The main activity of the season was the renewed excavation at the Natsargora mound. The site had been excavated in the 1980s by the late Alexander Ramishvili, who had mainly investigated its Late Bronze levels. A number of soundings had however revealed the presence of EBA levels, and unearthed important materials of both the Kura-Araxes and the Bedeni periods, according to the excavator in association with each other. These had raised the question of the possible coexistence, at the site, of both cultures, which would have had important consequences on the general EBA chronology of the region. After analyzing, during the 2009 and 2010 field seasons, the finds from the old excavations and the relevant documentation, we realised that the site’s stratigraphy was so complex and confused that no firm conclusions about this question could be reached without undertaking new excavation there.
    With this aim, we opened a ca 200 square meters excavation area, corresponding to eight 5 x 5 m quadrants, on the present top of the mound, S and E of Ramishvili’s EBA soundings (Fig.1 e Fig.2). Contrary to our expectations, it turned out that, after completing his excavations, the Georgian archaeologist had created an artificial horizontal surface under the level reached by his excavation, which in most squares approximately corresponded to the base of the Late Bronze occupation (Fig.3). Since, however, the archaeological levels were heavily sloping in E direction, this had caused the disappearance, in the W part of the mound, of the upper EBA levels as well. In addition, the whole investigated area was heavily disturbed by a huge number of Late Bronze Age pits, partially excavated by Ramishvili, which had cut into the earlier levels, as well as by numerous animal holes and plant roots. For all these reasons, the EBA levels were rather poorly preserved.
    In the SE part of our excavation, in particular (quadrants 099.099b), less than half a meter of anthropic deposits had been left over the natural soil, which was met in a small sounding (Sounding no. 2) at a maximum elevation of ca 775.90 a.s.l. The remaining stratification consisted of a series of at least six thin superimposed layers marked by the presence of slightly sloping floors of yellowish or whitish colour. Three of them have been excavated so far: the upper ones appeared to have been external surfaces, since only scanty remains of installations (fireplaces, ash-filled areas) were found lying on them. The third layer, which was less disturbed by the later pits, on the contrary yielded some remains of more substantial constructions. These belonged to two main types: subcircular structures of small dimensions (Loci 0376, 0364, 0309) – one of them measured ca 200 x 150 cm –, built in mud-bricks (?) or blocks of reddish yellowish clay, one of which contained an in situ fireplace and an in situ grinding stone, and smaller rectangular structures with walls of dark brown mud, which appeared to host peculiar depressed combustion features (Loci 0389, 0371, 0374 etc.) (Fig.4). None of the walls was preserved for a height of more than 7 cm; as can be seen from the section of some pits, the same is valid for the structures of the underlying layers as well. It seems, therefore, that architectural structures in the area were rather ephemeral in nature, and had often been re-built within a short lapse of time.
    The ceramic inventory recovered in this area of the excavation was purely Kura-Araxes in date; in particular, no Bedeni sherds were found in good contexts there. This confirms our hypothesis that at least the lowest EBA levels at the site definitely belong to the pre-Bedeni period. An interesting example of complete Bedeni vessel was found, however, in a later pit cutting one of the Kura-Araxes structures (Fig. 5).
     Part of a larger construction (Locus 0431), presumably a typical Kura-Araxes hut of quadrangular shape with rounded corners, was discovered in quadrants 100.099b-d to the N (Fig.6). It had two phases of use, during the second of which a rounded installation was built approximately in its centre, over a thick layer of ashes. The area to the N of this building had unfortunately been cut into by the stone foundations of some Late Bronze constructions (Locus 0406), which had partially been removed by Ramishvili.
    In the E part of the excavated area (quadrants 100.100a, b, 099.100a,b), the situation was quite different. Due to the natural slope of the mound, archaeological levels were much deeper here, as proved by a small test trench (Sounding 1), in which virgin soil was reached at alt. 774.30 a.s.l., i.e. more than one and half meter deeper than in the W part of the excavation. It is possible that the natural slope had been terraced in ancient times, as shown by the presence of one or two possible low steps visible in the EW sections.
    In quadrant 100.100c, a series of superimposed Late Bronze surfaces were preserved just under the present top soil. These were underlain by a more than 50 cm thick filling, which did not contain any significant feature. The following layer was characterised by a small stone ramp (Locus 0234) associated with a round fireplace, which appear to be still LB in date. Excavation stopped at the top of the underlying layer, which yielded mixed LB and EBA pottery, and might therefore represent the transition to the 3rd millennium BC occupation in this area.
     On the contrary, no LB occupational layer was preserved in quadrant 099.100a to the S. After emptying a large number of LB pits, we reached, under the disturbed sub-surface soil, the top of the EBA (apparently Kura-Araxes) occupation, which yielded decomposed remains of pisè or mud-brick walls, and a very fragmentary fireplace.     In the easternmost part of the area (quadrants 100.100d, 099.100b), excavation did not proceed beyond emptying the LB pits, with the exception of the already mentioned small test trench (Sounding 2) which was continued down to the natural soil (Fig.7). The sounding showed the presence, in the area, of a ca 60 cm thick LB occupational level, underlain by a ca 40 cm thick EBA, presumably Kura-Araxes, level. The latter appears have been characterised by a thickly packed sequence of undisturbed surfaces of yellowish and whitish colour directly overlying the natural soil, which we will hopefully investigate in the course of next year’s campaign. From the uppermost of these surfaces, a pit was dug, in which a circular stone feature of unclear function, covered by stones, (Locus 0176) was built (Fig.8).To sum up, the evidence collected so far indicates the presence at the site of a substantial Kura-Araxes occupation, though apparently characterised by rather ephemeral architectural structures, overlying the virgin soil all over the excavated area. The presence of Bedeni layers, on the contrary, remains rather elusive. The association of Kura-Araxes and Bedeni materials in the very same contexts encountered in Ramishvili’s investigations was not confirmed by our own excavations. It must be noticed, in this respect, that only a handful of Bedeni sherds was recovered this year in the whole excavated area. This suggests that the abundant Bedeni material collected by Ramishvili mainly derived from pits. These were probably not deep enough to exceed the bottom level of the Georgian excavation, and had therefore been already completely removed before the beginning of our own investigations.