Le fait urbain en Asie centrale préislamique : approche diachronique, approche synchronique, III : la crise urbaine et la réurbanisation (IIIe-VIe s.), un processus général ? (continued) (5)

Abstract

Nothing of the kind has been transmitted for Central Asia. On the other hand, Sogdiana provides data for these periods testifying to the rapid assimilation of successive waves of newcomers, particularly among the merchant class then expanding on the trade routes. In Sogdian inscriptions from the upper Indus, dating from around the^{5th century}, "Khūn" accounts for a third of the names, including sixteen whose fathers bore Sogdian names. In the early 8th^century, the ruler of Bukhara was still called Khūnak. The painting in the cella of the Dzhartepe temple (^{5th century}, pre-Hephtalite?) probably shows the new mixed aristocracy hunting in the "paradise" of the rulers of Samarkand. Linguistically, for the pre-Hephtalite period we have a few rare onomastic clues indicating an Altaic language: Khunkhas = "khan of the Huns", "Huns Oghlar" *Oghullar, "princely Huns" or "Huns Oghur" (the Oghur would later become the other major division of the Turks along with the Oghuz).

For the Hephtalite period, textual data become even more numerous and diverse, but they leave us with very contrasting impressions. There are two ways of looking at these contrasts: a hybrid politico-social structure or a chronological evolution. A bit of both, no doubt.

In Tokharestān, the ruling stratum of the Hephtalites, the milieu that the Chinese envoys encountered on the eastern foothills, long retained a nomadic lifestyle, including funeral practices, fully confirmed in this region by archaeology. Chinese chronicles admit that information on their origins is contradictory; that which dates back to the first contacts, the Weishu, has them coming like their predecessors from the great migration of 350-370, but unlike them the onomastics of the sovereigns is not Altaic. Once established, they remained behind the scenes for a century, which undoubtedly enabled an initial acculturation. The polyandry described by the Chinese has led to a false Himalayan/Pamirian trail. It is in fact attested by a marriage contract of Rōb as early as 332, before any nomadic invasion: it is a local custom. The last Chinese account (in the Zhoushu) for the "imperial" period of the Hephtalites mentions a large fortified capital, Badiyan(bǝit-tɛj-jian), where there are "many temples (non-Buddhist?) and pagodas"; although it has been doubted this is undoubtedly a suffixed form of the name Balkh, which was then pronounced Vakht.

In Sogdiana, they appear not as nomads but as emissaries of an already stabilized state, which set up garrisons in the towns and probably also dynastic branches. As early as 484, the Hephtalite king confronted at Pērōz is called Akhshūndhār, a Sogdian royal title.

It is unclear to which of these two parts of the empire the testimony of the Byzantine historian Procopius (Wars, I.iii.3-6) applies, which clearly tends towards the sedentary version:

[...] they are not nomadic like the other Hun peoples, but have long since settled on good land [...]; since they possess a standardized constitution they observe law and justice in their affairs, both among themselves and with their neighbors, to no less a degree than the Romans and Persians.

Does this mean tribal law (as the Mongolian yasaq would later become), or the adoption of the Iranian law of the subjugated countries? His testimony is corroborated by his continuator Menander: in 567, just after the Turks had defeated the Hephtalites, the Sogdian Maniakh, the Qaghan's envoy to Byzantium, when asked: "Do the Hephtalites live in cities or in villages?" replied: "In cities".

The appropriation of a state culture is marked by their fearsomely efficient taxation (Rōb, contract J, in 517: "the Hephtalite lords' tax on our house was heavy, and I have no domestic assets on which we can pay it", which recalls Songyun's testimony for the Gandhāra Hephtalites at the same date). This must be taken into account when explaining the size of their armies. For military strength, this is the only time we have figures, which do not seem implausible(Beishi : "about 100,000 troops" - the Chinese character bu can mean both "tribe" and "troops", but it's probably the second meaning) [10]. Their military victories over the Persians brought them control of the eastern provinces (Merv, Herat, Sistān) and enormous resources in precious metal: between 476 (first defeat of Pērōz) and 531 (interruption of tribute by Khosrow ^I), they would have, if the figures transmitted for tribute are reliable, received most of the money supply issued by the Sassanid state. Indeed, numismatists note massive issues of Pērōz from 476 onwards, and at Chaganiān (Northern Tokharestān) all silver coinage in circulation for decades consisted of countermarked Pērōz coins. Consequently, the Hephtalites had no incentive to issue original-type coins en masse in Tokharestān; they did so only in India.

References

[10] Bear in mind, however, a correction made by É. de la Vaissière, "Early medieval Central Asian population estimates", JESHO, vol. 60, 2017, p. 788-817, here p. 813: the round figures transmitted by foreign sources, up to the Mongol period, correspond to tümen, theoretical mobilization units of 10,000 men, whose effective fill rate may not have exceeded 50% or even less.