Skip to main content

Early China General Mailbox:
Email: earlychina@Dartmouth.EDU
Comments: Contact Us
Please submit manuscripts to this address:
Sarah Allan, Editor of Early China
HB 6191
Dartmouth College
Hanover, NH 03755

Please send books for review to this address:
Anne Behnke Kinney, Book Review editor for Early China
Department of East Asian Languages, Literatures and Cultures P.O. Box
400781 Charlottesville, VA 22904-4781

Please send correspondence concerning monographs to this address:
Andrew Meyer, Monograph editor
Please submit items for the website to this address:
Charles Sanft, webmaster

Research Notes

The Administrative Documents from Yinwan: A Summary of Certain Issues Raised

By Michael Loewe 魯惟一


As the large consignments of strips found at Dunhuang 敦煌, Juyan 居延 and other sites in Gansu came from the garrison stations built as part of the lines of defense, they are consequently largely concerned with military administration. Of the many and highly varied documents found in the sites of central China in recent years, very few have derived from the administrative procedures of local government that were practiced at the level of commandery (jun ) and county (xian ). Tomb no. 10 Fenghuang shan 鳳凰山, Jiangling CC, Hubei included records such as the names of members of a household, the extent of the arable land or the figures of tax that had been received. [1]  No more than a preliminary notice has so far been available for the 170 strips found at the tomb of Wu Yang 吳陽, Huxi shan 虎系山 Yuanling 元陵 (Hunan), which are described as a Huang bu 黃簿 and are said to give details of the communications and distances between the nobility of Yuanling and both Chang'an and the capital of Changsha kingdom; they are also said to carry details of matters such as the numbers of inhabitants, taxation, economic resources and exemptions granted to the elderly and the sick. [2]  Of considerably greater importance are the documents of this type that were found at Yinwan 尹灣, Donghai xian 東海縣, Lianyungang 連雲港, Jiangsu. These are far more extensive and varied than those found elsewhere; they include information of a specialized nature that is not otherwise known; and full photographs and transcriptions have been made available since 1997. [3] 

Other documents of a completely different type that came from Yinwan have attracted considerable attention. This article will concentrate on those that concerned provincial and local government and which happily can almost certainly all be dated to the comparatively short time span of 16 to 11 BC, during the reign of Han Chengdi. They may thus be taken to refer to conditions of government that remained largely uniform in those years and they may be compared with certain accounts which are retained in the Han Shu of very much the same period (AD 1 - 2), during Pingdi's reign. Studies of many of the points of detail that the strips mention may be found elsewhere. [4]  The following pages are addressed to possible conclusions of a general nature that apply to the administration of the Former Han empire.

During the reign of Chengdi the area where the strips were found was part of commandery of Donghai 東海, which was governed as one of the regular eighty-three commanderies of the empire. This had previously been an integral part of the Qin empire, to be named Tanjun CC in the restored kingdom of Chu (210 BC); this region remained within the succeeding kingdoms of Chu that were ruled by Xiang Yu 項羽  from 206, and by Liu Jiao 劉交, as part of the Han empire, after 201. In 155 Yinwan became part of the re-founded commandery of Donghai. By the time when the administrative documents were being drawn up this had enjoyed a long period of peaceful government. Involved slightly in civil warfare over the years, it had not been subject to unrest.  One of the largest commanderies of Former Han times, it comprised a total number of thirty-eight sub-units, of county (xian ), nobility (hou ), and estate (yi ); the average number of sub-units of a commandery was fifteen.

