Hostname: page-component-745bb68f8f-grxwn Total loading time: 0 Render date: 2025-02-11T07:43:12.644Z Has data issue: false hasContentIssue false

A Study of the Tang Dynasty Tax Textiles (Yongdiao Bu) from Turfan

Published online by Cambridge University Press:  19 July 2013

BINGHUA WANG
Affiliation:
School of Chinese Classics, Renmin University of China
HELEN WANG
Affiliation:
School of Chinese Classics, Renmin University of China
Rights & Permissions [Opens in a new window]

Extract

The Tang dynasty tax system is often referred to as zuyongdiao 租庸調, with each of these three terms having a specific meaning. Zu was the annual collection of 2 piculs, roughly 120 litres of grain (su 粟) per head. Yong was the annual corvée (labour) duty of 21 days per head, which could be substituted and paid for in silk or cloth, or the cloth-paid-in-place-of-annual-corvée tax. Diao was the tax in kind, payable in textiles (substitutions of other goods were sometimes permitted). In silk-producing areas the diao tax per head was 2 decafeet of silk, payable in ling-twill, juan-silk or silk thread (si 絲) and 3 ounces of silk floss (simian 絲綿). In areas that did not produce silk, the diao tax was 2 decafeet, 5 feet of hemp cloth and 3 pounds of hemp yarn.

Type
Research Article
Copyright
Copyright © The Royal Asiatic Society 2013 

The Tang dynasty tax system is often referred to as zuyongdiao 租庸調, with each of these three terms having a specific meaning. Zu was the annual collection of 2 piculs, roughly 120 litres of grain (su 粟) per head.Footnote 1 Yong was the annual corvée (labour) duty of 21 days per head, which could be substituted and paid for in silk or cloth, or the cloth-paid-in-place-of-annual-corvée tax.Footnote 2 Diao was the tax in kind, payable in textiles (substitutions of other goods were sometimes permitted). In silk-producing areas the diao tax per head was 2 decafeet of silk, payable in ling-twill, juan-silk or silk thread (si 絲) and 3 ounces of silk floss (simian 絲綿). In areas that did not produce silk, the diao tax was 2 decafeet, 5 feet of hemp cloth and 3 pounds of hemp yarn.

Since the 1950s the archaeological team at the Xinjiang Museum has carried out several seasons of clearing and excavation at the Astana and Karakhoja cemeteries in Turfan, and has unearthed a substantial amount of woven hemp cloth in the form of sheets, bedcovers, socks/leggings, grain sacks and other items, ranging in date from the Jin dynasties (Western Jin 265–317; Eastern Jin 317–420) to the Tang dynasty (618–907). There are also some woven hemp pieces which have a handwritten inscription in black ink, giving the date (year and month), place name, description of the textile, name of the taxpayer, quantity paid and stamped impressions of the seal of the prefecture, county and storehouse.Footnote 3 These are clearly tax textiles (yongdiao bu 庸調布), and provide us with a valuable body of primary evidence that reflects the Tang dynasty system of tax collection. Other woven hemp pieces may also have served as tax textiles but do not have inscriptions and are not discussed here.

The author worked together with Li Zheng on a preliminary examination of the woven hemp textiles unearthed at the Astana cemetery between 1972 and 1973, and selected 12 tax textiles with inscriptions for further study. To these he added another five similar tax textiles in hemp and three pieces of tax textiles in silk that had been published or exhibited previously. A total of 20 inscribed pieces are listed below, followed by a discussion. With reference to the Tang historical literature, the author discusses the inscriptions (dates, types of inscriptions and seals, names of tax textiles and place names), how hemp and silk textiles arrived in Xinjiang and the significance of the loom width and weave density.

List of Specimens

A. The 17 woven hemp tax textile documents from the Astana cemetery

A.1 Hemp bed-sheet (73TAM192:5)Footnote 4

Width 59.5 cm, length 73 cm (incomplete). 25 warp and 18–25 weft threads per sq cm. This specimen has a particularly dense weave. Pale yellow, almost white, in colour. An epitaph from the same tomb names the occupant as yiwei-guard (翊衛) Zhang Daliang 張大良 and is dated 724 (Kaiyuan year 12).Footnote 5 There is one seal impression, in red, 5.4 × 5.5 cm, in seal script, reading Changshui xian yin 長水縣印, “Seal of Changshui county”. The inscription reads:

河南府長水縣歸仁鄉劉元楷 (black ink, regular script) // 行達 (black ink, upside-down, grass script)

Liu Yuankai, Guiren canton, Changshui county, Henan prefecture //[Signature of] Xing Da

A.2 Hemp bed-sheet (72TAM225:1)Footnote 6

Width 52.5 cm, length 115 cm. 15 warp and 11 weft threads per sq cm. With five stamped seal impressions, three of which are stamped over the inscription. Documents dated 700 and 702 were found in this tomb and provide a date for this piece.Footnote 7 There are three seal impressions on this piece. The first is red, square, 5.5cm, in relief seal script, reading Xuanzhou zhi yin 宣州之印, “Seal of Xuanzhou [prefecture]”. The second is also red and square, reading Liyang xian zhi yin 溧陽縣之印, “Seal of Liyang county”. The third, stamped at the bottom, is illegible. The inscription reads:

宣州溧陽縣 (on right-hand side, black ink, regular script) // 超 (black ink, grass script) // 定中 (black ink, grass script) // 宗慎 (on reverse, in upper left-hand corner, black ink, grass script)

Liyang county, Xuanzhou prefecture //[Signature of] Chao //[Signature of] Dingzhong // [Signature of] Zongshen.

A.3 Hemp bed-pad cover with purple juan-silk edging (72TAM218:17)

Width 55.5 cm; length 96 cm. 11 warp and 8 weft threads per sq cm. This piece is dated 721 (Kaiyuan year 9). There is one seal impression, in red, square, 5.5 cm, in seal script, reading [. . .] dudufu yin 都督府印, “Seal of the [. . .] dudufu”. The inscription reads:

梁州都督府開元九年八月日 (on left-hand side, black ink) // 侗 (Signature, black ink, regular script) [blank space left so that the day could be filled in later]

Liangzhou area command, Xth day of the 8th month of the 9th year of the Kaiyuan reign // [Signature of] Dong

A.4 Hemp bed-pad cover with juan-silk edging (72TAM157:4)

Width 56 cm; length 155 cm (incomplete). 10 warp and 9 weft threads per sq cm. Three seal impressions. The construction of the tomb, with its ‘heaven-well opening’[ 天井 tianjing], and other burial goods are characteristic of the mid-Tang. There are three seal impressions in red, all illegible. The inscription reads:

[. . .][. . .] 州西鄉縣云 [. . .] 鄉庸調布一端 (at edge, black ink, regular script) // 庭 [. . .]

