Hostname: page-component-78c5997874-mlc7c Total loading time: 0 Render date: 2024-11-11T14:27:20.212Z Has data issue: false hasContentIssue false
  • English
  • Français

SINGLE- AND MULTI-PIECE MANUSCRIPTS IN EARLY IMPERIAL CHINA: ON THE BACKGROUND AND SIGNIFICANCE OF A TERMINOLOGICAL DISTINCTION

Published online by Cambridge University Press:  31 July 2018

Thies Staack
Thies Staack*
Affiliation:
Thies Staack 史達, University of Heidelberg; email: thies.staack@zo.uni-heidelberg.de.
Get access Rights & Permissions [Opens in a new window]

Abstract

Received and excavated sources from early imperial China employ various terms for pieces of bamboo or wood that served as writing support. In many cases, neither the exact meanings nor diachronic differences in usage of these terms are sufficiently clear. What kinds of concrete objects the terms actually referred to in a certain period accordingly turns out to be quite an intricate question. This article focuses on the terms du 牘 and die 牒, which not only occur most frequently in the sources, but can also be considered as a complementary pair. Investigating differences in form and function that can be gathered from the way the terms are employed in both administrative documents and legal prescriptions of the Qin and Han period (including a newly published Qin ordinance) it argues that du and die were connected to two conceptually different types of manuscripts, namely single- and multi-piece manuscripts. It shows that these two types also entailed differences in how the manuscripts were kept for storage and transport, which were likewise reflected by special terminology. Finally, it proposes that the increasing use of multi-piece manuscripts instead of single-piece ones, especially since the time of Emperor Wu of Han 漢武帝 (r. 141–87 b.c.e.), probably had both pragmatic and economic reasons, which fit well into the setting of a gradually consolidating empire with an ever-growing volume of bureaucratic record keeping.

提要

秦漢時期的傳世和出土資料，記載了不少與用作書寫載體的竹或木片相關的術語。很多時候，它們用法上的確切含義以及歷時性差異，皆未得到充分了解。究竟這些術語在特定時間內，指涉哪些具體事物，成為一個複雜的問題。本文聚焦於「牘」與「牒」，不但由於這兩個詞在資料裏最常出現，還因為它們可被視作互補的一對。爬梳行政文書和法律規範（包括一條新見的秦令）中「牘」和「牒」在形制和功能方面的差異後，本文認為「牘」和「牒」實際上連繫着兩種不同概念的寫本，即單獨簡和編冊簡。本文的分析又顯示，這兩種寫本也意味着儲藏方式的區別，其同樣反映在特殊的術語上。最後，本文提出漢代（特別從漢武帝時起）編冊簡使用之所以日益增加，可能出於實用和經濟原因，這也符合當時漢帝國日漸穩固、官僚系統簿記數量不斷增加的情況。

Keywords

bamboo and wood manuscriptscodicologyterminologyQin ordinances簡牘寫本學術語秦令
Type
Articles
Information
Early China , Volume 41 , 2018 , pp. 245 - 295
Copyright
Copyright © The Society for the Study of Early China and Cambridge University Press 2018 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

This paper emerged from my work at the Heidelberg Collaborative Research Centre 933 “Material Text Cultures” (Subproject B09 “Bamboo and Wood as Writing Materials in Early China”), which is financed by the German Research Foundation (DFG). I would like to thank Enno Giele, Ulrich Lau, Tang Junfeng 唐俊峰 and Paul Nicholas Vogt as well as the two anonymous reviewers for many valuable suggestions and comments, which helped to improve a previous draft of this work in various ways. I am also indebted to Chen Songchang 陳松長 and his editorial team for giving me the opportunity to participate in their editorial work on the Yuelu Academy Qin manuscripts, which also greatly benefitted the writing of this paper.

References

1. See Guowei, Wang 王國維, Jiandu jianshu kao jiaozhu 簡牘檢署考校注, ed. and comm. Pingsheng, Hu 胡平生 and Yuehua, Ma 馬月華 (Shanghai: Shanghai guji, 2004 [1914]), 14–27 and 37–41Google Scholar, respectively. For an earlier work see Chavannes, Édouard, “Les livres chinois avant l’invention du papier,” Journal Asiatique 10th ser., 5 (1905), 5–75Google Scholar.

2. See, for example, Junming, Li 李均明, Gudai jiandu 古代簡牘 (Beijing: Wenwu, 2003)Google Scholar. Cf. the broader term jianbo zhidu 簡帛制度 used in Xiancheng, Zhang 張顯成, Jianbo wenxian xue tonglun 簡帛文獻學通論 (Beijing: Zhonghua, 2004)Google Scholar. See also the narrower term jiance zhidu 簡冊制度 in Xiao Yunxiao 肖芸曉, “Qinghua jian jiance zhidu kaocha” 清華簡簡冊制度考察 (MA thesis, Wuhan University, 2015).

3. By “writing support” (German “Beschreibstoff”) I mean pieces of wood, bamboo, or other material intended or used to carry writing. “Shape” and “format” are understood to be sub-categories of the more general “form.” While “shape” is used to distinguish, for example, flat rectangular pieces of writing support from polygonal ones, the category “format” serves to subdivide pieces of the same basic shape according to different measurements.

4. See, for example, Li Junming, Gudai jiandu, 135–37.

5. For a recent discussion of terms for pieces of writing support see Yuqian, Pian 駢宇騫, Jianbo wenxian gangyao 簡帛文獻綱要 (Beijing: Beijing daxue, 2015), 41–62Google Scholar. The overview of different interpretations of the terms jian 簡, zha 札, and die 牒 in that book illustrates the remaining uncertainties. See Jianbo wenxian gangyao, 41–43 and 46–47.