The documents bring out the importance of the officials who were entitled Gong cao 功曹, or occasionally Gong cao shi 功曹吏. Low grade officials with the rank of 100 shi C, the Gong cao were posted at different levels of both central provincial government. When, in the commanderies, they were attached to the Governor (Tai shou 太守), they could hold responsibility for handling the reports and recommendations on which an official's chances of promotion depended. Sometimes they acted as the personal assistant of the Governor. One of the documents from Yinwan is a record of the journeys that the Gong cao  undertook as part of his official duties in the year Yuanyan 元延2 (11 BC); in this particular case its purposes possibly included the duty of conveying the news of the death of the Governor to officials of neighboring commanderies. [5]  The appearance of this official as the subject of several wall paintings may indicate the importance that was attached to his office. [6]

Some of the documents, which originated form the office of the Governor, give lists of the subordinate units of the commandery. These may be compared with precisely the same type of list which is included in chapter 18 ('Di li zhi' 地理志) of the Han shu for the year AD 1-2. The degree of correspondence is very high, thus corroborating the value of the Han shu as a reliable record, but the order of the units differs as between the two lists. One of the documents [7] from Yinwan give the counties, nobilities and estates in descending order of their size, as measured by the number of the registered inhabitants. Other details show how important this distinction was, as the grade of the senior official of these units varied according to their size. Thus, the senior officials of the counties (magistrate) were termed ling for the larger xian with a population of 10,000 households or more, and zhang for those of less than 10,000. As is known from the Han shu, the ling were entitled to a higher salary than the zhang, and we now see how the numbers of their supporting staff varied correspondingly; the ling had a staff of 60-107 subordinates, the zhang a staff of 86-27; in very small counties this figure was no higher than 22-66.

In principle officials were not permitted to serve in areas of which they were natives and the documents show how this principle was implemented in the case of Donghai jun. They name 119 subordinate officials whose place of origin had been in a number of commanderies, many of which were close to Donghai; exceptionally, one came from Donghai itself. That many of the officials came from the adjoining commanderies may perhaps have been due to practical reasons, in so far as they would have had little or no difficulty in understanding the dialect spoken by the men and women of Donghai itself.  Had the officials posted there been men whose native area had been further a field, e.g. from Shu or the deep south, they might have encountered difficulties in making their demands fully understood. Possibly there were other considerations of a local nature, as there is a considerable variation in the distribution of officials among the different commanderies; no less than eighteen had been natives of Shanyang jun 山陽郡, while the average number of those from other commanderies was four or five.

The register of officials serving on the staff of Donghai commandery suggests that it is necessaryt to distinguish between three categories of those of junior rank, although these differences are not so obvious in the historical texts. [8]  There were:(a) Officials posted in the commandery as part of the regular complement.(b) Officials serving on the staff of a senior official, in this case the Governor.(c) Others who were supernumerary, being present in the commandery for purposes that are not clear and which may have been somewhat varied.

The Han shu gives the situation of the thirty-six agencies for salt and forty-eight for iron that were established throughout the empire in AD 1-2. The documents from Yinwan name three others for salt and two others for iron that are not given there, and which were presumably closed by the time when the list for AD 1-2 was drawn up. For the first time we learn, from the documents, something of the types and numbers of officials who were responsible for overseeing the work of the agencies, which was undertaken by conscript servicemen. There were perhaps twenty-seven officials at the salt agencies and twelve at the iron agencies.

Of particular value is the information that the documents provide regarding the annual administrative procedures whereby the whole empire was governed, and which concerned one of the primary and basic tasks of any government, that of the collection of tax. Specially designated officials were responsible for conducting these procedures, which were termed Shang ji 上計, and consisted of the submission of reports to the central government at Chang'an.  Based on information that the officials assembled from the subordinate units of the commandery, these reports included a statement of financial accounts; figures of the registered population; figures for the land that was under cultivation and for that which had been reclaimed; and the total number of all types of subordinate unity, whether county, nobility or state. Curiously enough, as late as 10 BC they were still using the calendar of Qin, that had otherwise been replaced some hundred years earlier.

