Материал: Aratashen_mélanges Burney_2004

1

Publié in : Antonio SAGONA (éd.), A View from the Highlands. Archaeological studies in honour of Charles Burney. Ancient Near Eastern Studies, Supplement 12. Peeters Press, Leuven, pp. 399-420.

The Neolithic and Chalcolithic phases in the Ararat plain (Armenia):

The view from Aratashen

Ruben BADALYAN, Pierre LOMBARD, Christine CHATAIGNER, Pavel AVETISYAN

The Armenian part of the plain of Ararat, situated on the great east-west route joining the region of Erzerum in Turkey to that of Tabriz in northwest Iran, is characterised by a mountainous, seemingly restrictive environment. Except for this large natural axis (which follows the course of the Arax to Nakhichevan, then turns south towards Lake Urmia), the traditional itineraries which might insure contact with the other neighbouring regions of eastern Turkey, Transcaucasia or the northern Near East are much more rare. They should not, however, be dismissed (map).

The Anatolian zone of Lake Van remains accessible, at least in summer, by a route which must cross a pass at an altitude of 2500 m. To the north, the middle valley of the Kura can be reached by skirting the Aragats range, either in the west by following the Akhurian river and then crossing the high plateaus of northern Armenia to southern Georgia, or in the east by the Razdan valley, then that of the Agstev which flows to northern Azerbaijan. To the east of the plain of Ararat, the mountain ranges of Gegham, Vardenis and Sjunik present obstacles to communication with the more southern steppes of Azerbaijan. However, today as well as in the past, the populations of these steppes and those of the plain of Ararat summered their herds in these very mountains, thus favouring contact and exchange.

This potential for contact with regions which have been occupied since the Neolithic or the early Chalcolithic (Shulaveri-Shomutepe culture of the Kura basin and the Karabakh plain, Kül Tepe of Nakhichevan culture in the lower basin of the Arax), even the Palaeolithic (western Georgia), is in total contrast to the lack of information observed in the plain of Ararat1.

For the periods before the Early Bronze (Kura-Araxes culture or Early Transcaucasian Culture), information is very fragmentary. A single site, Tekhut, was excavated at the beginning of the 1970’s and published. This site produced circular mud brick habitations, as well as a lithic industry and pottery attributed to the late phase of the

Ministry of Culture and A. Kalantaryan, director of the Institute of Archaeology at Yerevan, for their administrative support. We are most grateful to all the researchers who took part in the excavations and in the study of the material : G. Artine, A. Hayrapetyan, S. Melkonyan, I. Gagnon (excavators), J. Chabot (lithic specialist), H. Sarkissyan (architect), J. Chekyan (topographer), L. Manoukyan (restorer), A. Karakhkhanyan and V. Trifonov (geologists). We thank also G. Poplevko (St Petersburg, Russia) and B. Lyonnet (CNRS) for information about the data coming from Aslanian’s sounding and kept in St. Petersburg Museum.

2

Chalcolithic (Torosian 1976). At another site, Khatunakh, an exploratory trench revealed an earlier phase, considered to be Neolithic because of the scarceness of pottery (Torosian et alii, 1970, p. 386). A few other deposits were surveyed (Ada Blur, Masisi Blur, etc…), but this late Chalcolithic material, kept at the regional museum of Vagharshapat (former Echmiadzine), remains almost unpublished2.

For several years, the Franco-Armenian archaeological mission created in 1977 has endeavored to fill this lack in the knowledge of the late prehistory of Armenia. The mission has paid particular attention to the collection of chronological, stratigraphic, and material data pertaining to the crucial phases preceding the flowering of the Kura-Araxes culture, the expansion of which in the third millennium had a strong impact on Transcaucasia and the whole of the northern Near East. The site of Aratashen, briefly explored in the 1970S by Sardaryan, then in 1993 by Aslanyan (note), was again chosen in 1999 for a large program of excavations over several seasons.

The site of Aratashen

Situated some 40 km west of Yerevan and 1.5 km southwest of Vagharshapats, the settlement of Aratashen lies in a loop of the Kasakh River, which flows into the Araxes a few km to the south (Figure 1).

The central part of the site consists of a little elliptical elevation (blur) of about 90 x 60 m which dominates the surrounding plain at a maximum of 3 meters, at an absolute altitude of 852 meters. A thick alluvial layer of pebbles, vestige of an ancient and extensive spate of the Kasakh, covers the immediate vicinity of the blur. It probably hides the exact borders of the site, which possibly extended more to the north and the west originally. During its seasonal inundations, the river eats regularly into the embankment of the site to the north, where layers containing obsidian and pottery are exposed in section over about ten meters. During the 1950’s substantial leveling work was carried out in the zone of Artashen, which appears to have greatly modified the original profile of the blur. Intended to counter the winter flooding of the Kasakh, two segments of an elongated earth dike are still visible to the east of the central elevation; it is now known that this dike was constructed of earth removed by bulldozer from the main elevation and its immediate surroundings. It is possible that this recent work could be superimposed upon an older protective construction, perhaps contemporary to the prehistoric site.