6. See, for example, Jingzhou bowuguan, “Hubei Jingzhou Jinan Songbai Han mu fajue jianbao” 湖北荊州紀南松柏漢墓發掘簡報, Wenwu 文物 2008.4, 24–32; Jingzhou bowuguan, “Hubei Jingzhou Xiejiaqiao yi hao Han mu fajue jianbao” 湖北荊州謝家橋一號漢墓發掘簡報, Wenwu 文物 2009.4, 26–42; Changsha shi wenwu kaogu yanjiusuo, “Hunan Changsha Wuyi guangchang Dong-Han jiandu fajue jianbao” 湖南長沙五一廣場東漢簡牘發掘簡報, Wenwu 文物 2013.6, 4–26.

7. In German, the words “Tafel”/“Täfelchen” and “Leiste” or “Streifen” with similar meanings are most common. Enno Giele has recently proposed to subdivide slips/strips produced from wood and bamboo by the German words “Leiste” and “Spleiße” (English “splice”), respectively, to reflect different production techniques. See Enno Giele and Jörg Peltzer, with the assistance of Trede, Melanie, “Rollen, Blättern und (Ent)Falten,” in Materiale Textkulturen: Konzepte—Materialien—Praktiken, ed. Meier, Thomas, Ott, Michael R., and Sauer, Rebecca (Berlin: de Gruyter, 2015), 678Google Scholar.

8. For example, it has been proposed that du can be used individually but can also be tied together with other pieces. See Ling, Li 李零, Zhongguo fangshu xukao 中國方術續考 (Beijing: Dongfang, 2000), 456Google Scholar. According to Ōba Osamu 大庭脩, in early imperial China the term du simply referred to individual tablets or boards used for writing. To use du to designate comparatively wide tablets (as opposed to narrower pieces) was, according to him, a later development. This view is cited in Takeyuki, Takamura 高村武幸, “Shin Kan jidai no toku ni tsuite” 秦漢時代の牘について, Jinbun ronsō: Mie daigaku jinbun gakubu bunka gakka kenkyū kiyō 人文論叢：三重大学人文学部文化学科研究紀要 30 (2013), 57–71Google Scholar. For the view that du are both comparatively wide and used individually see Mengjia, Chen 陳夢家, Han jian zhuishu 漢簡綴述 (Beijing: Zhonghua, 1980), 314Google Scholar.

9. See, for example, Xiejiaqiao 謝家橋 Han tomb no. 1, where a list of grave goods written on 197 narrow pieces of bamboo was apparently tied together with three wider pieces of bamboo carrying a so-called “announcement to the world below” (gao di shu 告地書). See Jingzhou bowuguan, “Hubei Jingzhou Xiejiaqiao yi hao Han mu fajue jianbao,” 36, 41; cf. bowuguan, Jingzhou, ed., Jingzhou zhongyao kaogu faxian 荊州重要考古發現 (Beijing: Wenwu, 2009), 191, 194Google Scholar.

10. In Han administrative documents, jian were called either die or zha 札, depending on whether their function or their form was referred to, see further below. It is as yet unclear whether zha was used to refer to writing support already in the Qin period. The term jian was uncommon in administrative documents and in other sources often referred especially to pieces of bamboo. See, for example, Suanshu shu 算數書 70–71. For this manuscript, which was excavated from the early Han tomb no. 247 at Zhangjiashan 張家山, see Zhangjiashan ersiqi hao Han mu zhujian zhengli xiaozu 張家山二四七號漢墓竹簡整理小組, ed., Zhangjiashan Han mu zhujian [ersiqi hao mu] 張家山漢墓竹簡 [二四七號墓] (Beijing: Wenwu, 2001)Google Scholar.

11. Throughout this article a basic distinction between single- and multi-piece manuscripts is drawn. The former are manuscripts consisting of only one piece of inscribed bamboo or wood, while the latter are all manuscripts that consist of at least two pieces connected consecutively with the help of two or more binding strings. Possible cases where, for example, a label or tag may have been temporarily attached to a single-piece manuscript with a string are not counted as multi-piece manuscripts.

12. It has been proposed that shu can be seen as a specific “literary form” or more precisely “any text which claims to be a contemporaneous record of a speech of an ancient king.” See Allan, Sarah, “On Shu 書 (Documents) and the Origin of the Shang shu 尚書 (Ancient Documents) in Light of Recently Discovered Bamboo Slip Manuscripts,” Bulletin of the School of Oriental and African Studies 75.3 (2012), 557CrossRefGoogle Scholar. Although this specific meaning of shu may have existed, the generic meaning “writings” seems to have been very common, at least by the early imperial period. See, for example, the various mathematical, legal, and medical texts in the manuscripts excavated from Zhangjiashan tomb no. 247 that are designated as different kinds of shu (Zhangjiashan ersiqi hao Han mu zhujian zhengli xiaozu, Zhangjiashan Han mu zhujian [ersiqi hao mu]).

