic canon. For it seems now just as “certain” that there was no matching eastern apron. The reasons are two.

First, there is a wall that sits astride the staircase itself (Fig. 7). This wall had been exposed last year already, but we assumed, too hastily as it turned out, that it was a later addition. This year we decided (at the insistence of Federico Buccellati, who felt that the situation may not be as we had concluded last year) to probe more closely the base of the wall. And it turned out that the wall was in fact bonded with the staircase itself, thereby indicating that it was part of the original construction. But if so, this strongly suggested that there was no matching apron on the other side, and that the wall was the eastern boundary of the staircase itself, continuing, as it were, the line of the revetment wall – but jutting sharply to the south.
Second, a sounding just to the east of this same wall found no trace whatsoever of the steps that should have been there had the staircase been constructed symmetrically. The excavations went well below the level of the steps immediately to the west, and there we found immediately levels dating to the fourth millennium.

Now, it is still conceivable that an original symmetrical eastern wing of the staircase may have been removed in antiquity, and that the north-south wall was bonded at that point in time with the portion of the staircase that had been left to the west. But this would have had to have occurred at a time when the full staircase was still entirely visible, i.e. the early second millennium at the latest. It seems strange that (a) a visible monumental staircase should have been so curiously halved, and (b) that fourth millennium layers should be found at a level equal or even slightly higher that the presumed matching steps to the west. To these questions we should be able to find an answer in the next season of excavations.

2.3 Conservation

2.3.1 Wall conservation

During the 2007 season we made significant progress in the documentation of our wall conservation efforts, which is a mainstay of our overall project. Two Italian graduate students, Beatrice Landini and Marta Lorenzon, took over the multiple tasks involved, to wit:

continuing the photographic and descriptive monitoring of the Palace wall, and harmonizing the earlier stages of the attendant documentation;
overseeing the repairs and maintenance of the Palace walls and of their protective shelters, including the implementation of a new type of roofing for the same shelters;
developing strategies for the protection of sections;
starting a monitoring system for the conservation of the stone architecture, includ-ing the repair and maintenance of the portions already exposed;
completing the documentation of the test walls constructed in 2004 as part of spe-cial study for the Getty Conservation Institute.

A full separate report is in preparation for these various activities.

2.3.2 The conservation lab

Beatrice Angeli, of the Opificio delle Pietre Dure, joined us in the field to complete the conservation work on a group of metal objects from previous seasons which we were then able to turn over to the Der ez-Zor Museum.

Her contribution was also invaluable for the stabilization and conservation of the frit necklace describe above (5.1).