4. Interpretation

The discovery last year of Early Chalcolithic pottery immediately below the glacis in J3 had led us to suggest that the material may have been brought in as fill from some other part of the ancient settlement. Now, however, it appears that this earlier occupation is in fact to be understood as a core situated immediately behind the revetment wall and below the glacis. We have in fact not only ceramics from three distinct places below the third millennium surfaces, but also sealings from pits cut directly into the glacis in J3. The details are given above in sections 3.2 and 4.2.

Here, four stratigraphic points need to be stressed.

1. The wall situated on top of the monumental staircase is in fact bonded (section 2.3), and hence contemporary, with the staircase itself. This suggests that behind it there is the same core that we found in J3. In other words, this wall may be functionally the same as the revetment wall in J1-J3-J2.
2. Such an interpretation is reinforced by the fact that the sounding we opened behind it (originally with the expectation that we may find the continuation of the apron and the steps of the monumental staircase) showed in fact a sloping surface that may correspond to just such an early glacis.
3. At the bottom of the staircase (in locus J2k100) there lurks an earlier staircase. The sherds associated with its last use date to phase 1 (ED III). While we could not date the time of construction, the presence of fourth millennium materials just below the surface in J1, J3 and J4 suggests that this earlier staircase in J2 may well date to the same period.
4. The nature of the revetment wall is such as to suggest that it framed a pre-existing glacis, and that the amount of earlier fill within the wall would be minimum. But such pre-existing glacis at such a high elevation implies that it would have served already a sacral function not dissimilar from the one of the later Temple Terrace.

The significance of these considerations is that from them we can infer a functional continuity between the earlier and the later uses of the area. In other words, the fourth millennium would already have seen a Temple Terrace just where the third millennium one stood, and practically of the same height above the surrounding plain. This is staggering because (a) it would push back the beginning of Hurrian history to at least the middle of the fourth millennium, (b) it would give evidence of a well established Hurrian sacral center at such an early date, and (c) it would give evidence for a full grown urban settlement capable of supporting such a shrine.

• Temple ovals: structural and functional remarks
• Temple Terraces in the Khabur plain
• Tell Chuera
• Tell Khazna
• Tell Banat
• Temple Terraces in the northern highlands
• Tepe Haji Nebi
• Mokha Blur
• Temple Terraces at a farther remove
• Monte d’Accoddi
• Greek acropolis

The picture that emerges from our excavations of Urkesh adds a significant new element to the typology of settlements in the kingdom of Mittani. A recent description of settlement patterns (see Akkermans – Schwartz 2003, pp. 346-348.) describes a territory characterized by isolated high status residences installed on sites that were by then largely depopulated, with an extensive ruralization trend (people living in small short-lived hamlets) and a corresponding de-urbanization.

The upper strata that we are getting to know from the excavations in the central High Mound present us with a very different type of settlement. We have here a major regional sanctuary, the monumental integrity of which is safeguarded in spite of the progressive retrenchment in size. At the beginning of the Mittani period the Temple Terrace stands at its highest, with the full staircase and the revetment wall still showing: it thus retains the full monumentality of the third millennium Terrace, except for the escarpment at its base, which has presumably already been covered by Khabur period accumulations. Even as further accumulations during the Mittani period eventually (presumably by the end of the reign of Sauštatar, around 1400 B.C.) covered the entire face of the wall, the upward slope of the glacis and the flat area in front of it (the ancient Plaza) remained unencumbered by even minor intrusions (pits, tannurs, burials), preserving, as it were, the ideological monumentality of the Terrace.

Some of the excavated structures to the west of the Temple Plaza in A17-A18 (above the Palace of Tupkish) can best be interpreted as service quarters in function of the temple administration. Also, a stone pavement at the edge of eastern end of the structures may be interpreted as marking the western border of the Plaza. We have discussed this in our report on the 2004 season of excavations (Buccellati & Kelly Buccellati 2005).