13. Liye 里耶 8-375. If not stated otherwise, the transcriptions of the manuscripts from layers 5, 6, and 8 of Liye well no. 1 follow Wei, Chen 陳偉, ed., Liye Qin jiandu jiaoshi (di yi juan) 里耶秦簡牘校釋（第一卷） (Wuhan: Wuhan daxue, 2012)Google Scholar. For the respective photographs see Hunan sheng wenwu kaogu yanjiusuo, ed., Liye Qin jian (yi) 里耶秦簡（壹） (Beijing: Wenwu, 2012)Google Scholar. For transcriptions and photographs of manuscripts from layers 7, 9, 10, 11, 12, 14, 15, 16, and 17 see Liye Qin jian bowuguan and Chutu wenxian yu zhongguo gudai wenming yanjiu xietong chuangxin zhongxin Zhongguo renmin daxue zhongxin, eds., Liye Qin jian bowuguan cang Qin jian 里耶秦簡博物館藏秦簡 (Shanghai: Zhongxi, 2016)Google Scholar. For additional manuscripts from layers 7, 9, 14, 15, and 16 see Hunan sheng wenwu kaogu yanjiusuo, Liye fajue baogao 里耶發掘報告 (Changsha: Yuelu shushe, 2006) 180–99, color plates 16–40Google Scholar.

14. Although the word juan might at times refer to silk manuscripts, at least the following passage provides quite reliable evidence that it was likewise used to refer to bamboo or wood manuscripts. Otherwise, the fact that rolls became “disarranged” or “mingled with each other” would be difficult to explain. See Xiang, Liu 劉向, Zhanguo ce 戰國策 (Shanghai: Shanghai guji, 1985), 1195Google Scholar (“Fulu: Liu Xiang shulu” 附錄：劉向書錄): 臣向言所校中戰國策書：中書餘卷錯亂相糅莒。 “Your servant [Liu] Xiang … reports on the writings of the Zhanguo ce from the inner [palace], which he checked by means of comparison: The more than [?] rolls with writings from the inner [palace] had become disarranged and were mingled with each other.” Punctuation modified, translation adapted from Friedrich, Michael, “Der editorische Bericht des Liu Hsiang zum Chan-kuo ts’e,” in Den Jadestein erlangen: Festschrift für Harro von Senger, ed. Gänßbauer, Monika (Frankfurt am Main: Lembeck, 2009), 246Google Scholar.

15. In some cases a cloth bag or bamboo basket must have contained several distinct documents, but only one jian (or jie 楬 “label”) was attached to that container, while in other cases a jian was probably directly attached to one particular document. On jian see Li Junming 李均明, “Fengjian tishu kaolüe” 封檢題署考略, Wenwu 文物 1990.10, 72–78; Osamu, Ōba 大庭脩, Han jian yanjiu 漢簡研究, trans. Shihong, Xu 徐世虹 (Guilin: Guangxi shifan daxue, 2001), 176–204Google Scholar.

16. Qinlü shiba zhong 秦律十八種 131. Throughout this article all references to the Shuihudi manuscripts—except for Qinlü shiba zhong also Falü dawen 法律答問, Feng zhen shi 封診式, and Wei li zhi dao 爲吏之道—are according to Shuihudi Qin mu zhujian zhengli xiaozu 睡虎地秦墓竹簡整理小組, ed., Shuihudi Qin mu zhujian 睡虎地秦墓竹簡 (Beijing: Wenwu, 1990)Google Scholar. Translation adapted from Hulsewé, A. F. P., Remnants of Ch’in Law: An Annotated Translation of the Ch’in Legal and Administrative Rules of the 3rd Century B.C. Discovered in Yün-meng Prefecture, Hu-pei Province, in 1975 (Leiden: Brill, 1985), 76 (A 77)Google Scholar.

17. It may be doubted that fang refers to the kind of polygonal rods (gu 觚) that were occasionally found among other Han manuscripts in the northwest of China, as originally proposed by the Shuihudi editors. See Shuihudi Qin mu zhujian zhengli xiaozu, Shuihudi Qin mu zhujian, 50–51 (transcription part).

18. Hulsewé, Remnants of Ch’in Law, 76 (A77).

19. See Yunmeng Shuihudi Qin mu bianxie zu 雲夢睡虎地秦墓編寫組, Yunmeng Shuihudi Qin mu 雲夢睡虎地秦墓 (Beijing: Wenwu, 1981), 25–26Google Scholar, plates 167–68; Xinfang, Liu 劉信芳 and Zhu, Liang 梁柱, eds., Yunmeng Longgang Qin jian 雲夢龍崗秦簡 (Beijing: Kexue, 1997), 45, plate 27Google Scholar; Sichuan sheng bowuguan and Qingchuan xian wenhuaguan, “Qingchuan xian chutu Qin geng xiu tianlü mudu: Sichuan Qingchuan xian Zhanguo mu fajue jianbao” 青川縣出土秦更修田律木牘——四川青川縣戰國墓發掘簡報, Wenwu 文物 1982.1, 11; Hubei sheng Jiangling xian wenwuju and Jingzhou diqu bowuguan, “Jiangling Yueshan Qin Han mu” 江陵岳山秦漢墓, Kaogu xuebao 考古學報 2000.4, 549. For the proposal that fang means “rectangular” in the aforementioned passage see Yates, Robin D. S., “The Qin Slips and Boards from Well no. 1, Liye, Hunan: A Brief Introduction to the Qin Qianling County Archives,” Early China 35–36 (2012–13), 314n65Google Scholar.

20. Yi li zhushu 儀禮注疏 (Shisan jing zhushu 十三經注疏 ed., 1815; rpt. Taipei: Yiwen, 2001), 24.283. Translation adapted from Steele, John, The I-Li or Book of Etiquette and Ceremonial, vol. 1 (London: Probsthain & Co., 1917), 232Google Scholar.