For the first time we are able to learn something of the contents of these regular annual returns. They could include maps; they might give the length of the boundaries between commanderies; and they could include the total number of officials, from the Governor to the lowest grade of clerks (Zuo shi 左史), actually in their posts in the commandery, i.e. 2,203, or 2,202 for Donghai. [9]  As in Chapter 28 of the Han shu, these documents give the total figures for the registered population in terms of households (hu ) and individuals (kou ), as follows:(1)Document YM6D1: 266,290 households; 1,397,34 individuals (2)Han shu 28A, p. 1588 : 358,414 households; 1,559,357 individuals. In the figures that are given in the documents, the ratio of household to individual stands at 1:5.2; this is somewhat higher than the general ratio of 1:4.6 that derives from the figures given in the Han shu for the whole empire at AD 1-2. [10]

The reports also gave the extent of the tax that had been collected, either in grain or in cash. It is of considerable interest to note that the figures for the population are broken down according to variations of age, for example for the very young and the very old, thus allowing for corresponding differences of tax assessment and the grant of special privileges. In addition the documents make a special point of stating the numbers of vagrants who had been reclaimed for the register and were thus liable to the statutory duties of tax and service. Such increases stood to the credit of the Governor.

It is clear that these reports were prepared with considerable care, but while they would seem to have been drawn up with consistency, there is no immediate check on how far the figures that were recorded corresponded accurately with the facts of the actual situation. Some fifty years previously an imperial decree had referred to the deception as practiced in their preparation or presentation. [11]  As seen above there is some difference in the figures given for the population of Donghai commandery in the documents and Han shu. They are however reasonably close, and it could hardly be expected that the counts would remain identical over the fifteen years that lay between them.

As has been noted the documents give the total number of officials posted in the commandery at 2,202 or 2,203, and fortunately these derived from a not inconsiderable number of subordinate units (38). With all allowances that must be made for basing a general estimate on a particular sample, it may be calculated that the probable number of officials posted in the 83 commanderies and 20 kingdoms of the Han Empire and their 1,577 subordinate units in 15-10 BC was 99,214. This figure may be compared with that given in the Han shu for the total number of officials of the central and provincial offices of government for 5 BC, which is 130,285. [12]  From these two sets of figures, it may be supposed that the number of officials of the central government, posted in Chang'an, was of the order of 30,000.  At the time the figures for the population of Chang'an itself are given as 80,800 households, 246,200 individuals, and Professor Bielenstein interprets these as deriving from taxation lists rather than from actual counts of the population. [13]  For purposes of comparison, it may also be noted that the estimate for the total number of officials of the Later Han empire, as suggested by Du You 杜佑(735-812) was 152,986. [14]

21 November 2001



[1] See Wenwu 1974. 6, p.7, 1974.7, pp. 49-63, and T'oung Pao 63 (1977), p. 125.
[2] Wenwu 2001.5, p. 95.
[3] Yinwan Han mu jiandu 尹灣漢墓簡讀(Beijing: Xinhua shuju, 1977).
[4] See the articles collected in Yinwan Han mu jian zong lun 尹灣漢墓簡綜論 (Beijing: Xinhua shuju, 1999).
[5] Yuanyan er nian riji 元延二年日記 (strips nos. 1-76).
[6] E.g., see Han Tang bi hua 漢唐壁畫, Plates 9 and 11, for paintings from Wangdu tomb no. 1, and  Helinge'er Han mu bihua 和林格爾漢墓壁畫, p. 49.
[7] YM6D2.
[8] YM6D5 verso; transcription, in Yinwan Han mu jiandu, p. 100.
[9] These two figures derived from two separate documents, i.e. YM6D1 (for 2,203) and YM6D2 (for 2,202).
[10] These are given in Han shu 28, p. 1640 as 12,233,062 households and 59,594,978 individuals. As Professor Bielenstein has shown the sum of the figures given for the 103 commanderies and kingdoms actually amounts to 12, 366,470 and 57, 671,400.
[11] Han shu 8, p. 273.
[12] Han shu 19A, p. 743, where it is given as 120,285; see Han shu bu zhu 19A.31b (Wang Xianqian's note) and Bielenstein, The bureaucracy of Han times, pp. 156 and 205 note 1 for the correction.
[13] Han shu 28A, p. 1543. Being given in round numbers these figures may not necessarily be based on an actual count. See Bielenstein, 'Lo-yang in Later Han Times', in the Bulletin of the Museum of Far Eastern Antiquities 48 (1976), p. 19.
[14] Tong dian 通典 36, p. 205c.

Last Updated: 1/21/12