1. duan of cloth-paid-in-place-of-the-annual-corvée-tax from Yun [. . .], Xixiang county, [. . .] zhou prefecture //[Signature of] Ting [and Illegible inscription]

A.5 Fragment of hemp cloth (72TAM157:8)

Width 27 cm; length 33 cm (incomplete). 10 warp and 10 weft threads per sq cm. Same date as no. 4 (mid-Tang). The inscription reads:

何儀 (upper-right corner, black ink) // 李藝 (upper-left corner, black ink)

[Signature of] He Yi //[Signature of] Li Yi

A.6 Pair of hemp cloth socks/leggings (72TAM157:5)

Leg length 28 cm; foot length 30 cm. 11 warp and 9 weft threads per sq cm. Date as nos 4–5 (mid-Tang). The inscription reads:

庸布一端伯 (black ink)

1.duan of cloth-paid-in-place-of-the-annual-corvée-tax //[Signature of]] Bai

A.7 Fragment of hemp cloth (73TAM232:15)

Width 52.5 cm; length 80 cm. 11 warp and 12 weft threads per sq cm. This piece is dated 680 (Diaolu year 2). The inscription reads:

調露二年八月 (on right-hand side, black ink, regular script) // 違言 (grass script) // 惠 (grass script)

8th month of the 2nd year of the Diaolu reign //[Signature of] Wei yan //[Signature of] Hui

A.8 Hemp bed-sheet with printed pattern sha-silk edging (72TAM167:4)

Width 56 cm; length 220 cm. 14 warp and 10 weft threads per sq cm. The construction of the tomb and burial goods is characteristic of the mid-Tang. The inscription reads:

梅思慎布保玉鄉 (black ink)Footnote 8

Cloth [paid by] Mei Sishen // Baoyu canton

A.9 Hemp cloth (72TAM194:13)

Width 56 cm; length 156 cm.12 warp and 9 weft threads per sq cm. An epitaph names the tomb occupant as Zhang Xinglun 張行綸 and is dated 719 (Kaiyuan year 7). There are four seal impressions in red, all illegible. The inscription reads:

杜 //均州 [. . .] 鄉[. . .] 租丁杜國布一端 // 炣 (right-hand side, black ink)

Du // 1 duan of cloth from Du Guo, zu-tax payer of [. . .], in [. . .] canton, Junzhou // [Signature of] Ke

A.10 Hemp cloth (72TAM194:9)

Width 56 cm; length 175 cm. 12 warp and 8 weft threads per sq cm. A seal impression, in red, is stamped over inscription. The inscription reads:

婺州通 (black ink)

Wuzhou [prefecture]//[Signature of] Tong

A.11 Hemp bed-sheet covered with juan-silk (72TAM214:129)

Blue juan-silk cover; hemp lining. Width 50.5 cm (incomplete); length 95 cm. 11 warp threads and 11 weft threads per sq cm. An epitaph from this tomb names the occupant as Madam Qu, wife of Mr Zhang (Zhang Jun qi Qu shi 張君妻麴氏) and is dated 665 (Linde 麟德 year 2). There are two red seals, both illegible. The inscription on the bed-sheet reads:

[. . .][. . .] 縣新成鄉祖華里戶主弘政 (black ink)

Head of household Hong Zheng, of the Zuhua ward, Xincheng canton, [. . .] county

A.12 Fragment of hemp cloth (72TAM191:107)

Width 55 cm; length 31 cm. There is one seal stamped in red over the signature, readingAnji xian yin 安吉縣印, “Seal of Anji county”; and two seals stamped in black, both illegible. The inscription on the cloth reads:

湖州安吉縣無 [. . .] 鄉清鏡里施恩景布 (black ink, regular script) // 森 (black ink, signature, written in crude style)

Cloth from Shi Enjing, of Qingjing ward, Wu [. .] canton, Anji county, Huzhou //[Signature of] Sen

B. The following four pieces have been published previouslyFootnote 9

B.13 Tax textile document (68TAM108:16)Footnote 10

This document is dated 721 (Kaiyuan year 9). Inscription in black ink. There are three seal impressions stamped over the county name: all are in red, and read Yunxian zhi yin 鄖縣之印, “Seal of Yun county”.

西浦里賀思敬 //鄖縣光同鄉賀思敬庸調布一端 //開元九年八月日 // 專知官主簿苑 (in black ink)

He Sijing, of Xipu ward // 1 duan of cloth-paid-in-place-of-the-annual-corvée-tax from He Sijing of Guangtong canton //[. . .]th day of the 8th month of the 9th year of the Kaiyuan reign // Yuan, Specially Appointed Recorder

B.14 Jiaobu ‘transport tax-textile’ (67TAM96:4)Footnote 11

Width 55.5 cm; length 209 cm. The tomb construction and burial goods date this piece to the Tang dynasty. Two seal impressions, black, illegible, to the left of the inscription. The place name 婺州蘭溪縣 (Wuzhou Lanxi xian) is written in larger characters, the rest in smaller characters. The signature of Bao Liang 鮑良 is written in a larger, darker hand. The inscription reads:

婺州蘭溪縣歸德鄉 [. . .] 招里吳德吳護兩人共一端作腳布鮑良 (in black ink)

  1. 1. duan of transport tax textile jointly submitted by two people Wu De and Wu Hu, of [. . .] chao ward, Guide canton, Lanxi county, Wuzhou [prefecture]. [Signature of] Bao Liang.

B.15 Hemp cloth (68TAM106:5)Footnote 12

The tomb construction and burial goods date this piece to the Tang dynasty. The inscription reads:

奉 [. . .] 鄉申[. . .] 里陳禮[. . .][. . .] (black ink) //詷 (signature, black ink, to side of main inscription)

[. . .] Chen Li of Shen[. . .] ward, Feng[. . .] locality //[Signature of] Tong

B.16 Painting on hemp cloth (67TAM76:11)Footnote 13

Painting of Fu Yi and Nü Wa. Burial goods from this tomb date this piece to the Tang dynasty. The inscription is positioned at the four corners on the reverse, and reads:

陵州//師

Lingzhou [prefecture]//[Signature of] Shi

B.17 Hemp cloth (60TAM340)Footnote 14

Width 53.5 cm; length 263 cm. Inscription in black ink. This piece is dated 681 (Yonglong year 2). The personal name Tian Yuanqing 田元卿 appears twice, and in a different hand (different style, darker ink) from the rest of the inscription. The inscription reads:

蠻田元卿 // 澧州慈利縣讓德鄉永樂里住戶主田元卿調 //布一端永隆二年//八月日覽

Tian Yuanqing, member of the Man people // 1 duan of cloth-paid-as-tax-in-kind from head-of-household Tian Yuanqing, of Yongle ward, Rangde canton, Cili county, Lizhou // [. . .]th day of the 8th month of the 2nd year of the Yonglong reign // Inspected

C. The following three silk pieces have been published or exhibited previously

The following three silk tax documents (yongdiao juanling) were found in tombs at the Astana cemetery, and are of the same date.

C.18 Fine silk paid-in-place-of-the-annual-corvée-tax (72TAM226:16)Footnote 15

Width 4.5 cm, length 25 cm. This is only a small fragment of the original full piece. White, slightly yellow. Plain weave background, with vivid flower pattern. This piece is dated 710 (Jingyun year 1).