21. Zheng Xuan 鄭玄 glosses ce 策 in this passage as jian 𥳑 “slip(s)” of bamboo or wood (Yi li zhushu, 24.283). It seems likely that ce actually meant slips that were tied together, because ce 策 and ce 冊 “mat” are often used interchangeably. See Li, Wang 王力, Wang Li gu hanyu zidian 王力古漢語字典 (Beijing: Zhonghua, 2000), 60, 876Google Scholar; Qiyong, Feng 馮其庸 and Ansheng, Deng 鄧安生, Tongjia zi huishi 通假字彙釋 (Beijing: Beijing daxue, 2006), 35, 733Google Scholar.

22. The translation of the chapter name follows Boltz, William G., “I li,” in Early Chinese Texts: A Bibliographical Guide, ed. Loewe, Michael (Berkeley: Society for the Study of Early China, Institute of East Asian Studies, 1993), 235Google Scholar.

23. Chong, Wang 王充, Lun heng jiaoshi 論衡校釋, ed. and comm. Hui, Huang 黃暉, 4 vols. (Beijing: Zhonghua, 1990), 12.551Google Scholar.

24. See the gloss du pu 牘樸 “wood block for [the production of] du” for qian 槧 in Shen, Xu 許慎, Shuowen jiezi zhu 說文解字注, ed. and comm. Yucai, Duan 段玉裁 (Shanghai: Shanghai guji, 1981), 6.265Google Scholar.

25. Schuessler, Axel, ABC Etymological Dictionary of Old Chinese (Honolulu: University of Hawaiʻi Press, 2007), 155Google Scholar.

26. Scraping pieces of wood was one of many tasks assigned to garrison soldiers during the Han period. See Guihai, Wang 汪桂海, “Handai guanfu jiandu de jiagong, gongying” 漢代官府簡牘的加工、供應, Jianbo yanjiu 簡帛研究 2009 (2011), 144–45Google Scholar.

27. See Liye 8-1369+8-1937 as well as Falü dawen 88.

28. Shuowen jiezi zhu, 7.318.

29. This probably refers to the passage already cited above: Yi li zhushu, 24.283.

30. However, it should be noted that fang actually occurs in other meanings such as “recipe” (Liye 8-876, etc.), “surface” (Liye 8-1369+8-1937) or “side” (Liye 12-1784).

31. For die see Liye 8-5, 8-42+8-55, 8-135, 8-164+8-1475, 8-170, 8-175, 8-183+8-290+8-530, 8-225, 8-234, 8-235, 8-317, 8-369+8-726, 8-528+8-532+8-674, 8-551, 8-602+8-1717+8-1892+8-1922, 8-645 (2x), 8-651, 8-653, 8-677, 8-686+8-973, 8-768 (2x), 8-804, 8-1041+8-104, 8-1069+8-1434+8-1520, 8-1511, 8-1514 (2x), 8-1539, 8-1559, 8-1565, 8-1715, 8-2003, 8-2035, 8-2543, 9-1869, and 9-2352. For du see Liye 7-4, 8-302, 8-487+8-2004, 8-499, 8-1019, 8-1203, 8-1494, 8-1517, 8-1566, 8-1654, and 8-2146.

32. Tsuneko, Sumiya 角谷常子, “Riya Shin kan ni okeru tandoku kan ni tsuite” 里耶秦簡における単独簡について, Nara shigaku 奈良史学 30 (2012), 109Google Scholar. An important question, of course, is whether the text written on the pieces, including the terms du or die, was originally drafted on exactly these pieces or originally drafted on other pieces and copied onto the pieces of writing support we have before us. In the latter case, the direct connection between the terms du or die and pieces of writing support they refer to may have been lost. As all pieces discussed in this section and gathered in Table 1 constitute documents that were very likely addressed to the prefectural court of Qianling 遷陵—at the original location of which they were also excavated—and in some cases (e.g., Liye 8-1566) were also furnished with notes of receipt, they are tentatively considered as “original” documents that were actually sent to the Qianling court. Cf. the discussion in Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 111–15. For more in-depth studies on the problem of so-called originals and copies, see Maxim Korolkov, “Criteria for Discerning Individual Writing Habits of the Liye Scribes: Originals and Copies of the Official Documents and the Bureaucratic Politics in the Qin Empire” (paper presented at the Fifth European Association for the Study of Chinese Manuscripts Conference, Heidelberg, July 11–13, 2014); and Yitian, Xing 邢義田, “Handai jiandu gongwenshu de zhengben, fuben, caogao he qianshu wenti” 漢代簡牘公文書的正本、副本、草稿和簽署問題, Zhongyang yanjiuyuan lishi yuyan yanjiusuo jikan 中央研究院歷史語言研究所集刊 82.4 (2011), 601–78Google Scholar.

33. Xiqi, Xu 徐錫祺, Xi-Zhou (Gonghe) zhi Xi-Han lipu 西周（共和）至西漢曆譜, 2 vols. (Beijing: Beijing kexue jishu, 1997), 1250Google Scholar.

34. The formulation shu shu 疏書 refers to writing something in form of a list or in separate entries. See Shuihudi Qin mu zhujian zhengli xiaozu, Shuihudi Qin mu zhujian, 163n3 (transcription part). This means a special kind of layout in which every entry of a list of items is started in a new column. It probably implied that the space on a certain piece of writing support was separated into several registers. On principle this separation into registers could be used on both wider tablets with space for several columns of writing (as in the present example) as well as on narrow slips with only one column, see, for example, the Wei li zhi dao from Shuihudi. However, it is likely that the expression shu shu mainly referred to a particular layout for single-piece manuscripts. At least with regard to the Liye finds, all occurrences of shu shu are on single-piece manuscripts. See Liye 8-487+8-2004, 8-686+8-973, 8-1069+8-1434+8-1520, 8-1517, and 8-1566. For the view that shu shu is not confined to writing support of a particular shape or format, see Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 122.

