The urban phenomenon in pre-Islamic Central Asia : diachronic and synchronic approaches (continued) (8)

Listen to audio

The site of Chirik-rabat, south of the lower Syr-darya, seems to offer a nomadic environment (Strabo's Apasiacs?) an echo of these Chorasmian royal or proto-city enclosures. It was in use from the end of the ^4th century to the ^2nd century BC. A 42-hectare rampart, reduced in a second phase, encircles an oblong citadel. The technique used probably involved Chorasmian architects and even masons. Prior to these constructions, the site appears to have been a funerary site, possibly royal, as it incorporates pre-existing kurgans, after which two mausoleums were built, one of which was taken from the final rampart of the site. The latter yielded the remains of cataphract armor identical to that found at the Aï Khanoum arsenal, suggesting that these semi-sedentary "Saka" tribal chieftains could have served as auxiliary troops or individual mercenaries in the heavy contingents of the armies of southern empires: the Parthians, or even the Greeks? In this respect, the nomads of the Chorasmian periphery could have been an important factor in relations with the southern empires, in the same way as trade, of which little is known at this time, apart from a few ceramics and small imported objects.

The "Kushan" period (Minardi: "Antique 3"), ^1st - early ^4th c. b.c.

Toprak-kala (fig. 3) is clearly the site that replaced Akchakhan-kala as the royal site, 20 km to the northeast, according to a methodically executed decision and not under the pressure of external events. Certain elements (woodwork, stone bases) even seem to have been moved from one site to another.

Following the discovery of the site in 1938, the palace within the walls was extensively excavated in 1945-1950, after which more limited excavations continued until 1985.

The city proper occupies 17.5 ha, of which around a north-western quarter is taken up by the inner enclosure, while the north-western quarter is occupied by the palace - roughly the same distribution of space as at Akchkhan-kala, but three times smaller. However, there is an evolution towards the monumental: the city's towers are 15 m high (instead of 8), and the palace itself is perched on a 15 m base. In addition, certain functions were "externalized": the immense enclosed racecourse, the out-of-wall palace, and eventually the city of Darkhas, on the Oxus, 20 km to the west, which disappeared into the river but may have been the "real" economic city, in a relationship reminiscent of that of Old and New Nisa. Digger Jurij Rapoport aptly quotes a text by Appian(Mithridates, 84) on Tigranocerte in Armenia:

"At this place the king founded a city in his own honor. He gathered the principal nobles there, threatening to confiscate anything they didn't bring with them. He surrounded it with walls 50 cubits high (...) On the outskirts he built a palace with vast parks, hunting grounds and lakes. A mighty fortress was built beside it"

The entire site was built in a single phase, in the second half of the ^2nd century AD. The main north-south axis of the out-of-wall palace is aligned with the throne room of the upper palace; the inner walls of the palace platform retained sighting slits, oriented at 333 degrees, Babylonian north(iltanu). Water channels were used for levelling. All the clan signs known in Khorezm at the time are represented on the bricks, indicating that the whole community was mobilized for the construction. However, the site had a shorter existence than its predecessor: it lost its royal function at the beginning of the ^4th century, again by a concerted decision, which we are tempted to see as the effect of the change of dynasty attested at the time. The town survived modestly until the ^6th century.

The racecourse measures 1,250 x 1,000 m, bounded by an earthen levee. It is not an Iranian-style "paradise", since there is no trace of planting or water supply. The excavators' interpretation of the site as a racecourse and army parade ground is obvious. The length corresponds to that estimated by Henning for the čarǝtu avestique, the maximum unit of the race.

The palace outside the walls extends over 9 ha, with structures preserved only on the ground and foundation. It comprises fourteen buildings, each perched on a platform, linked by raised brick causeways. Here, open spaces were used for gardens, and the overall layout is clearly that of an Iranian-style chahâr-bâq, the only probable precedent for which has been identified at Pasargadae. One of the buildings (V) could have been a temple, but its state of preservation makes it impossible to be certain.

The city has only been partially excavated [1]. All that is known of the inner enclosure, outside the palace, is that there was probably a fire temple, identified by its accumulation of ashes (and matching the monumental platform at the center of Akchakhan-kala?). To the south are twelve identical islets. The inhabited islet excavated here is reminiscent of the Dzhanbas-kala casernal dwelling, but here we can see the circulation system more clearly; the rooms are grouped into units whose boundaries could move. A calculation gives 2,000 to 2,500 inhabitants, considered as dependents of the palace. This is the major difference with Vieille Nisa, a royal city of the same dimensions, but uninhabited (because, unlike Toprak-kala, it was in the immediate vicinity of the "real" city?). The artefacts found show the openness of trade in the country at that time, at least in this royal site: glassware, some of which is thought to have been imported from Rome, and Chinese lacquerware.

An islet adjacent to the inner enclosure contains two temples whose function seems to be beyond doubt, although their plans have no parallels. The main one has an entrance iwān followed by three long rooms, each with two supposed altars. It was probably peripteral. Ashes were stored in a well. In the last phase, a cult of the bucranium is attested.

The palace in the north-west corner has been completely excavated and is now the best known and best published of all pre-Islamic Iranian palaces [2]. The interpretation of the architectural layouts and decorations has been the subject of a controversy between myself and Jurij Rapoport, the publication's main author, who recently passed away, and who wanted to attribute a religious or funerary function to almost all the rooms, whereas I reserve this type of function for the "Hall of Kings" alone, which does indeed appear to be a gallery of cult images (perhaps, more precisely, of calendar deities). As the main points of the debate are available in French and English, in online publications, I'll simply refer the interested reader to them [3].

To conclude: at the end of this year's journey, we finally came across a city in the Khorezm, admittedly modest in size and rather artificial in existence, but which concentrated genuine residential functions; we also encountered for the first time a palace that was clearly intended to be inhabited (as evidenced by the harem apartments on the first floor and those upstairs, both decorated with figurative paintings, and the ablutions room near the entrance). There's a change in cultural orientation: whereas Akchakhan-kala betrayed an Arsacid influence, Toprak-kala's palace is a Kushan-style achievement, presumably built or inspired by Kushan architects and decorated by Kushan artists. The city's orthogonal, modular layout, not directly attested in Bactria at the time, could be a local creation.

References

[1] E.E. Nerazik, Ju.A. Rapoport (ed.), Gorodishche Toprak-kala, Moscow, 1981.

[2] Ju.A. Rapoport, E.E. Nerazik (ed.), Toprak-kala. Dvorets, Moscow, 1984.

[3] F. Grenet, "Palais ou palais-temple? Remarques sur la publication du monument de Toprak-kala", Studia Iranica, 15, 1986, pp. 123-135; Iu.A. Rapoport, "The palaces of Topraq-Qal'a", Bulletin of the Asia Institute, 8, 1994 [1995], pp. 161-185.