The urban phenomenon in pre-Islamic Central Asia : diachronic and synchronic approaches (continued) (4)

Dzhanbas-kala, a site probably roughly contemporary with the previous one and comparable in general layout, but much smaller (3, 4 ha), lies on the north-eastern border, in an area reclaimed by the desert in later periods. Explored as early as 1939, it has played an important role in the history of Khorezm archaeology, as the mission made its mark both in terms of method (surface surveys with very limited excavation) and interpretation. It features a hollow, tiered rampart with multiple sets of triple archways spaced 1.30 m apart, a barbican but no towers. The major difference with the sites previously examined is that it is almost entirely built indoors, with two compact housing blocks, each with around 200 rooms and a few courtyards, on either side of an irregularly wide axial "street". Tolstov states categorically that there are no perpendicular alleys, so the analogy with a Hippodamian plan is superficial, even though the occupation corresponds to the Hellenistic period. It makes individual rooms the only dwelling unit. In light of the excavation of Kampyr-tepe (see last year's lecture), we can now question these restitutions.

This site has generated a number of patterns [1], later applied to other sites:

1) The matriarchal survivals of the "primitive community": assuming one adult man per cellular dwelling, i.e. 400 men, they could only have lined (and still incompletely) the arches of a single wall; this implies that women took part in the defense (cf. the ancient notices on the "Amazons" of the Massagetes, a people to whom Strabo links the Chorasmians). Today, we would tend to think that the many archers had a dissuasive effect.

2) Bipartite urban societies, another survival of the "primitive community" in ^19th-century anthropological literature, an idea taken up by Engels. Tolstov sees confirmation in the fact that the brigade marks on the bricks form distinct groups from one rampart to the next, indicating that each section of the rampart was built by the inhabitants of the adjoining massif. According to ^10th-century Arab geographers, survivals of these dual societies can be found in the medieval towns of Central Asia, under the guise of affiliation to various sectarian or Sufi movements. Considered without preconceptions, the layout of Dzhanbas-kala is more reminiscent of a barracks, which is perfectly in keeping with its frontier location.

3) The social role of fire worship: this was the first "fire temple" to be recognized in Khorezm, followed by a host of others. Cautiously, Tolstov speaks of a "house of Fire", using Biruni's expression beyt al-nirān, which emphasizes the communal rather than strictly religious aspect (he cites Biruni in connection with the Sogdian festival of Mithras, where people feast in the beyt al-nirān). The building, located at the end of the street to the south, comprises a corridor and three rooms, one of which (112^m2 ) is littered with crockery and animal bones, obviously used for banquets. In the next room, a mud-brick platform with a bench was built in the middle, a feature that was present at all three stages. Given its poor state of preservation, there are no decisive clues to help us choose between two interpretations:

• a fire temple where the permanent fire would have burned in a metal vase, as in modern Zoroastrian temples; the large room would have been the place for religious assemblies, according to the usage described by Biruni and again in the 1960s by Mary Boyce (admittedly in private homes and not in the temple);
• a building for secular use, probably the residence of the local commander; the remains observed could be explained by the practice of sandali (or korsi) once found in Iranian "tea houses", with guests sleeping on benches and sharing a blanket diffusing the residual heat of the central fireplace.

The same terms of discussion will be found in many later monuments in Khorezm and elsewhere in Central Asia.

[1] Described in great detail in Drevnij Khorezm, pp. 88-100.