35. The text in the cover letter gives shu shu zuo tu ri bu yi die 疏書作徒日薄（簿）一牒 “we recorded in separate entries the daily register of the convicts assigned to work [in our office] on one die” instead of the expected shu shu zuo tu ri bu die bei 疏書作徒日薄（簿）牒北（背） “we recorded in separate entries the daily register of the convicts assigned to work [in our office] on the back of this die.” However, it is clear from the text on the other side of Liye 8-1069+8-1434+8-1520—a list introduced by ku Wu zuo tu bu 庫武作徒薄（簿） “register of the convicts assigned to work [in the armory, drafted] by Wu, [head] of the armory”—that the “one die” mentioned in the cover letter must refer to the very same piece of wood on which it is written.

36. Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 109, 122.

37. The measurements provided in Table 1 are based on hand measurement of the photographs in Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian (yi). In most cases the pieces did not have exactly the same width over the whole length. The table gives the maximum width. Cf. the slightly lower measures for Liye 8-1517 (3.6 cm), 8-1566 (1.7 cm) and 8-1069+8-1434+8-1520 (2.8 cm) provided in Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 109–10. The years refer to the reign of King Zheng of Qin 秦王政, who became king of Qin in 246 b.c.e. and assumed the title First Emperor of Qin 秦始皇帝 in 221 b.c.e.

38. Feng zhen shi 91–92. Translation adapted from Hulsewé, Remnants of Ch’in Law, 206 (E 24).

39. The character 諜 was occasionally used to write die 牒. Several examples of this can be found in received literature. See Feng Qiyong and Deng Ansheng, Tongjia zi huishi, 956; Hui, Wang 王輝, Gu wenzi tongjia zidian 古文字通假字典 (Beijing: Zhonghua, 2008), 773Google Scholar.

40. An example of an actual administrative document with a textual structure similar to this Feng zhen shi model is Liye 8-439+8-519+8-537. There we also find a protocol (yuanshu 爰書) that begins with a report (gao 告), albeit without the details on the reporting persons on the back of the tablet. Only a description of the reported person and his personal belongings is recorded on the same side as the protocol.

41. The formulation shang jie die 上解牒 on Liye 8-804, which could be understood as “to submit those die [the binding strings of] which have become loose,” might be seen as a case in which several die were originally tied together with binding strings.

42. Heng, Gao 高恒, Qin Han jiandu zhong fazhi wenshu jikao 秦漢簡牘中法制文書輯考 (Beijing: Shehui kexue, 2008), 365n8Google Scholar. See also further below.

43. This is definitely the case for Liye 8-183+8-290+8-530 (with traces on fragments 8-183 and 8-530) and possibly also for Liye 8-551 and 8-1559. See the photographs in Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian (yi). Overall, traces of binding strings are rarely observable on the available Liye photographs, which in large part are infrared scans better suitable for identification of the carbon ink writing. Even the photographs of pieces Liye 8-755 to 8-759, which according to the editors were once tied together, do not exhibit any obvious traces of binding strings. See Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian (yi), 49 (transcription part). With this in mind, it seems likely that the circumstances in well no. 1 may have promoted the complete disintegration of binding strings to the extent that barely any traces of them are left.

44. Liye 8-164+8-1475 with a width of 4.2 cm represents the widest example. Liye 8-170, 8-768, 8-1514, 8-1559, and 8-1565 are each narrower but likewise have three columns of writing. Of the overall eighteen pieces in Appendix A on which the text probably refers to an attachment, twelve carry one or two columns, while only six carry three columns of writing. Note that three of four single-piece manuscripts on which die refers to the same piece carry four columns of writing, at least on one of their sides (see Table 1).

45. The cover letter pieces in Appendix A furthermore usually do not carry writing on their verso apart from brief notes of receipt or the name of the scribe who drafted the respective document. Cf. similar notes on confirmed multi-piece manuscripts, like Xie shou 歇手 “handled by Xie” on the verso of Liye 8-755.

46. Wei yu deng zhuang si zhong 爲獄等狀四種 148. All references to this manuscript are according to Hanmin, Zhu 朱漢民 and Songchang, Chen 陳松長, eds., Yuelu shuyuan cang Qin jian (san) 嶽麓書院藏秦簡（叁） (Shanghai: Shanghai cishu, 2013)Google Scholar. Translation adapted from Lau, Ulrich and Staack, Thies, Legal Practice in the Formative Stages of the Chinese Empire: An Annotated Translation of the Exemplary Qin Criminal Cases from the Yuelu Academy Collection (Leiden: Brill, 2016), 225CrossRefGoogle Scholar.

47. Wei yu deng zhuang si zhong 168–69. Translation adapted from Lau and Staack, Legal Practice, 244.

48. Another possibility is that the original case files were written on slips of a different length, which could therefore have accommodated a higher or lower number of characters, leading to a different number of slips necessary to make a complete copy.