景元元年折調細綾一匹 //雙流縣以同官主火愉 (in black ink)

1st year of the Jingyun reign, 1 bolt of fine silk twill paid-in-place-of-the-annual-corvée-tax // Huo Yu, Head Official of Yitong, Shuangliu county.

C.19 Red juan-silk (72TAM227:4)Footnote 16

Width 25 cm; length 33.5 cm. Red juan-silk, fragment of a skirt which had been used to dress a wooden figurine. This piece bears a seal impression, 6 × 5.5 cm, but no inscription written in black ink.

益州都督府之印

Seal of the Yizhou army command

C.20 White juan-silk (64TAM36:11)Footnote 17

Loosely woven. This piece is dated 713 (Xiantian year 2). It bears a seal impression, in red, in seal script [but the author does not give details]. The inscription in black ink reads:

先天二年八月 (right-hand side, black ink)

8th month of the 2nd year of the Xiantian reign

Discussion

1. The importance of the inscriptions

The inscriptions not only identify these pieces as tax textiles but also allow us to compare the archaeological evidence with the historical record.

Dates

Five of the 20 specimens listed above are dated, and give the year and month in which the tax payment was made:

  1. (A.7) 8th month of Diaolu year 2 (680)

  2. (B.17) 8th month of Yonglong year 2 (681)

  3. (C.20) 8th month of Xiantian year 2 (713)

  4. (A.3) 8th month of Kaiyuan year 9 (721)

  5. (B.13) 8th month of Kaiyuan year 9 (721)

They show that there was strict adherence to the early Tang regulation that: “Transport of all yongdiao items should take place annually in the first half of the eighth month, and should be completed within 30 days”.Footnote 18 Indeed, all the dated tax textiles we have seen record that the payment was made in the eighth month.Footnote 19 After the Tianbao period (742–756), the date of collection was eased and extended to the end of the ninth month, in order to facilitate agricultural activity. All the dated tax textiles we have seen refer to tax collection before the Tianbao period, and are in accordance with the regulations.

The function of inscriptions and seals

Regulations concerning the collection of yongdiao stated that the various types of cloth and silk were to be inspected by county officials, who were to mark them with an inscription and stamp them with a seal.Footnote 20 According to the Da Tang liudian [Compendium of administrative law of the six divisions of the Tang bureaucracy]:

凡物之精者與地之近者供御, 謂之 (支) 納司農, 太府, 將作, 少府等物. 物之固者與地之遠者以供軍, 謂支納邊軍及諸都督, 都護府.

The finest goods and produce of nearby areas are to be presented to the imperial court. These goods are to be passed on to the Court of the Imperial Granaries, Court of the Imperial Treasury, Directorate for the Palace Buildings and Directorate for Imperial Manufactures. Coarser goods and produce of more distant areas are to be presented to the military, and are to be called “goods for the military at the borders and for the various Area Commands and Protectorates”.Footnote 21

Some of the 20 specimens have inscriptions written by county-level officials upon receipt and inspection of the textiles:

  1. (A.6) 庸布一端伯 Yong bu yi duan bai

  2. (A.8) 梅思慎布保玉鄉 Mei si shen bu Baoyu xiang

  3. (B.15) 奉 [. . .] 鄉申[. . .] 里陳禮 [. . .][. . .] Feng [. . .] xiang shen [. . .] li Chen Li [. . .][. . .]

Others have inscriptions recording their transfer from county and prefectural level up to central government:

  1. (A.1) 河南府長水縣歸仁鄉劉元楷 Henan-fu Changshui xian Guiren xiang Liu Yuankai

  2. (A.2) 宣州溧陽縣 Xuanzhou Liyang xian

  3. (A.4) [. . .][. . .] 州西鄉縣云 [. . .] 鄉庸調布一端 [. . .]zhou Xixiang xian Yuna[. . .] xiang yongdiao bu yi duan

When yongdiao hemp cloth and silk were received by the central government, they entered the Left Storehouse (the palace depository for valuables) and became the responsibility of the Left Storehouse Office (左藏署 zuocang shu). This Office also served as a transport centre:

凡天下賦調先於輸場簡其合尺度斤兩者, 卿及御史監閱, 然後納於庫藏, 皆題以州, 縣, 年, 月, 所以別粗良, 辨新舊也

All goods received in tax payment throughout the empire are to be taken to the transport centre for measuring and weighing, and for inspection by the Chief Ministers and Censors. They are to be deposited in the national storehouses and the imperial purse, and each piece is to be given an inscription recording the name of the prefecture and county, year and month. On that basis the goods are to be sorted by quality and age.Footnote 22

The inscriptions on the textiles state the prefecture, county, locality, month and year. To the side of the inscription there is sometimes a signature (in regular or grass script), written in a hand different from the main inscription. The signature certifies that the piece has passed inspection. Official seals were stamped in red or black. The red seal impressions were made at prefectural and county level, and reflect the process of receiving textiles and transferring them up to a higher authority. The black seal impressions certify that the pieces have been approved by the Left Storehouse Office. According to the Da Tang liudian, when the Left Storehouse Office subsequently removed any cloth or silk from the storehouses, official sanction was required:

先勘本契, 然後錄其名數及請人姓名, 署印送監門衛, 乃聽出, 若外給者, 以墨印印之

First check the original record, then record the name and quantity, and the name of the person making the request, then stamp it with his seal and deliver it to the Palace Gate Guard for permission for it to be removed. If it came from a storehouse outside the imperial city, then it is to be stamped with his seal in black.Footnote 23

Names of tax textiles

Different names for tax textiles appear on the 20 documents above: for example, diaobu 調布 (B.17), yongdiao bu 庸調布 (A.4; B.13) and its abbreviated form yongbu 庸布 (A.6), and jiaobu 腳布 (used to cover transportation costs) (B.14). The term zubu 租布 (cloth paid in place of grain tax) is also seen on a tax textile found by Aurel Stein in Turfan (see below).

Yongdiao bu is a combination of the two terms yongbu and diaobu. Tang regulations stated that each man had to do two 10-day periods (旬 xun) of corvée duty per year. Or he could transmute the corvée duty and pay its equivalent in yong, this being 3 feet [of silk] per day, with an additional fifth if paid in hemp cloth (and not in silk).Footnote 24 Yongbu was the name given to cloth-paid-in-place-of-annual-corvée tax.

Zubu is seen on a hemp tax textile dated 684, found by Aurel Stein in Turfan.Footnote 25 The inscription on this piece reads:

婺州信安縣顯德鄉梅山里祝伯亮租布一端光宅元年十一月日

1 duan of cloth-paid-in-place-of-the-grain-tax from Zhu Boliang of Meishan ward, Xiande canton, Xin'an county, Wuzhou. [. . .]th day of the 11th month of the 1st year of the Guangzhai reign (684–685 ce).

The inscription tells us that on this occasion cloth was paid as a substitute for the zu-tax. This is unusual and may relate to exceptional circumstances. An edict issued some 50 years later, in 737 (Kaiyuan year 25), suggests a plausible explanation: namely, that there was a shortage of silkworms and mulberries in the Guanfu 關輔 region around the capital of Chang'an and therefore yongdiao was paid instead of the zu-tax. The inscription reveals the hardship in the area and of the very high costs of transporting goods:

常賤糶貴買損費愈深

People (in the Guannei Circuit) often had to sell [grain] for a low price and to buy [yongdiao bu] for a high price [in order to pay their yongdiao tax obligation].