An obvious consequence of this partial scraping is profound alteration of the stratigraphy of the site and the acceleration of erosion. The first exploratory trenches in 1999, then the two excavation seasons carried out in 2000 and 2001, demonstrated the immediate proximity (-0.20 to -0.30 m beneath the present surface) of a series of three distinct horizons (Levels I to III) which represent the earliest occupation of the site, preserved here in a depth of a little more than 2.5 m. The lower Level III rests on virgin soil. No earlier deposit has been found, while the intrusive pottery of late phases (Iron Age, classical and medieval periods) can be found near the surface.

3

To better determine the obviously complex geo-archaeological development at Aratashen, a strategy of excavation in intermittant longitudinal trenches (A to K) was adopted and applied to a north-south median axis covering the whole of the blur over almost 80 meters. This strategy proved effective, as the stratigraphic sections in line provided a profile of the whole of the elevation.

An obvious dichotomy appeared very rapidly between the situation observed in the central exploratory trenches (A, B, D, E, F and H) and that revealed by the peripheral trenches (mainly C, J and K). The three successive archeological horizons which were clear in the former were nowhere evident in the latter, which presented a series of overturned destruction levels several of which contained abundant pottery material of Chalcolithic type, which is practically absent from the levels excavated on the highest part of the excavation.

A thorough study of the stratigraphic sections, combined with a sedimentological analysis of the exposed layers, explains this unusual situation as well as reconstructing the geomorphological sequence of Aratashen as a whole (Figure 2).3

Geologists and archaeologists agree today that the occupation at Aratashen was without a doubt still active after the period represented by levels I-III. As we suspected at the beginning of our excavations, these later levels have completely disappeared today, partly victims of recent leveling work for agricultural purposes, but mainly because of long natural erosion of the blur; the products of this destruction have accumulated at the foot of the elevation in a thick layer of nearly 1m where the abundant Chalcolithic pottery recovered in 1999 abd 2000 in trial trenches C and J was found. This displaced and accumulated evidence of an occupation level which no longer exists today thus explains the scarcity of pottery of the same period at the top of the blur (presence limited essentially to Level I, and only in low quantity). This reconstruction influences the dating of levels I to III, older than this ruined phase, dated by its pottery to a late phase of the Chalcolithic.

The thick alluvial layers which cover the low areas of the site are thus due to a sudden rise of the waters of the Kasakh (from 2 to 2.5 m), after a probable tectonic accident (a sudden fault, in a seismically unstable region). These alluvions then covered the destruction/accumulation level, but not the upper part of the blur, which remained above the water. Somewhat later (probably very quickly, the river having rapidly modified its course by circumventing the obstacle), the level returned to its initial position, causing the formation of Terrace I of the Kasakh. It is not possible at the moment to date this sudden inundation of the Kasakh.

The three archaeological horizons identified at Aratashen are identified briefly.

Level I

This level is cleared over more than 170 m2. Its proximity to the surface (average depth of structures: -0.70 to -0.30 cm) renders its preservation extremely unequal: the surface vegetation, composed of little thorny bushes, produces here a deep, strong root system which once it breaks the coherance of the the bricks in place, causes rapid disintegration of the construction, already weakened by humidity and winter frost. A strategy of extensive

4

excavations has however enabled discovery of the vestiges of four circular structures (and probably those of a fifth, to the east). In the south of the sector excavated at present, there are also the vestiges of three little orthogonal walls, separating two constructions of circular plan. With the exception of one of these, which is smaller, these round constructions present the same reconstituted diameter of about 5 m. All are built in flat, rectangular mud bricks of even size (about 45 x 25 x 8 cm). Remains of clay floors are evident is each of them, as well as in the orthogonal structure.

As Figure 3 indicates, two long, straight structures with coating cross the excavated zone from east to west. They are preserved respectively over 9.50 and more than 10 m in length, with an average width of 1.8 to 2 m; they have a slightly concave profile, in a “U” form, and appear to be edged here and there with a kind of border; they are carefully coated along all their length. In the waiting of a complementary excavation, we reserve our interpretation for the moment. We can of course see two areas intended for circulation in the middle of the constructed space, but this remains a simple and unsatisfactory hypothesis. The stratigraphic relation between the northern structure and the circular construction situated at its western extremity is not clear; it is possible that future work will demonstrate that these structures, with their slightly concave profile, could constitute the lower part of a construction dug from a higher level which no longer exists today; the nature of this possible entity remains unknown. A zone of hearths, dug into the ground and stone-lined, as well as two mud brick silos (?) have also been discovered outside the habitations.