49. A count of the number of slips on which the case records are written gives six slips (Wei yu deng zhuang si zhong 142–47) for case II.9; the record for case II.10 is written on seventeen slips (Wei yu deng zhuang si zhong 150–66) to which one assumed missing slip would have to be added. The count includes only the slips from the beginning of the respective case record until the passage where the punishment for the offender is noted (Yi lun zhe X 已論磔X。 “Judgement has already been passed to quarter X.”), as this is probably where the original case file ended. What follows is a letter of recommendation attached to a copy of the original case file. It should be noted that even in the copies of the original case files there are a few short narrative passages that are likely to be later interpolations. On narrative elements and literary embellishments in the Zouyan shu 奏讞書 criminal case records excavated from Zhangjiashan tomb no. 247 see Barbieri-Low, Anthony J. and Yates, Robin D. S., Law, State, and Society in Early Imperial China: A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no. 247, 2 vols. (Leiden: Brill, 2015), 32–33, 98–109Google Scholar.

50. Yuelu Lüling 嶽麓律令 138–45. See Chen Songchang 陳松長, ed., Yuelu shuyuan cang Qin jian (wu) 嶽麓書院藏秦簡（伍） (Shanghai: Shanghai cishu, 2017). If not stated otherwise, in the following, Yuelu Lüling refers to the slips published as part of that volume, following the numbers of the edition. For yet unpublished Yuelu Lüling slips, the original item numbers are given.

51. On the title of this ordinance see the discussion in Chen Songchang 陳松長, “Yuelu Qin jian zhong de ji ge lingming xiaoshi” 嶽麓秦簡中的幾個令名小識, Wenwu 文物 2016.12, 61–64. Chen considers zu 卒 to have the meaning gong 共 “to supply, distribute” in this context.

52. See Yuelu Lüling 0378, 0581, J21, 0682, 0710, 0362, 0330, and 0476 (unpublished).

53. That is, with the exception of titles, which were often written on the verso, or other brief notes such as scribal “signatures” in administrative documents. However, the use of both sides for writing was even in these cases normally restricted to the very first or last piece(s) of a multi-piece manuscript. On one- vs. two-sided writing see also n. 10 in Momiyama Akira 籾山明, “Amu koto to tabaneru koto: Senryō ken ni okeru bunsho hokan to gyōsei jitsumu (2)” 編むことと束ねること——遷陵縣における文書保管と行政實務 (2), January 13, 2014 (http://www.aa.tufs.ac.jp/users/Ejina/note/note06%28Momiyama%29.html), accessed November 21, 2017.

54. One may hypothesize that ban refers to wood pieces with a comparatively large format and/or thickness such as the boards excavated from Qin tomb no. 1 at Fangmatan 放馬灘. The tomb contained four wooden boards with maps drawn on them. One of the maps consisted of three connected boards, the other three of one board each. The maps measure 26.7×18.1×1.1 cm, 26.6×15×1.1 cm, 26.5×18.1×2.1 cm, and 26.8×16.9×1 cm, respectively. See Gansu sheng wenwu kaogu yanjiusuo, ed., Tianshui Fangmatan Qin jian 天水放馬灘秦簡 (Beijing: Zhonghua, 2009), 73–76, 119–20Google Scholar, as well as color plates 3–7. Cf. the comparatively large piece Liye 8-461, measuring roughly 28x13 cm. It is noteworthy that all four Fangmatan maps as well as Liye 8-461 have a similar format, which may hint towards a certain standard for production. However, there are other possibilities. For example, if the hypothesis that the term fang refers to plane or levelled pieces (see above) is correct, this would imply that the respective pieces at least ideally had only right angles and six plane faces. A ban then could be anything similar that does not fulfil the criteria “right angles” and/or “plane faces,” for example, because of rounded faces on pieces of wood cut from the outermost part of a tree trunk. On the other hand, note Zheng Xuan’s Yi li commentary, which states that a fang is a ban with five, seven, or nine columns of writing (Yi li zhushu, 39.463).

55. Qinlü shiba zhong 131. Cf. the translation in Hulsewé, Remnants of Ch’in Law, 76 (A 77): “In prefectures where there are many sedges in the hills, write on sedge tied together.”

56. See, for example, Shi ji 史記 (Beijing: Zhonghua, 1959), 20.1059, 47.1937; Han shu 漢書 (Beijing, Zhonghua, 1962), 51.2367. Judging from his translation of the passage, Hulsewé appears to understand chan 纏 in the same way as bian 編.

57. Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 123. The Shuihudi editors translated chan as chan shu 纏束 “to tie together in a bundle.” See Shuihudi Qin mu zhujian zhengli xiaozu, Shuihudi Qin mu zhujian, 51 (transcription part). Enno Giele brought to my attention that in principle chan would not necessarily have to be restricted to single-piece manuscripts. It is also possible that rolled-up multi-piece manuscripts were tied together with a string that was wrapped around the outside. Accordingly, both chan and bian might be expected to occur in the context of multi-piece manuscripts. Note that a Qin ordinance (Yuelu Lüling 103–4) also refers to strings that are wrapped around a document and to which sealing clay is attached as chan. Still, it seems that only chan could have been used, whenever several single-piece manuscripts were tied together.

58. The latter is proposed in Hunan sheng wenwu kaogu yanjiusuo, “Liye yi hao jing de fengjian he shu” 里耶一號井的封檢和束, Hunan kaogu jikan 湖南考古輯刊 8 (2009), 68–69Google Scholar. Cf. Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian (yi), preface, 2.

