Abstract
Procopius also mentions that "rich citizens have the habit of attaching to themselves friends numbering twenty or more, who become permanent banqueting companions and have a share in all their property"; to which he adds the custom of collective suicide on the death of the master, which is perhaps to be taken cum grano salis. We have seen here the prefiguration of the Sogdian chākar, a military retinue attached to the lord by a sort of bond of adoption. Is this the origin of the accusations of sodomy that the Persians attributed to the Hephtalites, a stereotype or reality often associated with military fraternities?
The best-known military episode, recounted identically by Procopius and by Arab-Persian historiography, is the Sassanid Azincourt, the final battle against Pērōz in 484: by a feigned flight a Hephtalite detachment lures the Sassanid army into a deep trench covered with branches and earth; the heavy armored cavalry perish entirely, along with the king. False flight and trapping are the practices of nomadic hunters (see the huge gazelle traps all along the edge of the Ustjurt plateau west of the Aral Sea). Procopius mentions the archery skills of the Hephtalite troops temporarily employed by the Sassanids (fig. 1).
Having eliminated the Sassanid threat, the Hepthalites knew that the main danger would now come from the north. The systematization of Sogdian oasis walls into a continuous chain of long walls, the Kampir-Diwal, probably dates from this period, and is currently being studied by Sören Stark in the Bukhara oasis using new physical dating methods.
Khosrow I wanted to repeat the reverse alliance that, early in the reign of Pērōz, had brought the Hephtalites to power, this time with the newly formed empire of the Western Turks. In 556, the Turks conquered Sogdiana. In 560, Tokharestān found itself divided between them and the Sassanids, but friction soon began. As with previous invasions, there are no accounts of sieges or damage to civilian populations in Central Asia, apart from one account in a recently published Buddhist source, the account of the Indian pilgrim Narendrayaśas [11] : coming from Khotan and arriving at Beshbaliq, he was unable to continue his journey to Sogdiana because of the Turkish invasion, which blocked the roads. Compared with previous conquests, the originality of this new conquest from the steppe is that it waited a century to extend south of the Hindukush.
With this last wave, we're bringing the historical narrative to a temporary halt, to take a closer look at the archaeological terrain. In doing so, we need to keep in mind the major frameworks defined by the foregoing: migrations; sedentarization processes; militarization of society; monetary influx.
Archaeological reconstructions: richness and limitations
Sergey Tolstov's theoretical framework placed fortified systems at the heart of archaeological analysis, and more specifically the phenomenon ofencastellamento which, according to him, characterizes the period between the 4th and 7thcenturies [12]. They have remained so ever since, for the simple reason that they constitute the bulk of documentation in this field, both in terms of quantity and accessibility. Sometimes, particularly in Khorezm, we can - or rather, in Tolstov's time, we could - make a lot of observations without having to excavate.
The main synthesis today is G. L. Semënov, Sogdijskaja fortifikatsija V-VIII vekov, St. Petersburg, Gos. Èrmitazh, 1996, unrivalled for the abundance of information and the intelligence of its reflection [13].
The overall inventory of sites is well advanced in all regions (except on the Afghan side), and the most important task today is the refinement of chronologies.
This is a complex task, as was made clear at a symposium in memory of Semënov held specifically on this subject in 2015 [14].