又江淮苦變造之勞, 河路增轉輸之弊, 每計其運腳, 數倍加錢

Furthermore, Jianghuai [region] suffered from the costs of converting one tax commodity to another, [and the people on] the Grand Canal suffered the additional abuse of shipping, so that each [year], calculating the transport costs, their payments would be multiplied many times.

均其餘以減遠費, 順其便使農無傷

Adjust the payment of cloth in place of corvée tax so that people in different regions can pay using whatever commodities they have in abundance. This will fit local conditions, and there will be no harm to peasants.

關內諸州庸調資課, 并宜準時價變粟取米, 送至京, 逐要支用其路遠處, 不可運送者, 宜所在收貯, 便充隨近軍糧, 其河南河北, 有不通水利, 宜折租造絹, 以代關中調課

As for the yongdiao, zi, and ke taxes in the Guannei region, it is appropriate to pay using grain and rice that have been converted according to the current rate, send to the capital, and subsequently pay out. In remote places, where it is difficult to transport goods, then the goods are to be kept in storehouses, and used to supply troops. In Henan Circuit and Hebei Circuit, where water transport is poor, it is appropriate to pay the tax using juan-silk, . . .Footnote 26

The inscription on the tax textile dated 684 suggests that this practical solution to difficult circumstances may have been in use long before the Kaiyuan reign, and already established in the reign of Wu Zetian. Indeed, a number of the tax textiles may indicate a similar usage. For example, if zuding 租丁in document A.10 refers to the per head tax payment in grain, then zuding bu 租丁布 would indicate that this payment was transmuted and paid in cloth.

Jiaobu requires some explanation. It may, as some have suggested,Footnote 27 be a ‘catch-all’ name for the various transport costs incurred as tax grain, tax textile and other official items passed through the various levels on their way up to higher authorities. Transport costs (运脚 yunjiao) were a significant burden to the people. A Tang dynasty regulation stated that:

凡天下舟車水路載運, 皆具為腳值輕重貴賤平易險澀而為之制

All transport throughout the empire, on land or water, incurs specific transport costs. These are determined by the weight and value of the goods transported, as well as the risks involved.

河南河北河東關內等四道諸州運租庸雜物等腳, 每馱一百斤一百里一百文山坡處一百二十文車載一千斤九百文黃河等水運腳費從幽州運至平州上水十六文, 下水六文餘水上十五文, 下五文從澧荊等州至揚州四文山陵險灘驢少處不得過一百五十文, 平易處不得下八十文

The costs for transporting zuyong goods in the various prefectures within the four circuits of Henan, Hebei, Hedong and Guannei are as follows: 100 coins per packload of 100 pounds carried a distance of 100 third-miles. In mountainous areas this rises to 120 coins; and 900 coins per 1,000 pounds when carried by cart. There are specific jiao fees for transporting goods by river: on the Huang He (Yellow River) from Youzhou to Pingzhou it costs 16 coins to travel upstream (against the current) and 6 coins to travel downstream (with the current). On other rivers it costs 15 coins and 5 coins, respectively. A river journey from Lizhou and Jingzhou to Yangzhou costs 4 coins. In mountainous regions, and wherever donkeys are scarce, the cost cannot rise above 150 coins; and in regions with easier terrain the cost cannot fall below 80 coins.Footnote 28

In document B.14 above, two people, Wu De and Wu Hu, shared the payment of 1 bolt of jiaobu, and their record survives as evidence of such additional charges and exploitation. It also provides irrefutable evidence validating the rule that when collecting yongdiao taxes:

若當戶不成匹端屯綟者, 皆隨進 [the correct character is 近] 合成

Any householders that do not make up a full bolt or a full hank [of silk floss, cotton balls, or hemp not yet woven into cloth] should join up with others to make a full payment.Footnote 29

2. The origins of the tax textiles

The Tang dynasty tax textiles from Astana were mostly unearthed during the clearing of the tombs in 1972–1973. The inscriptions on these documents reveal that they came from six provinces: Henan, Shannan East, Shannan West, Jiangnan East, Jiangnan West and Jiannan, which correspond with the modern-day provinces of Henan, Shaanxi, Hubei, Hunan, Sichuan, Jiangsu and Zhejiang. The inscriptions reflect the high level of unity in Tang dynasty China, and the close political and economic relationship between Xinjiang and central China. The historical records for these locations confirm that they were well known for their textile industries.Footnote 30

Changshui county in Henan-fu (see A.1) corresponds with modern-day Luoning county in Henan province. In the Tang dynasty it was in Henan province.

Liyang county in Xuanzhou (see A.2) corresponds with modern-day Liyang county in Jiangsu province. It is located on the southern side (yang 陽) of the Li River, hence the name Liyang. In the Tang dynasty Liyang was in Jiangnan province, and came under the jurisdiction of Xuanzhou prefecture. Xuanzhou prefecture was known for its woven hemp products, including a kind of hemp cloth (baizhu bu 白纻布) which was submitted as tribute to the court during the Kaiyuan reign (713–741). Later, during the Guangqi reign (885–888), Liyang came under the jurisdiction of Shengzhou.Footnote 31

The Liangzhou government-general (dudufu) (see A.3) included the four prefectures of Liangzhou, Yangzhou, Jizhou and Xingzhou. This was an important region, with the political seat of the Liangzhou government-general located at modern-day Nanzheng county in Shaanxi province.Footnote 32 During the Kaiyuan reign (713–741) the Liangzhou region submitted “red flowers” (紅花 honghua), tax [hemp]cloth (賦布 fubu), hemp (麻 ma), silk floss (綿 mian) and juan-silk as tribute.Footnote 33 Tax textile A.3 provides evidence that diaobu was indeed supplied by Liangzhou in 721.

Xixiang county (see A.4) corresponds with the modern-day Hanzhong region in Shaanxi province. During the Tang dynasty, it was one of the four counties (Xixiang, Huangjin, Xingdao and Zhenfu) under the jurisdiction of Yangzhou in Shannan West province, and was the political centre of Yangzhou. It was known for its woven hemp cloth, and tribute goods included ‘white interlocked shuttle [-cloth]’ (白交梭 baijiaosuo), hemp cloth (火麻布 huoma bu) and wild ramie (野苧麻 ye zhuma).Footnote 34

Junzhou (see A.9, A.13) was part of Wudang jun in Shannan East province. There were three counties under the jurisdiction of Junzhou: Wudang, Yunxiang and Fengli.Footnote 35 The inscription on A.9 refers to Yunxiang county, which corresponds with modern-day Yun county in Hubei province. Junzhou paid tribute in tax silk (賦絹 fu juan), cloth (布 bu), silk floss (綿 mian) and hemp (麻 ma),Footnote 36 which matches the inscription in A.9.