59. Momiyama Akira, “‘Taba’ to hyōdai kan no kankei ni tsuite: Senryō ken ni okeru bunsho hokan to gyōsei jitsumu (1)” 「束」と表題簡の關係について——遷陵縣における文書保管と行政實務 (1), January 13, 2014 (http://www.aa.tufs.ac.jp/users/Ejina/note/note05%28Momiyama%29.html), accessed November 21, 2017. The pieces cited by Momiyama are Liye 8-204+8-1842, 8-306+8-282, 8-1242, 8-1556, 8-1728, 11-14, and 16-38. The top of Liye 8-204+8-1842 is fragmented and therefore no judgement on the existence of a blackened top is possible. For the reconstruction of Liye 8-306+8-282 see He Youzu 何有祖, “Liye Qin jiandu zhuihe (liu)” 里耶秦簡牘綴合（六）, June 4, 2012 (http://www.bsm.org.cn/show_article.php?id=1708), accessed November 21, 2017. For Liye 11-14 and 16-38 see Hunan sheng wenwu kaogu yanjiusuo, “Liye yi hao jing de fengjian he shu,” 68.

60. Momiyama, “Amu koto to tabaneru koto.”

61. As a notable coincidence, the modern word shu 束 is actually used by the archaeologists to describe the bundles. At the same time they note that “some of the tablets carry writing on both sides” but also that “most of them show traces of two binding strings.” See Hubei sheng wenwu kaogu yanjiusuo and Yunmeng xian bowuguan, “Hubei Yunmeng Shuihudi M77 fajue jianbao” 湖北雲夢睡虎地M77發掘簡報, Jiang Han kaogu 江漢考古 2008.4, 35. It is as yet unclear exactly which tablets show traces of binding strings. However, it seems possible that the pieces with writing on both sides were not tied together as multi-piece manuscripts but stored in a form that at least in Qin times may have been called “bundles” (shu), while the others might have been part of multi-piece manuscripts. One should bear in mind that even in bundles the outermost two pieces could also exhibit traces of binding strings that may be difficult to distinguish from those used in multi-piece manuscripts. It is also noteworthy that in other cases the existence of traces of binding strings, for example, on both sides of Liye 16-5 (see the photos in Liye Qin jian bowuguan et al., Liye Qin jian bowuguan cang Qin jian, 68, 142), points to the fact that even comparatively wide pieces with writing on both sides were occasionally tied together with other pieces consecutively in a multi-piece manuscript. See also the discussion on Liye 8-159, etc. in Momiyama, “Amu koto to tabaneru koto.” However, considering the number of examples for this practice, it seems to have been an exception rather than the rule.

62. Jingzhou bowuguan, “Hubei Jingzhou Jinan Songbai Han mu fajue jianbao,” 29. Note that six tablets without writing were apparently used as covers on the top and bottom of certain bundles. Another example of stacked single-piece manuscripts may be found in Yinwan 尹灣 tomb no. 6. At least the archaeological drawing suggests that the overall twenty-three wood tablets in the tomb were found in two stacks. See bowuguan, Lianyungang shi, Donghai xian bowuguan, Zhongguo shehui kexue yuan jianbo yanjiu zhongxin, and Zhongguo wenwu yanjiusuo, eds., Yinwan Han mu jiandu 尹灣漢墓簡牘 (Beijing: Zhonghua, 1997), 172, nos. 17, 18Google Scholar. I thank Enno Giele for pointing out this example.

63. These are Liye 8-154 to 8-159, 9-1 to 9-12, and 16-5 to 16-7. Cited according to Liye Qin jian bowuguan et al., Liye Qin jian bowuguan cang Qin jian, 15. Liye 8-154 to 8-159 are described as having stuck together.

64. Yitian, Xing 邢義田, “Hunan Longshan Liye J1(8)157 he J1(9)1-12 hao Qin du de wenshu goucheng, biji he yuandang cunfang xingshi” 湖南龍山里耶J1(8)157和J1(9)1-12號秦牘的文書構成、筆跡和原檔存放形式, Jianbo 簡帛 1 (2006), 275–96Google Scholar; Momiyama, “Amu koto to tabaneru koto.” For a neat presentation of the twelve documents and the mirror-inverted imprints see Liye Qin jian bowuguan et al., Liye Qin jian bowuguan cang Qin jian, 12–13.

65. Momiyama, “Amu koto to tabaneru koto.”

66. Shuowen jiezi zhu, 13.647.

67. Wei yu deng zhuang si zhong 240, 243. Cf. the translation in Lau and Staack, Legal Practice, 298–99, 302–3.

68. See the formulation shu quan shu 鼠券束 (Liye 8-1242); cf. the explanation in Chen Wei, Liye Qin jiandu jiaoshi (di yi juan), 298. In contrast to the interpretation of Chen Wei and his team, who suspect shu quan 鼠券 to be “tallies regarding rats/mice,” shu 鼠 more likely means “to give, bestow” here. Before the character 予 became the standard way of writing “to give, bestow,” the form 鼠 was frequently used. See Chen Kanli 陳侃理, “Liye Qin fang yu ‘shu tong wenzi’” 里耶秦方與 “書同文字”, Wenwu 文物 2014.9, 77–79. For an actual example of a quan concerning a bestowal see Liye 10-1157. Furthermore, Liye 8-1554 mentions a tally in connection with a bestowal of slaves, grain, money, etc. (see n. 73 below). Hence, although the existence of “rat/mouse tallies” can of course not be ruled out, it seems more suitable to understand shu quan shu as “bundle of bestowal tallies.”

