The Georgian-Canadian Excavations at Ats’quri
in South-Western Georgia
The archaeological expedition described in this short piece was
a project in which a Canadian team from McGill University in
Montréal, where I had been a student before joining the faculty
at Bilkent, integrated to an on-going Georgian research project
conducted by archaeologists from the Centre for Archaeological
Research in Tbilisi. The team went to Georgia as the result of
an International Congress held a few years earlier, to which
(then Soviet) Georgian scholars had been invited. Fieldwork was
conducted between the summers of 1993 and 1997, with one final
study season in 1998. These years were a very difficult period
in Georgia, since the total collapse of its economy soon
followed the political implosion of the Soviet Union itself. The
chaotic state of the country during that time made working
conditions far from ideal, but we were nevertheless able to
complete a modest research programme with the indispensable
assistance of our Georgian colleagues.
The site of Ats’quri where we excavated is located in the area
of Samtskhe-Djavakheti, which borders Turkey to the north of
Ardahan. It is a region of stunning beauty, combining the craggy
mountain landscape of the lesser Caucasus massif, interspersed
with green plateaux and the Mtkvari river, flowing
northwestwards through a deep steep-sided gorge which widens at
places, leaving enough flat or gently sloping land to allow
cultivation and, therefore, settled life. Ats’quri itself, today
a modest village, population about 2000, is built at the head of
one of these small plains (fig. 1), in a position that allows
control of circulation along the river, hence its importance at
all periods of Ancient and Mediaeval history.
Although the Mtkvari river valley is encased on either side by
mountains, relatively easy communications with the outside world
are possible. This is true in particular with the Black Sea
coast to the West, via several relatively low passes between the
highest peaks of the Surami ridge. The Mtkvari valley lays about
100 km from the Euxine coast, as the crow flies. As Classical
archaeologists, our interest in the area centred in the
potential for contact, communications and influence between the
local populations and the Greek settlers of Colchis, along the
Georgian segment of the Black Sea coast. Strabo alludes to such
influence in his description of Iberia, to the north of our
area. The region, somewhat on the margins of Iberia proper, lay
on one of the possible routes linking it with the Black Sea
coast. Our main goal in Georgia was thus to investigate patterns
of acculturation between the local populations and their Greek
neighbours, as reflected by the import and consumption patterns
of Greek materials in the region (pottery and other commodities,
such as wine), as well as the integration of Greek cultural
elements to the local culture (for example local imitation of
pottery types, architectural forms, burial customs, or cult
practices).
The ancient settlement of Ats’quri is located along the south
west shore of the river, at the foot of a series of low hills
constricting the valley to a narrow passage. Ceramic and other
evidence show that the site was inhabited at the very least from
the 6th century BC to the 1st century AD;
chance finds made after our project ended extend the occupation
of the site into the Bronze Age. At some unknown point after the
1st century AD, habitation shifted across the river,
congregating at the foot of a fortress built atop a massive rock
formation. From this point on, the fort, as well as a sequence
of churches (the latest being, in fact, a
large 13th century AD cathedral), became the focus of
the settlement. The former inhabited area was not totally
abandoned, but activity there was minimal, perhaps limited to
agricultural exploitation, as is shown by the few Mediaeval
sherds collected during excavation at the surface of the topmost
cultural stratum. The fortress constitutes the earliest
undisputable sign that the area had integrated formally to an
extensive state: Kartli (Iberia to the Greeks and
Romans).
The ancient site is quite extensive. We cannot measure its
actual extent because, for the most part, its surface lays
buried under a thick accumulation – 2 to 5 meters deep – of
Mediaeval and post-Mediaeval colluvium, which slid from the
slopes of the valley and today blankets not only the settlement,
but the entire valley floor also. Stray finds made along natural
incisions on the river shore and spot soundings, dug by our
Georgian colleagues over the years, demonstrate that the
built-up area stretched over perhaps five to 10 hectares. The
nearly total lack of cultural material in the colluvium
overlying the site, as well as the nature of its matrix (low in
organic content and light in colour), show that the hills were
never extensively inhabited. It is difficult to compare this
site with other settlements in the region to determine whether
it was typical or exceptionally large, unfortunately, because
very few other settlements have been explored at any length. The
surface area of the valley supporting its subsistence, however,
suggests that it must have been among the largest in the
immediate vicinity, a hypothesis backed by the Georgian
historical traditions.
Because of the presence of the late colluvium over the site,
combined with the impossibility of recruiting sufficient local
workmen, excavations could only be conducted on a very modest
scale. On most of the site, it proved impossible to extend
trenches laterally, because this would have meant removing
several meters of very hard clayey sediment to reach the
culture-bearing strata. Such an operation could be done only at
the beginning of each season when the only mechanical excavator
in the village was available. As a result of this awkward
situation, we were able to dig a surface area of approximately
125 m2 in toto over four full seasons of
excavation in the sector under our responsibility (Area V), in
addition to a few small sondages in other areas.
Since there is no place in this format to provide a detailed
account of the finds, I will limit myself to three themes only,
relating them to the general objectives of the expedition: the
house excavated in Area V, the Greek material collected during
the excavation, and the cathedral.
|
Our work centered mainly on the excavation of one house (fig.
