Back to Dunhuang

Dunhuang Social, Economic and Political History

This material cannot pretend to be a full and coherent treatment of the subject but will instead offer several short essays, which taken together may at least provide useful snapshots. The subjects treated are:

1. Population.

2. Economic life and obligations of the local population.

3. The Sogdian colony at Dunhuang.

1. Population.

As reported by Lionel Giles, we have the following census figures for the Dunhuang district (chün):


2 CE

140 CE

280 CE

609 CE

740 CE













There is reason to believe that in the middle of the ninth century under Tibetan occupation, the population was about that of a century earlier.

Naturally it is a bit difficult to generalize from data spread so widely apart, since events during a relatively short period could cause the population to fluctuate widely. The overall trend seems to be a decline from the middle of the Han to the eve of the Tibetan conquest during the T'ang period. One likely explanation for the decline would be that traffic on the route west going south of the Tarim Basin had declined substantially by the the time of the T'ang. At the same time though, one should not assume that the re-routing of trade of itself would lead to a decline in population. As Giles further notes, with reference to some fragments of a census from the year 416, it appears as though household size in Dunhuang was smaller than in the more densely populated regions of eastern China.

It is likely that the bulk of the Dunhuang population was engaged not in long-distance trade but in agriculture or in the military operations which agricultural pursuits supported. The other main "employer" of the region seems to have been the Buddhist religious establishment, numbering 100-1500 monks during the T'ang period, and relying not only on donations from the faithful but on substantial numbers of "enserfed" laborers. We shall deal separately with the religious establishment and its economy. Census fragments from the year 416 record households a military cantonment of the village of Kao-ch'ang, the cantonment apparently being a fortified settlement (Chao Yü's Rampart). The census indicates the names and ages of all the family members, their relationship to one another, and, for a male head of household, his profession. The breakdown for the households with complete data is:

1. 65-year-old soldier, apparently a widower, with a household of four including his adult sons.
2. 15-year-old "settler", living with his aged mother.
3. 66-year-old soldier with a household of 7 including his wife and grown sons and their families.
4. 56-year-old "settler" with a household of 5 including his wife, two sons and one daughter.
5. 45-year-old soldier, his wife, three sons and one daughter.
6. head local official with his wife, sister-in-law, adult son and daughter-in-law.
7. 26-year-old settler, living with his mother.
8. 24-year-old settler with his wife and young daughter.

A household recorded in the year 747 was headed by a 29-year-old man, and included his paternal grandmother (aged 69), his widowed mother, his wife, and his younger siblings (a brother and two sisters).

There seems to have been a relatively close "clan" ties amongst all the individuals mentioned in the census of 416 CE--that is the circle of families with whom marriage ties were established was small.

The term "settler" probably is a generic one for a farmer but it may owe its origins to a time when in fact many of the local population were new colonists. Stein noted that the documents in the Dunhuang fortifications suggest many of the soldiers were from provinces far away in eastern China; we know that the system of military colonies was established in the Dunhuang region beginning in the Han. Over time, however, the staffing of the garrisons seems to have come mainly from the local population. There is ample evidence that inhabitants of Dunhuang came from various regions though in T'ang times and included Koreans, Uighurs and other colonists from north of the Tarim (e.g., from the area of Karashahr).

2. Economic life and obligations of the local population.

Even though the excavations along the Dunhuang limes have turned up quite a bit of coinage, it seems as though the local economy in the Dunhuang area was largely organized around barter or payment in kind. Fiscal obligations of the farmers were generally specified in terms of deliveries of products some labor duties (corvee). If there was a common "currency," it seems either to have been measures of wheat or bolts of silk. A document regarding the sale of a female slave in the year 991 is revealing with regard to the latter. A "functionary", who lacked the necessary supply of silk to meet his expenses, sells a female slave to a monastery dependent for the price of five bolts of silk, finished and unfinished. An addendum to the document indicates that then an appropriate amount of serge was to be substituted for one of the bolts of silk. Another document from the year 723 regarding a family inheritance dispute, set as a fine for violations a bolt of fine silk for the government and fifteen bushels of wheat for the military.

The local economy was largely based on agriculture, with wheat apparently the dominant crop. Animal husbandry was relatively unimportant, probably be cause of the lack of accessible pasturage, even though the sources mentions some of the standard domestic animals (goats, sheep). Since agriculture (in fact, survival) in the region depended on irrigation, considerable attention was paid to developing and maintaining the system of canals. In the first half of the eighth century, there were seven major irrigation canals in the region. Immediately after the liberation of Dunhuang from Tibetan rule in the middle of the ninth century, one of the first measures of the regional commander was to reconstruct and improve the irrigation system. Irrigation canals sizeable enough to be recorded by name numbered more than 100 in the documents of the eighth through tenth centuries. There were ateliers of specialists assigned to maintenance of the irrigation system.