Wuzhou (see A.10, A.15) corresponds with the modern-day Jinhua region in Zhejiang province. In the Tang dynasty Wuzhou was in Jiangnan province, and had seven counties under its jurisdiction, including Jinhua and Lanxi. Wuzhou paid tributes of silk floss (綿 mian), kudzu vine (葛 ge) [the fibres of which were used to make cloth] and hemp cloth (纻布zhubu).Footnote 37

Anji county (see A.12) corresponds with modern-day Anji county in Zhejiang province. In the Tang dynasty it was one of five counties under the jurisdiction of Huzhou (in Wuxing jun) in Jiangnan province. Huzhou had flourishing silk and hemp industries and paid tribute in ‘imperial robes’ (御服 yufu), ‘bird's-eye twill’ (鳥眼綾 niaoyan ling), ‘split black cloth’ (折皂布 zhezao bu), silk (綿綢 mianchou) and ‘cloth hemp’ (布纻 buzhu).Footnote 38

Yun(?) county appears in a tax textile document (see A.13), but does not appear in the historical records.Footnote 39 The character clearly has the elements 貝 and 阝, but the 口 is not clear. If the character were read as 鄮 mao, then it could refer to Maoxian in Mingzhou (in Yuyao jun) in Tang dynasty Jiangnan province. Maoxian was the political seat of Mingzhou, and corresponds with modern-day Ningbo in Zhejiang province.

Cili county (see B.17) corresponds with modern-day Cili county in Hunan province. During the Tang dynasty Cili was one of four counties (Liyang, Anxiang, Shimen and Cili) under the jurisdiction of Lizhou (in Liyang jun) in Shannan province. Lizhou's tribute goods included patterned damask (紋綾 wenling) and ‘hemp binding cloths’ (纻綀縛巾 zhushu fujin).Footnote 40 The four characters at the beginning of the inscription on B.17 identify the ethnicity of Tian Yuanqing as Man. In the past, Hunan was an area populated by the Man people. Tian was one of the most common family names of the Man, and is seen frequently in historical records.Footnote 41 Tian may have been the name of the leaders of the Man, or the name of an important person. During the Northern and Southern Dynasties (420–589), the Man paid lighter taxes than the Han people in this region. A regulation stated that the Man and those attached to them were required to send in several hu of grain per household and nothing else.Footnote 42 This explains why, during the Liu-Song period (420–479), many Han people escaped heavy taxation by fleeing to the mountainous areas and joining the Man. But, as tax textile B.17 shows, by the time of the Tang dynasty the Man were also paying yongdiao.

Lingzhou (see B.16) corresponds with modern-day Renshou county, in Sichuan province. During the Tang dynasty Lingzhou was in Jiannan province, and had five counties under its jurisdiction, including Renshou, Guiping and Jingyan. Its tribute goods included panned gold from river beds (麩金 fujin), ‘goose stream silk’ (鵝溪絹 exi juan) and fine kudzu vine fabric (細葛 xige) and medicines.Footnote 43

Shuangliu county (see C.18) corresponds with the modern-day Chengdu region, in Sichuan province. In the Tang dynasty Shuangliu was one of the 10 counties under the jurisdiction of Chengdu-fu in Jiannan province.Footnote 44

Yizhou government-general (see C.19) corresponds with part of modern-day Sichuan province. In the Tang dynasty the Yizhou government-general was in Jiannan province, with its political seat close to modern-day Chengdu. Originally called Chengdu-fu, it was changed to Yizhou in 618 (Wude year 1), and had 17 prefectures under its jurisdiction, including Yizhou, Mianzhou and Lingzhou.Footnote 45 Most of the silks (juan and ling) with inscriptions unearthed in Turfan came from the Chengdu plains, and reflect the importance of this area as a centre of textile production in the Tang dynasty.

3. How hemp cloth and silk arrived in Xinjiang

During the Tang dynasty, hemp cloth and silk textiles arrived in Xinjiang in the form of military expenses, meritorious awards and as money.

The Tang histories give details of the geography and political administration of Xinjiang. Xinjiang was in Longyou province, with the key administrative centres being the Anxi high-protectorate (da duhufu) in Kuche, the Beiting protectorate (duhufu) in Beshbaliq and the prefectures of Yizhou (Hami) and Xizhou (Turfan). There were four military bases: the Yiwu army stationed at Gan Lu Chuan north-west of Yiwu (Hami); the Tianshan army stationed at Xizhou; the Hanhai, Qinghai and Jingsai armies stationed within the Beiting protectorate; and the Baoda army stationed at Suiye (modern Tokmak, Kyrgyzstan).Footnote 46 In addition to these military bases, Tang dynasty documents unearthed in tombs in Turfan also refer to the Shule, Yuhe, Jinya and Suxiang armies.Footnote 47 These armies, which were stationed at the borders or sent to deal with special emergencies, are not mentioned in the official histories. Similarly, the names of other military organisations such as guards (守捉 shouzhuo), brigades (鎮 zhen), beacon troops (烽 feng) and garrisons (戍 shu) also appear frequently in the excavated documents from Turfan, although they are not specifically mentioned in the official histories.

The needs of the border troops were met by taxes levied and paid in central China. The Tang histories tell us that

貞觀中. . .侯君集平高昌, 阿史那社爾開西域, 置四鎮. . . 於是歲調山東丁男為戍卒, 繒帛為軍資, 有屯田以資糗糧, 牧使以娩羊馬大軍萬人, 小軍千人, 烽戍邏卒, 萬里相繼, 以卻於強敵.

In the Zhenguan reign [627–649]. . . Hou Junji defeated Gaochang. Ashinashe'er (d.655) opened up the Western Regions and established the Four Garrisons. . . Thereafter, east of the mountains, adult males were sent to be soldiers on guard at the frontier (戍卒 shuzu), silks (繒帛 zengbo) was sent for military expenses, military farms (屯田 tuntian) provided grain and fodder and stockmen (牧使 mushi) reared sheep and horses. There were larger armies of 10,000 men, and smaller armies of 1,000 men. These men were stationed one after another over 10,000 li, working at the beacons, on patrol and on guard (烽戍邏卒 feng shu luo zu), providing a strong defence against the enemy.Footnote 48

The costs of sustaining such a military presence were not insubstantial. The Tang histories speak of the tens of thousands, and hundreds of thousands, of lengths of hemp cloth and silk (布絹端匹 bu juan duan pi) supplied for military expenses, of which a very large proportion was for the Xinjiang region. There are numerous references to these costs in the Tongdian [Encyclopaedic history of institutions]. For example, during the Tianbao reign (742–756):

諸色資課及勾剝所獲不在其中. . . 租稅庸調每歲錢粟絹棉布約得五千二百二十餘萬端疋屯貫石. . . 其度支歲計粟則二千五百餘萬石, 布絹綿則二千七百餘萬端屯疋. . . 千三百萬入西京, 一百萬入東京, 千三百萬諸道兵賜及和糴並遠小州便充官料郵驛等費.

