2011: p. 406 fig. 15). Despite the impressive analogies to buildings in Persepolis, the spacious architectural ensemble at Karacamirli rather reminds us of Cyrus' residence in Pasargadae. While in the vicinity, some older settlements of the local population apparently were abandoned in the late sixth or early fifth century BCE, contemporaneously with the palace pit houses were built on virgin soil just 950 m north of Gurban Tepe at a place called Dara Yatax. It is obvious that the Persian conquerors enforced a kind of synoikismos in order to have enough labor force available nearby.
Figure 22.2 Karacamirli: palace on Gurban Tepe.
Figure 22.3 Gurban Tepe: bell‐shaped column bases.
The older indigenous architecture lacks not only the monumental size and complex ground plan but also a number of constructional details of the buildings in Sari Tepe and Karacamirli: regular mudbricks (33 × 33 × 12 cm), recesses, stepped walls, and stone masonry. The quality of execution leaves no doubt that at least some of the stone masons were foreigners. Before the arrival of the Persians there were no large supraregional political institutions, no recognizable states in this region, which would have been able to erect such magnificent buildings. Ground plans as well as column bases and capitals show that the builder‐owners had close relationships to Achaemenid Persia; architects and craftsmen must have been trained in Achaemenid workshops at least.
In eastern Georgia we find similar structures as in western Azerbaijan. The remains of a monumental mudbrick building in the Alasani valley at a site called Gumbati bear resemblance to the palace on Gurban Tepe at Karacamirli situated just 70 km to the south (Furtwängler 1995; Furtwängler and Knauß 1996; Knauß 2000). The bell‐shaped column bases from Gumbati show marked similarities with bases from the site in Azerbaijan. Petrological examinations have proven that both of them come from the same quarry.
In Sabatlo, situated midway between Gumbati and Karacamirli in the Alasani valley, at an important junction chance finds of fine column bases hint at another important Achaemenid center.
Further west, in Zikhia Gora, central Georgia, a fragment of a column base of the same type came to light (Kipiani 1987: pp. 6–11 Pl. 2–5; Zkitišvili 1995: pp. 88–89 figs. 5–6; Gagošidze and Kipiani 2000). More interesting is a bull‐protome capital from this site. It is a provincial adaption of prototypes well known from Persepolis and Susa. The dating of this capital still is a matter of debate (Knauß 1999c: pp. 180–181).
In the late fifth or early fourth century BCE a tower was erected in Samadlo, central Georgia (Gagoshidze 1979, 1996: p. 130 fig. 3). The building recalls Achaemenid models such as the Zendan‐e Sulaiman in Pasargadae and the Kaabah‐e Zardusht in Naqsh‐e Rustam. The archeological context supports a religious function of the tower in Samadlo.
Unlike in Georgia or Azerbaijan there already existed a local tradition of monumental architecture at least in the southern part of Armenia. The Urartian fortress Erebuni is situated on the eastern outskirts of Erevan. According to the archeological evidence, the fortress had not been destroyed at the time of the fall of the Urartian Empire, whereas Karmir Blur (Teišebai URU) on the northwestern border of Erevan, residence of the Urartian governor of Transcaucasia, was razed to the ground and abandoned in the second half of the sixth century BCE. It may be that in the southern part of Armenia a number of former Urartian residences has been reused with minor modifications during Achaemenid rule. This has been claimed for Erebuni and Argishtikhinily (Ter‐Martirossov 2001; Kanetsyan 2001). However, recent investigations have raised serious doubts whether Erebuni was an important administrative center in Achaemenid times (Stronach et al. 2010: p. 126).
The situation was different in the northern part of Armenia beyond the boundaries of the former Urartian Empire. There must have been an important Achaemenid administrative center at Benjamin, about 10 km southwest of Kumairi (Zardarian and Akopian 1994: p. 187 fig. 6; Santrot 1996: pp. 187–189; Ter‐Martirossov 2001: pp. 158–161). The shape of the column bases, worked in local black tufa, is reminiscent of the finds at Sari Tepe, Karacamirli, and Gumbati.
Bell‐shaped column bases are exclusively known from the Achaemenid era and only within the Persian Empire. Outside the major centers in Iran and Babylonia the only findspot so far is the Caucasus (Knauß 1999b: p. 104 fig. 15), nowhere else within or beyond the borders of the vast empire.
The discovery of monumental architecture (Kipiani 1987, 1993), closely related to models from the center of the empire, proves at least temporary Persian presence on the northwestern border of the empire. The distribution of sites with Achaemenid architecture (Figure 22.1) suggests that there existed a kind of network of administrative centers. As the capital of the satrapy (Armina?) must have been somewhere else (at Van Kale?), the resident of the palace in Karacamirli certainly was a Persian magistrate below the satrapal level (Khatchadourian 2016: pp. 146–150). But how can we explain the enormous size and grandeur of the building on Gurban Tepe? Xenophon (Anabasis 4.4.2) reports that the satrap Orontes had a palace (βασίλειον) in a village in Armenia. Maybe the palace served as the satrap's residence when he stayed in Karacamirli during a journey through the land under his control.
The Persian residence in Karacamirli is a clear instance of imitatio regis which Xenophon mentions in his Kyrou paideia (8.6.10–14). The edifice on Gurban Tepe is the largest ancient building in the whole Caucasus. Moreover, genuine Achaemenid architecture on the periphery of the empire is most remarkable, since there was hardly any similar building outside of Persia until recently. To this day we are completely lacking excavated satrapal palaces all over the Achaemenid Empire. Karacamirli may help us imagine their appearance. However, in other parts of the empire where the Persians met with developed cultures (e.g. in Babylonia, Egypt, Anatolia, Syria), the conquerors refrained from implementing their architectural models; instead, local traditions dominated (Knauß 2011: pp. 406–407).
The sheer number of Achaemenid and Achaemenizing artifacts found in Georgia is striking and it reflects a paradigmatic process of acculturation of the Colchians and Iberians. The local elite tried to copy the luxurious lifestyle of the Persian noblemen. Probably, it was not necessary to introduce the habit of the symposion in this region. However, the Colchian aristocracy appreciated the Persian (and Greek) paraphernalia. Furthermore, the iconographic evidence shows a spread of the oriental practice of balancing wine bowls on the fingertips (cf. Xenophon, Kyroupaideia 1.3.8–9). Such an attestation of Persian drinking manner can be found on a recently discovered silver belt as well as on the famous fourth‐century gold finger‐ring with the inscription “Dedatos,” both from Vani (Kacharava and Kvirkvelia 2008: pp. 41–42 fig. 17 [reversed image]; Sens 2009: p. 199 Pl. 53.4; Knauß 2009: p. 299 Pl. 3.6).
Future research in Georgia might concentrate on possible trade routes. In the light of silver phialai from workshops in Lydia especially, the archeological exploration of southern Georgia, e.g. in the Mtkvari River valley (Licheli 2007), and in southwestern Adjara seems to be promising. Of course, investigations in northeastern Anatolia would help significantly to complement our knowledge of the Caucasus in antiquity.
The lack of a similarly impressive archeological record in northern Armenia and Azerbaijan may be explained by a lesser degree of archeological fieldwork. An interesting site is Oglankala in Nahicevan. It is one of the largest late Iron Age fortified