Finally, a number of items discovered in earlier excavations also point in the direction of a settlement quite distinct from a rural village. We will recall here a dozen bronze and two gold objects found in A17-A18, decorated “Nuzi” style pottery from BH, the head of a small stone statue from A9^{Note 1} and a good Mittani style seal from A9. The typical common style Mittani seal found this year in J1 (see above, section 4.1) can now be added to this not inconsiderable corpus.

The immediate superposition of Mittani above Khabur seems to provide a general argument in support of the low chronology, in that it argues against a vacuum between the two periods. The low chronology dates the fall of Babylon to 1531 B.C. It has been argued (de Martino 2000) that the reason why Muršili I left the statues of Marduk and Sarpanitum in the region of Khana rather than taking it back to Hattuša may be found in the opposition of Hurrian principalities in the Jezirah and the upper Tigris, which would have made his return to the homeland potentially dangerous, and would have made accordingly for a relatively hasty retreat. This is relevant for our work in that it suggests that Urkesh, inasmuch as it was one of the most revered sanctuaries in the Hurrian world, might well have served as a beacon for Hurrian identity at a time when the Hurrians were recovering from the pe-riod (in Old Babylonian times) of subjection to external control (primarily of Mari). Hence the continued importance of the city as a religious center, particularly since at the beginning the Temple Terrace was still standing with the staircase and the revetment wall still in full view.

Table 3 shows a correlation between the stratigraphic sequence and the known kings of Mittani. It must be stressed that we do not have any epigraphic evidence from Urkesh in support of these correlations. However, the conclusions suggested seem inescapable, the main criteria and presuppositions for linkage to stratigraphy being as follows.

1. Urkesh is located almost exactly half way between Wasshukanni (Tell Fekheriyah) and Ta’idu (Tell Hamidiyah). Our excavations have shown that Urkesh had remained an active and respected cultic center, the ancient monumental structure of the Temple Terrace remaining in full use throughout the period that can be identified as Mittanian on the basis of the ceramics. Given this visibility of the sacral structure, and its proximity to the two major political centers of the Mittani kingdom, it seems likely that the fortune of Urkesh was not tied exclusively to the vestiges of its former glory, but that it served effectively as a state sanctioned sanctuary. This would have possibly enhanced the continued sense of Hurrian identity that seems important as a factor of social and political solidarity within Mittani.
2. The immediate superposition of Mittani strata above Khabur (especially in A17 and in J1) suggests that the earliest Mittani levels must date to the very beginning of Mittani history. In other words, since there is no gap in the stratigraphy, we assume that that the beginning of the stratigraphic sequence matches the beginning of the historical sequence. Since Parattarna is the earliest of the Mittani kings about whom we have some detailed information, we link our earliest stratum (s14: full visibility of the revetment wall and staircase) with his name.
3. The loss of the sacral aspect of the Temple Terrace occurs at the end of a long stratigraphic history. Throughout this history, the Terrace retained its integrity. Even as it lost its outward monumentality (on account of the thick accumulations that gradually covered entirely the face of the revetment wall), the Terrace and the Plaza remained privi-leged spaces, without any intrusions such as pits, tannurs or burials. This suggests a strong continuity of the religious identity of the Terrace as sacral space. Consequently, we are attributing the loss of this identity to the disappearance of the Hurrian constituency that made it possible in the first place –- i.e., to the final demise of Mittani (at the end of Tušratta’s reign, s11) as an autonomous political entity at the moment when the Assyrians (with Adad-Nirari I, s6-8) take full control of the area.
4. Given the preceding two assumptions (that the beginning and the end of the stratigraphic sequence span the time period between Parattarna and Tušratta, it is logical to assume further that the intervening strata must be assigned to the intervening rulers. The most notable among these are Sauštatar in the second half of the 15th century (to whom we attribute s13: upper portion of revetment wall and staircase still visible) and Šuttarna II (to whom we attribute s12: Plaza at the same level as the base of the glacis, with only rows of stone marking the boundary between the two).

• Note 1: However different in size and iconography, the style recalls the statues of Idrimi and the one from Tell Brak –- the latter being acknowledged (until the discovery of the Urkesh statuette) as the only example of stone sculpture from the Mittani heartland, see, e.g., Harrak 1997. _{Back to text}