Even after the people had been fleeced of all they owned, still enough was not raised . . . The Tang government collected zu shui yong diao taxes in coin, grain, silk, cotton and hemp cloth to an annual amount of 50.22 million cash (錢 qian). . . The Tang government estimated expenditure at over 25 million piculs of grain and over 27 million bolts and bundles of bu-hemp, juan-silk and silk floss (綿 mian); 13 million entered the Western Capital, 1 million entered the Eastern Capital, and 13 million were used for military and meritorious awards in the circuits and heqin arrangements, and for official matters and communications in the smaller and more remote prefectures.Footnote 49

自開元中及於天寶. . . 每歲軍用日增其費糴米粟則三百六十萬疋段. . . 伊西北庭八萬, 安西十二萬. . . 給衣則五百三十萬. . . 伊西北庭四十萬, 安西三十萬.

From mid-Kaiyuan to Tianbao. . . the annual expenses for the military continued to grow. The cost of grain and rice was 3.6 million pi and duan [bolts of textiles], of which 80,000 were for Yizhou, Xizhou and Beiting, and 120,000 for Anxi. Clothing costs were 5.3 million, of which 400,000 were for Yizhou, Xizhou and Beiting, and 300,000 for Anxi.Footnote 50

These figures show that the expenses for food and clothing alone in the Western Regions totalled 900,000 pi/duan, equivalent to one tenth of such expenses for the entire nation.

All the figures above, quoted from the Tongdian, represent normal military expenditure, and not exceptional circumstances. They do not include expenditure relating to meritorious awards. Although there are no figures for expenditure on meritorious awards, these are also likely to have been substantial. In one Tang dynasty document unearthed in Turfan, we read:

. . . 其安北都 //護府諸驛賜物於靈州都督府給, 單于大都護符諸驛賜物於(朔)州給, 并請準往//例相知給付不得變. . . 安北都護府. . . 已數於靈州. . .

The gift-goods for various stations under the Anbei Area Command are to be distributed at Lingzhou prefecture, the gift-goods for Shanyu Protectorate are to be distributed at [Shuo]zhou, . . .Footnote 51

From this document we can see that the Tang government every year prepared precise ‘gift items’ (gei fu ci wu) for allocation to the ‘various stations’ under the Area Commands and Protectorates in the borderlands. These ‘gift items’ (ciwu) were collected together in a neighbouring region for subsequent distribution. Another Tang dynasty document unearthed in Turfan and dated 733 (Kaiyuan year 21) provides evidence of the transport of gift items. It mentions a certain Wang Fengxian of Huayuan county, Jingzhao prefecture, and loadmaster Xu Zhong who “[drove] a donkey to deliver weapons and gifts to Anxi”. The Anxi Protectorate then issued a pass (guosuo) for their return to the capital. Something unfortunate happened on the return journey and Wang Fengxian was detained. This document is a detailed account of the entire incident, prepared for the investigation, after which Wang was freed.Footnote 52

After the abolition of the government military system (fubingzhi), the Tang government replaced it with a mercenary system (mubingzhi). In the new system, border guards served for a period of three years, and did not return home until the end of that period. The mercenaries were called ‘strongmen’ (健兒 jian'er), and in order to pay for them the government levied a special tax known as ‘the 20 duan of gift goods’ (賜物二十段 ci wu ershi duan).Footnote 53 The Tang Liudian records:

天下諸軍有健兒. . . 丁壯情願充健兒, 常住邊軍者, 每年加長例給賜, 兼給永年優復其家口情願同去者, 聽至軍州, 各給田地屋宅.

All the armies have “strongmen” . . . who volunteered for the work, and who live for long periods in the border armies. Every year they are rewarded with more gift goods. They take their families with them, and all go willingly, in the knowledge that when they reach the military prefecture, they will be given farming land and a house to live in.Footnote 54

A loan contract written on paper, dated 782 (Jianzhong year 3) and unearthed at Dandan Uiliq, near Khotan, concerns the “strongman” Ma Lingzhi who, in urgent need of cash, borrowed from a local monk Qian Ying at a high rate of interest.Footnote 55 This proves that mercenaries existed and it is likely that their service was rewarded with gift goods.

During the Tang dynasty, silk and hemp functioned like bronze coins as money. An edict of 732 (dated 29th day of the 9th month of the 20th year of the Kaiyuan reign) reads:

綾羅絹布雜貨等交易, 皆合通用, 如聞市肆必須見錢, 深非通理自今後, 與錢貨兼用, 違者准法罪之.

Ling-twill, luo-gauze, juan-silk, bu-hemp and various goods can all be used together when exchanging goods. If market and border officials insist on the exclusive use of coins, this is deeply unreasonable. From now on, coins and goods are to be used together, and anyone found violating this will be charged in accordance with the law.Footnote 56

Many of the documents found in Turfan are ‘market vouchers’ (市券 shi quan) recording trade in livestock and slaves, and reckoning the values not only in copper and silver coins but also in juan-silk and bu-hemp. In this way, it is clear that hemp cloth also arrived in Xinjiang through trading activity among the people.

4. The significance of differences in loom width and weave density

There are huge differences in the loom width and warp and weft thread counts in the 17 tax textiles (yongdiao bu) listed above. In addition to those pieces bearing inscriptions, there are very many more pieces of hemp cloth which do not have inscriptions, ranging from large, complete sheets and covers to small grain bags and socks, etc. From our small selection of relatively complete woven hemp cloths, our analysis shows that they were from the Tang dynasty, and that the loom width and weave density varied enormously. Most pieces had a loom width of 50–60 cm; a smaller number had a loom width of 40–48 cm. The densest thread count was 15 warp and 13 weft threads per square cm; the loosest was 8 warp and 6 weft threads per square cm; with most pieces having 10–12 warp and 10 weft threads per square cm.

Diaobu played an important role in tax revenue for the Tang government. In the second month of 720 (Kaiyuan year 8) the government repeated its regulation relating to yongdiao:

頃者以庸調無憑, 好惡須准, 故遣作樣以頒諸州, 令其好不得過精, 惡不得至濫任土作貢, 防源斯在而諸州送物作巧生端, 苟若副於斤兩, 遂則加其丈尺, 有至五丈為疋者, 理甚不然闊尺八寸, 長四寸, 同文共軌, 其事久行立樣之時, 已載此數若求兩而加尺, 甚暮四而朝三宜令所司簡閱, 有逾於比年常例, 尺寸過多者, 奏聞.

In the past, the quality of unlabelled yongdiao had to be tested, and for this reason we have sent samples to all of the prefectures against which to measure quality. Good should not mean exceptionally fine; bad should not mean falling to pieces. Tribute goods are received according to locality, and should reflect that locality. Yet the prefectures try to cheat when they send in goods, such as short-weighting the goods, so that the length has to be increased. In some cases one bolt has measured five decafeet in length. This is absurd. The width should be 1 foot 8 inches, the length should be 4 decafeet, the inscriptions should be written in the same way and then the piece can be in circulation for a long time. When we established the model examples, these were the criteria. Adjusting the weight or the length is simply fiddling the system. So we have instructed officials to make inspections, and any discrepancies should be reported.Footnote 57

Loom width and weave density may reflect differences in local looms and traditional practices. But the main reason must be that the working people reduced the width of hemp cloth and the density of the weave as a means of resistance against the feudal government and economic exploitation. In regulating the standard of yongdiao bu, the Tang government's chief aim was to limit and attack this resistant behaviour. A wooden measure unearthed in a Tang dynasty tomb in Turfan shows that 1 foot measured 29 cm, and 1 foot 8 inches measured 52 cm.Footnote 58 Of the woven hemp cloth pieces that have been unearthed, the majority are not significantly different from the regulation size, although a small number are. Some of these have a very loose weave density indeed, just half that of a finer weave. Such pieces are an expression of the working people's resistance to feudal authority.