Some of the cadastral and judicial records from Dunhuang specify the amount and location of land held by families. The absolute quantities are probably less meaningful here than the fact that the land consisted of various separate parcels, scattered over the territory of a settlement. There seems to have been some system of land re-distribution to maintain some equality in the ability of the population to pay its obligations.

While documents concerning loans do not provide an adquate basis for judging the fiscal stability of the population as a whole, they do indicate that many of the farmers had to resort to borrowing grain in order to sow their crops at the beginning of the agricultural season or to eat until the next harvest. Most such documents which have been studied involve borrowing the grain from monasteries. The issue of whether interest was due is unclear, but failure to repay on time could bring harsh and arbitrary confiscation of portions of the debtor's property. There are also documents regarding loans in the form of bolts of cloth, in some cases to individuals who were heading off to other towns such as Turfan presumably for commercial purposes.

Those whose background in history is that of the "early middle ages" in the West cannot but be impressed by the extent to which bureaucracies governing the inhabitants of Dunhuang had elaborate written procedures at a much earlier time. As in any such circumstance, however, the use of and need for legal documents of itself does not provide evidence as to the extent of literacy. While there was some administrative reorganization under Tibetan rule, it seems as though most of the standardized Chinese chancery procedures continued for local matters.

3. The Sogdian Colony in Dunhuang.

The Han sources reporting on Inner Asia make particular note of the trading acumen of Sogdians, the ethnically Iranian merchants from Transoxiana. There is abundant evidence in documents, histories and petroglyphs that their trading activities from their homeland stretched as far as Luoyang (the "eastern" capital of China), north into Mongolia, and down into what today is Pakistan via the Hunza Valley. They were active on both the northern and southern branches of the Silk Road around the Tarim Basin. the chronological range of the evidence about the Soghdians in the East is from around the third century BCE to the eighth century CE. It should be no surprise that there seems to have been a sizeable colony of Sogdians at Dunhuang, dating from at least the early fourth century, when it probably numbered as many as 1000. Even on the eve of the Soghdians' dispersion or absorption into the local population in the mid-eighth century, they numbered some 300 households with an estimated 1400 individuals. It appears that the colony had increased in the early seventh century due to recent arrivals from Trasoxiana, since many of the residents have both Turkic and Sogdian names.

During much of their time in Dunhuang, the Sogdians seem to have been occupied largely in trade-related activities--either as agents in a network of Sogdian merchants or connected with the administration of trade and foreign merchants. The fourth century documents make it clear that they had direct contacts with Samarkand, and there is at least indirect evidence they probably had direct involvement with the trade south to India. Sims-Williams postulates a "triangle" of trade, where one side was the Tarim Basin route, a second the north Pakistan one, and the third from north Pakistan back to Transoxiana. the existence of such connections would help to explain why in the mid-5th century, a Chinese ambassador on his way to Samarkand traveled via the Hunza Valley, where he left an inscription on the Rock of Hunza. The early documents are fragmentary and clearly do not mention all the products of interest for this trade (notably, silk is absent), but among those that are listed are gold, musk, pepper, camphor, cloth of hemp or flax, and wheat.

It seems likely that the long-distance trade gradually gave way to more of a regional commercial focus, in part due to disruptions in the routes and due to political and economic changes. By the eighth century, there is increasing evidence that the Sogdian colony in Dunhuang was becoming sinicized; not long after that, evidence of its existence as a recognizable entity disappears. There are some documents indicating that the Sogdians must have fallen on hard times and became, among other things, dependents of Buddhist monasteries.

Although the building itself has left no traces, documents refer to a sizeable Sogdian temple compound in Dunhuang, where presumably they practiced their Mazdean religion. There is some evidence in the famous cache of documents from Mogao Cave 17 that sizeable numbers of Mazdean "icons" may have been produced for special religious holidays, and that the local government in Dunhuang was involved in their distribution (as it seems to have done for other faiths). Analogies can be seen between the Dunhuang pictures and those in the murals of pre-Islamic Panjikent (south of Samarkand). Such images, in particular involving deities holding the sun and moon, might have been quite acceptable in Dunhuang possibly because analogous ones developed in the Buddhist imagery of centers in the Tarim Basin and spread from there into China. In short, the information about the sogdians and their culture in the Dunhuang region reinforces other evidence we have regarding the cultural syncretism along the Silk Road.

© 1999 Daniel C. Waugh