2), dated to the 3rd century BC, which shows a number
of interesting features. It is almost square in plan (nearly 7 x
7 meters), with two rooms: a large room with a hearth, a narrow
bench running at the foot of three walls, and a 2 meter wide
platform to the south (the latter not illustrated here), and a
narrow room with a large storage pit in the floor and an oven in
the corner. The house was partially dug-out of a sloping
surface, so that its floor lay between about 1 meter and nearly
2 meters below ground level. The outside walls, made of sizeable
field stones, were very thick – more than 1 meter at their base
– and tapered towards their summit. The inner surface of the
walls, as well as the floor, were covered with a thick layer of
yellowish river clay. The structure of the roof remains unknown,
but the absence of any indications of the use of clay, combined
with the wet climate of the region and the abundant local
tree-cover, all suggest that wood was the main material. These
arrangements, apparently primitive at first sight, were meant to
ensure effective insulation of the house, given that winter
temperatures often dip far below freezing point, in this region.
This structure is the only house completely excavated in the
entire site; other fragmentary structures show, however, that it
is characteristic of the local domestic architecture, although
details may vary
from place to place. The house is at complete odds with the
description made by Strabo of Iberian structures, and shows
that, at this early date, Greek influence had not penetrated
East of Colchis.
A small amount of Greek pottery was collected during excavation;
in all, the number of Greek sherds does not exceed 50. This may
seem an insignificant amount of material, but the collection
from Ats’quri, in fact, corresponds to the largest assemblage of
Greek pottery in all of Iberia. Figure 3 presents a selection of
this material. Most of the Greek ceramics date from between the
5th and the 2nd century BC, and consist
mainly of Attic Black-glazed vessels. The range of shapes
represented is characteristic of the material generally traded
by the Greeks: a variety of drinking vessels, essentially cups
and bowls, with a few sherds belonging perhaps to pouring
vessels as well as transport amphora fragments – unfortunately
no stamped handles were found; in addition a number of small
sherds appear to belong to Hellenistic unguentaria. The interest
of this material rests in the contents of the closed vessels:
wine perhaps, and perfumed oil, as well as the consumption of
wine (which was not necessarily imported, since a long tradition
of viticulture is attested in Georgia). A few additional sherds
were dated to the 6th century BC: a fragment from an
East Greek conical cup, and three sherds perhaps from oenochoes.
These indicate that as soon as the Greeks had settled on the
Euxine coast, they made some form of contact with the
hinterland. The very small accumulation of material indicates
that contact was no more than sporadic. All of this material
originates from domestic refuse; none seem to have been
deposited in graves, which maintain an almost exclusively local
character throughout the history of the settlement. Figure 4
presents the typical contents of a “wealthy” local grave. Greek
material, although sparse, was found throughout the site,
suggesting that all inhabitants had access to it. What were
these products exchanged for remains an unanswered question,
however.
The Canadian team did not excavate in the Cathedral, but one of
our Georgian colleagues did. The interest of the cathedral for
us rests in the literary traditions that surround it: both the
Georgian Mediaeval chronicles and the text of Strabo. The
Georgian sources tell the story of Saint Andrew visiting the
site and converting the, then pagan, inhabitants to Christianity
through a series of miracles wrought by the intercession of the
Virgin and her image carried by the apostle. Furthermore,
following the conversion of the inhabitants, he replaced the
pagan cult with that of the Virgin Mary. Although the
historicity of Andrew’s presence at Ats’quri is most doubtful,
an element of the legend remains of interest. The pagan cult
identified in the Georgian sources is that of Artemis and
Apollo. This might suggest that the local inhabitants had, at
some point during antiquity, adopted Greek cults. The text of
Strabo mentions the presence, in the Moschic mountains (i.e.
today’s Surami ridge) of a cult (and oracle) of Leukothea – the
evil step-mother of Phryxos, of Golden Fleece fame – yet another
Greek legendary character. Ats’quri is one of the candidates for
the site of this sanctuary. The succession of Leukothea, Artemis
and the Virgin on the same site is possible on functional
grounds, since each of the three is associated with a younger
male character (Phryxos, Apollo, and Christ). Such a presence
would be a clear sign of acculturation, but the question is not
as simple as the succession might imply at first sight, for the
identification made by Strabo might very well be based on the
hellenization of a local divinity made by the Greeks for their
own use (such a phenomenon has parallels elsewhere in Strabo’s
description of Iberia), while the identification of Artemis in
the Mediaeval sources could be ascribed to the influence of
Greek culture on early Georgian Christianity. At any rate, the
excavations in the cathedral have failed to demonstrate the
presence of any activity,
let alone a cult, under the cathedral before Mediaeval times.
To sum up, although we set out for Georgia hoping to find
abundant signs of cultural interaction between the Greeks of the
coast and the local inhabitants, and found only very little in
terms of acculturation, we nevertheless can claim a certain
measure of success, in that we are better able now to see the
limits of the phenomenon in Transcaucasian Iberia.
Captions
Figure 1. General view of Ats’quri with Mediaeval castle
dominating the landscape
Figure 2. The house in Area V.
Figure 3. A selection of the Greek sherds found at Ats’quri.
Figure 4. Typical contents of a wealthy grave at Ats’quri.
Newsletter No. 2
- 2003, Pg. 49
< > |