References

1 This article was first published in two publications in the early 1980s, in essentially the same form but with slight variations in the wording: in hui, Zhongguo Tangshi yanjiu (ed.) 中 國唐史研究會, Tang shi yanjiu hui lunwenji 唐史研究會論文集 Vol. 9 (1983), pp. 822 Google Scholar (without illustrations) and in Wenwu 文物 (1981) No. 1, pp. 56–62 (with illustrations, and acknowledging the help of Zhu Lei 朱雷, Jiang Boqin 姜伯勤 and Li Zheng 李徵). The 1983 version was reprinted in Wang Binghua 王炳華, Xiyu kaogu lishi lunji 西域考古歷史論集 [as part of the series Xiyu lishi yuyan yanjiu congshu 歷史語言研究叢書], (Beijing, 2008), pp. 348–359, and was used for this translation. It has also appeared in the volume of Wang Binghua's collected essays, Xiyu kaogu wencun 西域考古文存 [Collected essays on the archaeology of the Western Regions] (Lanzhou, 2010), pp. 483–495. The translator has taken the liberty of occasionally adding basic explanatory information helpful for a non-specialist English reader. For this reason, the footnote numbering in the translation does not correspond to the original footnotes, and (WBH, fn) indicates the author's original footnotes.

2 Following Twitchett's, Denis translation, in his Financial Administration under the T'ang Dynasty (Cambridge, 1970)Google Scholar.

3 (WBH, fn 1) Museum, Xinjiang (ed.), Xinjiang chutu wenwu 新疆出土文物 [Excavated artifacts from Xinjiang] (Beijing, 1975), pl. 162 Google Scholar.

4 The identifying numbers give the date, findspot, type of site and number of each object. Hence, 73TAM192:5 means object no. 5 from TAM192 (Turfan Astana mu-tomb no. 192), found in 1973.

5 The epitaph reads 開元十二年歲次己酉甲子廿日故翊衛張大良之墓 Kaiyuan shi er nian sui ci jiyou jia zi nian ri gu yi wei Zhang Daliang zhi mu (the Wenwu 1981 version gives 張大良之靈 Zhang Daliang zhi ling)

“Tomb of yiwei-guard Zhang Daliang, 20th day of jiyou jiazi in the 12th year of the Kaiyuan reign”. There is a problem with this date; in the 60-year cycle, the year 724 was a jiazi year, not a jiyou year. Apparently, the author wrote jiyou by mistake, and someone – maybe the author? – wrote jiazi next to jiyou, so that it would indicate the jiyou month in the jiazi year. See the photo of the original epitaph in Hou Can 侯燦 and Meilin, Wu 吳美琳 (eds), Tulufan chutu zhuanzhi jizhu 吐魯番出土磚誌集注 [Annotated collection of brick inscriptions discovered in Turfan)] (Chengdu, 2003), pp. 638639 Google Scholar.

6 Illustrated in the Wenwu 1981 version of this article, fig.1, right.

7 The documents were dated 久視二年二月700 CE (2nd month of the 2nd year of the Jiushi reign) and 長安二年五月廿二日(22nd day of the 5th month of the 2nd year of the Chang'an reign). Wu Zetian introduced the Jiushi reign in 700 but also introduced the Dazu reign in the first month of the following year, and the Chang'an reign in the 10th month of that year. Technically, there was no 2nd year of the Jiushi reign. However, there are other documents and epitaphs from Turfan dated 2nd year of the Jiushi reign, and the continued use of this reign name is due to a break in communications with central China.

8 The Wenwu 1981 version of this article says tian bu 填布.

9 (WBH, fn 2) Xinjiang bowuguan kaogu dui 新疆博物館考古隊, “Tulufan Asitana – Helahezhuo gu muqun qingli jianbao” 吐魯番阿斯塔那—— 喀拉和卓古墓群清理簡報 [Brief report on the clearing of the Astana and Karakhoja cemeteries, Turfan], Wenwu 1972, No. 1, pp. 8–29; figs 21–23.

10 (WBH, fn 3) bowuguan, Xinjiang 新疆博物館, Sichou zhi lu 絲綢之路 [The Silk Road] (Beijing, 1972), pl. 62 Google Scholar.

11 Illustrated in the Wenwu 1981 version, Fig. 1, left.

12 (WBH, fn 4) See Xinjiang bowuguan kaogudui, “Tulufan Asitana mudi fajue jianbao”, p. 17, Fig. 23.

13 (WBH, fn 5) Fig. 22 in the 1981 Wenwu issue is captioned: “Jiaobu 腳布submitted by Wu De et al. of Lanxi county, from tomb TAM96, Astana”. Fig. 23 is captioned “Hemp cloth with inscription, from tomb TAM106, Astana”. This article mentions four hemp cloths with inscriptions.

14 Illustrated in the Wenwu 1981 version of this article, Fig. 1, centre. The inscription in the Wenwu 1981 version omits the character zhu 住.

15 Xinjiang Museum collection. This piece is illustrated in the Glossary in this special issue (Glossary no. C.13[2]).

16 Xinjiang Museum collection.

17 (WBH, fn 6) Xinjiang bowuguan 新疆博物館, “Tulufan xian Asitana – Helahezhuo gu muqun fajue jianbao” 吐魯番阿斯塔那— 喀拉和卓古墓群發掘簡報 [Short report on excavations at the Astana – Karakhoja cemeteries, Turfan], Wenwu (1973) No. 10, pp.7–27, especially. p. 17, Fig. 19.

18 (WBH, fn 7) Du You 杜佑 (734–812), Wang Wenjin 王文錦, Wang Yongxing 王永興 and Junwen, Liu 劉俊文 (eds), Tongdian 通典 [Encyclopaedic history of institutions] (Beijing, 1988), 6.109 Google Scholar; Xuanzong, Tang 唐玄宗 (685–762) and Linfu, Li 李林甫 (d. 753) Da Tang liudian 大唐六典 [Compendium of administrative law of the six divisions of the Tang bureaucracy] (Taipei, 1962), 3.68 Google Scholar; Pu, Wang 王溥, Tang huiyao 唐會要 [Important documents of the Tang] (Beijing, 1955), 83.1533 Google Scholar.