The collected archaeological material consists of a very well represented obsidian industry, a bone industry, ceramic material which is less evident, and grinding tools. Except for this last category (grinding stones, mortars, etc.), more associated with built structures, the archaeological material appears to be dispersed over the whole of the excavated zone; this appears to confirm the existence of zones of technical activity in the spaces separating the constructions.

Level II

This level was cleared over a surface of only 45 m2. This limited area does not yet permit a vision of the whole of this “horizon” which, unlike the two others, has so far produced hardly any architectural structures.

In the excavated sector, several isolated structures consisting of rounded, formless basaltic pebbles are found associated with large amounts of collapsed mud brick, apparently mixing pisé (?) and brick fragments, some of which are charred; these bricks are thicker than those of Level I, made from a more plastic clay, which in general have preserved impressions of its plentiful plant filler (stems and grains of cereals). These concentrations of pebbles do not generally present a regular plan, but are clearly anthropic in nature. Two little isolated round structures in mud brick (Figure 5) were also discovered for the first time this season, notably in Sondage E; as there are no traces of ash, they would not appear to be hearths.

These various features do not appear for the moment to form a truly coherent whole, nor do they seem to rest on a prepared floor or a simple exposed surface. However, clearly evident are the remains of earth constructions of which the foundations would have been built or reinforced with pebbles. We should also consider the possible association of structures built with pebbles and light constructions built with perishable materials. However, the particularly meticulous excavation of these levels has not revealed a single posthole, adding to our questions.

One statement based on the evidence is certain: this “horizon” is so far the richest in archaeological material (more than 55% of the inventory); there are large concentrations of artefacts in obsidian (mainly blades) and especially bone. Pottery is only represented by a few sherds, but also by a little ridged vase which is complete (Figure 4 :1). This level also

5

contains very rich faunal material, including many antlers; these sometimes present transverse perforations (Figure 4 :2), and are generally associated with the structures described above.

Level III

The earliest archaeological horizon at Aratashen has so far been found only in a narrow trial trench of 10 x 2 m. It is represented by a rather dense architectural network, made up of structures with rounded, usually irregular forms, often adjacent or attached by little curved walls (Figure 5). The main characteristic of this mud architecture is the use of pisé and not of brick, which is present in the upper levels; one also notes the presence of several ashy zones, of which at least one was associated with a circular hearth constructed of large basaltic pebbles (diameter: 15 to 18 cm); a little round structure with low elevation (0.13 m) and a diameter of about 0.70 m stands out in the western part of the trench. This horizon is dense with architecture and rich in objects (mainly obsidian and bone), and rests on a series of sandy virgin layers, partly reached by the level of the water table.

The objects discovered are apparently identical to that of Level II, but include no pottery.

Interpretation and discussion

With these preliminary results, we can propose a preliminary reconstruction of the occupation sequence at Aratashen. The village seems to have been founded by populations, which had no knowledge of pottery (Level III), and then progressively adopted (Levels II-I) the use of pottery which appears already developed. Moreover, the remains of one or several upper levels (Horizon “0”), which are completely eroded, have accumulated on the periphery of the site and present abundant pottery which is evidence of a cultural phase which is later but clearly precedes the Kura-Araxes culture.

With no radiocarbon dates for the moment, it is not possible to date this sequence precisely as dating still depends only on comparative data from the neighbouring regions of Transcaucasia and the northern Near East.

These analogies (or differences) will be determined according to the architectural data and the objects.

The pottery from Horizon “0” as terminus ante quem

The pottery from the eroded level or levels of the tell (sondages C and J near the river Kasakh) consists of two ensembles, both very far from Kura-Araxes productions: (a) an abundant production with plant filler mainly including hemispherical bowls with a scraped surface, which evoke the "Coba bowls" which became widespread in northern Mesopotamia in the first phase of the late Chalcolithic.4 (b) a production with various mineral tempers (sand, crushed obsidian, baked clay fragments), often containing chopped straw, made up of hemispherical bowls and jars with flared rims or little cylindrical necks; the decoration of the bowls consists of incisions on the lip, rounded protuberances on the edge of the lip or on the sides, and one or two rows of perforations beneath the lip; this pottery is close to that of the Sioni culture which developed in the Kura basin after the disappearance of the ShulaveriShomutepe culture.5

4Trufelli, 1997, p. 8-10.

5Dzhaparidze, 1989, p. 338-339 and fig. 132