19 To this list can be added the tax textile (Ast.ix.2a.07) dated 706 (8th month of Shenlong year 2) found by Stein at Astana. Its inscription reads 婺州蘭溪縣瑞山鄉從善里姚群 (?)庸調布一端 神龍二年八月 日 Wuzhou Lanxi xian Ruishan xiang Congshan li Yao Qun (?) yongdiao bu yi duan Shenlong er nian ba yue ri. 1 duan of yongdiao bu from Yao Qun (?) of Congshan ward, Ruishan xiang, Lanxi county, Wuzhou. [. . .]th day of the 8th month of the Shenlong reign. This tax textile is from a woman's coffin in the same tomb as Ast.ix.2b.011, which Wang Binghua discusses (see fn 26); Stein, A., Innermost Asia: Detailed Report of Explorations in Central Asia, Kan-su and Eastern Iran (Oxford, 1928), p. 708 Google Scholar (context of find), pp. 1044–1045 (translation), pl. CXXVII.

20 (WBH, fn 8) Xiu, Ouyang 歐陽修 (1007–1072), Xin Tangshu 新唐書 [New Tang history] (Beijing, 1975), 49.1319 Google Scholar; Tang huiyao 83.1531; Da Tang liudian, 3.68.

21 Da Tang liudian, 3.68.

22 (WBH, fn 9) Da Tang liudian 20.375.

23 (WBH, fn 10) Da Tang liudian 20.375. Translator's note: some editions of this text give mu 木 rather than ben 本; some omit the character wei 衛 and the character hei 黑.

24 (WBH, fn 11) Da Tang liudian 3.68; Tang huiyao 83.1531. (There is slight variation in the wording in these two sources.)

25 (WBH, fn 12) Ast.ix.2b.011. Stein, Innermost Asia, pp.708 and 1044, pl.CXXVII; Chinese translation of Innermost Asia by Wu Xinhua 巫新華, Xiyu kaogu tuji 西域考古圖記 (Guilin, 1999), Vol. 2, Appendix, p. 1472, and illustrated in Vol. 3, p. 127. The description, by Lionel Giles, notes that this piece of “cotton” from coffin b, belonging to Fan Yanshi (male, d. 689), was dated 684–685, and had three illegible square seal impressions in red, with each seal measuring 5 cm and having a four-character inscription. A second tax textile was found in this tomb, Ast.ix.2.a 07, also from Wuzhou; see Introduction in this special issue.

26 (WBH, fn 13) Tang huiyao 83.1533.

27 See, for example, D. C. Twitchett, Financial Administration, who translates zujiao 租腳 as “supplementary surcharge on taxes for transport costs” (p. 378) and jiaozhi 腳直 as “transportation money (surcharged on taxes)” (p. 370).

28 (WBH, fn 14) Da Tang liudian 3.73.

29 (WBH, fn 15) Da Tang liudian 3.68.

30 Twitchett, Financial Administration, 1975, pp. 104–105, explains the hierarchy of financial organisation in the first half of the Tang dynasty. The central government had direct dealings with the prefectures (州 zhou) or their equivalents, the administrations (府 fu), in centres of major importance; and the protectorates (都護府 duhufu) and area commands (都督府 dudufu) in areas of military importance on the frontiers. The provinces (道dao) were convenient divisions of the empire, and played no part as intermediaries in the dealings of the central offices with the prefectures. Below prefecture level was county (縣 xian) > locality (鄉 xiang) > ward (里 li).

31 (WBH, fn 16) Li Jifu 李吉甫, Cijun, He (ed.), 賀次君 Yuanhe junxian tuzhi 元和郡縣圖志 [Maps and gazetteer of the provinces and counties in the Yuanhe reign period, 806–814](Beijing, 1983), 28.681 Google Scholar, 684.

32 (WBH, fn 17) Xu, Liu 劉昫 (887–946), Jiu Tangshu 舊唐書 [Old Tang history] (Beijing, 1975), 39.1528 Google Scholar.

33 Yuanhe junxian tuzhi, 22.561.

34 Xin Tangshu, 40.1034.

35 Xin Tangshu, 40.1032.

36 Yuanhe junxian tuzhi, 21.543.

37 Xin Tangshu, 41.1063.

38 Xin Tangshu, 41.1058.

39 It does not appear in the Jiu Tangshu, Xin tangshu, or Yuanhe junxian tuzhi.

40 Xin Tangshu, 40.1029.

41 (WBH, fn 18) See, for example, Yue, Shen 沈約 (441–513), Songshu 宋書 [History of the Song)] (Beijing, 1974), 97.2396–98Google Scholar, Zixian, Xiao 蕭子顯 (489–537), Nan Qi shu 南齊書 [History of the Southern Qi] (Beijing, 1972), 58. 10071009 Google Scholar; Yanshou, Li 李延壽 [7th century, Nanshi 南史 [History of the Southern Dynasties] (Beijing, 1975), 79. 19811982 Google Scholar; Yanshou, Li 李延壽 [7th century, Beishi 北史 [History of the Northern Dynasties] (Beijing, 1974), 95.31503152 Google Scholar.

42 (WBH, fn 19) Songshu, 97.2396.

43 (WBH, fn 20). Xin Tangshu, 42.1091.

44 The other counties included Chengdu 成都, Huayang 華陽, Guangdu 廣都 and Xinfan 新繁. See Xin Tangshu, 42.1079–1080.

45 (WBH, fn 21) Jiu Tangshu, 41.1663.

46 Xin Tangshu, 40.1046–1048.

47 (WBH, fn 22) These are in the Xinjiang Museum collection of Turfan documents.

48 Jiu Tangshu, 196.5236.

49 Tongdian, 6.110–111.

50 (WBH, fn 23) Tongdian, 6.111; Jiu Tangshu, 38.1385–1389.

51 (WBH, fn 24) See document 72TAM230:46(2), which is an incomplete fragment in the Xinjiang Museum collection of Turfan documents.

52 (WBH, fn 25). Xinjiang bowuguan 新疆博物館, Xibei daxue lishi xi kaogu zhuanye 西北大學歷史系考古隊, “1973 nian Tulufan Asitana gu muqun fajue jianbao” 1973年吐魯番阿斯塔那古墓群發掘簡報 [Short report on the excavations at the Astana cemetery, Turfan, in 1973], Wenwu (1975) No. 7, pp. 8–26, especially plates 12–13. These documents were unearthed in tomb TAM509, which was one of eight tombs with an earth mound. It also contained a ‘paper bed-cover’ (紙衾 zhiqin). Arakawa Masaharu's article in this issue translates the complete document and discusses it at length.

53 (WBH, fn 26) Yinglin, Wang (ed.) 王應麟, Yuhai 玉海 [Ocean of jade] (Nanjing and Shanghai, 1987) 138.2568–69Google Scholar.

54 Da Tang liudian, 5. 116.

55 (WBH, fn 27) D.vii.2. Stein, M.A., Ancient Khotan, Detailed Report of Archaeological Explorations in Chinese Turkestan (Oxford, 1907), pp. 526527, pl. CXVGoogle Scholar.

56 (WBH, fn 28) Tang huiyao, 88.1618; see also Da Tang liudian, 3.74.

57 (WBH, fn 29) Tongdian, 6.107; see also Jiu Tangshu, 48.2090.

58 (WBH, fn 30) Xinjiang bowuguan, Xinjiang chutu wenwu, plate 171.