69. Yuelu Lüling 112–21 (with modifications).

70. From the coordination with two other types of submission that are directed towards a superior (qing 請 and zou 奏) and the use of the verb shang 上, it can be gathered that dui 對 must refer to answers to earlier enquiries, decisions, etc. by a superior authority. The exact difference between dui, qing, and zou is as yet unclear. If zou must be interpreted as a general term for “submission” (including both dui and qing), this would mean that a twofold distinction between dui zou “answer submission” and qing zou “request submission” is made in the present passage, instead of a threefold one. During the Qin period, a zou 奏 “memorial” was not yet necessarily directed to the emperor as in later times. For different meanings of zou in the Qin period see An, Tao 陶安 (Arnd Helmut Hafner), “Wei yu deng zhuang si zhong biaoti jian ‘zou’ zi zi jie dingzheng: jian lun Zhangjiashan Han jian Zouyan shu timing wenti” 《爲獄等狀四種》標題簡“奏”字字解訂正——兼論張家山漢簡《奏讞書》題名問題, Zhongguo gudai falü wenxian yanjiu 中國古代法律文獻研究 8 (2014), 22–48Google Scholar.

71. Note the use of bian 編, not chan 纏 or shu 束.

72. For lian shu 連屬 “line up continuously; form an unbroken line” see also Shi ji, 12.463 (cf. Nienhauser, William H. Jr., ed., The Grand Scribe’s Records, Volume II: The Basic Annals of Han China [Bloomington: Indiana University Press, 2002], 232)Google Scholar: 使者存問所給，連屬於道。 “[T]he envoys sent to present greetings and inquire after his needs formed a continuous line on the road.” The exact difference between tong bian 同編 “tie together” and lian shu “line up continuously” is as yet unclear. Could the first expression refer to a multiple-text manuscript (MTM) and the latter to a composite manuscript? (On these concepts see Friedrich, Michael and Schwarke, Cosima, eds., One-Volume Libraries: Composite and Multiple-Text Manuscripts [Berlin: de Gruyter, 2016]CrossRefGoogle Scholar, esp. the introduction.) Another possibility could be that the terms distinguish an “unsorted” from a “sorted” way (e.g., with regard to chronology) of tying manuscripts concerned with different official matters together.

73. It seems that a request connected with one and the same official matter (shi 事) could involve several items (wu 物). An actual example of a list of items can be found on Liye 8-1554r (punctuation modified): 丗五年七月戊子朔己酉，都鄉守沈爰書：高里士五（伍）廣自言：「謁以大奴良、完、小奴嚋、饒、大婢闌、願、多、□、禾稼、衣器、錢六萬，盡以予子大女子陽里胡。」凡十一物，同券齒。 “35th year, day jiyou of the seventh month with the first day wuzi, protocol of Shen, incumbent [Overseer] of the metropolitan district: Commoner Guang from Gao quarter stated on his own behalf: ‘I hereby request that the adult male slaves Liang and Wan, the minor male slaves Chou and Rao, the adult female slaves Lan, Yuan, Duo, and X, as well as grain, clothing, and vessels, and money to the amount of 60,000 cash are completely bestowed upon my daughter, the adult woman Hu from the Yang quarter.’ Overall eleven items, in accordance with the notches on [the side of] this tally.” As can be seen, wu can include different categories of things or even persons (e.g., slaves). Therefore, the abstract translation “item” appears most suitable.

74. The character 勶 is used in the Shuihudi Qin manuscripts to write the words che 撤 “to clear away, remove” (Qinlü shiba zhong 10) and che 徹 “to penetrate (e.g., a wall)” (Feng zhen shi 74, 76) or “to achieve, reach (intentions, goals, etc.)” (Wei li zhi dao 42.2–43.2, 48.2–49.2), but neither of these meanings appears to fit the present context very well. However, che 徹 obviously refers to different forms of “going all the way,” both in a concrete and in a more abstract sense. Therefore, a meaning such as “to be exhaustive/thorough” or “completely, thoroughly, exhaustively” may have been intended with regard to the drafting of official documents. As che must be a transitive verb in the present context, one would have to add another (implied) verb such as “to describe” or “to render” for the translation. Cf. Yuelu Lüling 252: 及諸作官府者，皆日勶（徹）溥（簿）之，上其廷。廷日校案次編，月盡爲冣（最），固臧（藏），令可案殹。 “as well as those who are working at government offices, in every case thoroughly record [these persons] in a register every day and submit [the records] to the [prefectural?] court. The court daily checks and ties [the submitted records] together [with the earlier ones] according to the [chronological?] sequence. At the end of the month [the court officials] compile a summary, and store [both records and summary] securely, to enable [later] consultation.” The Yuelu Academy editors propose that che in the present passage has the meaning “to spread/set out; enumerate, list,” citing a passage from the Fang yan 方言. See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 153n31. With regard to the fact that the use of che in that meaning would reflect a dialect spoken in the far east of the Qin Empire, which had only been conquered in 221 b.c.e., this interpretation does not seem likely, although the meaning “to list” would indeed fit the context.

75. Or “to make [the requested items] easy to distinguish/recognize.”

76. This is probably meant to stress that separate multi-piece manuscripts (bian 編) are not to be tied together in any way, even if they are concerned with the same official matter. The Yuelu Academy editors do not end the sentence after san 散 but rather read san and the following qu 取 continuously. See also the discussion of Yuelu Lüling 105 in n. 77 below.

77. Ji ci 急辤（辭）, literally “urgent words,” occasionally means abbreviated or concise wording in received literature. See, for example, Guliang zhuan zhushu 穀梁傳注疏 (Shisan jing zhushu 十三經注疏 ed., 1815; rpt. Taipei: Yiwen, 2001), 13.132. In the present context, it probably refers to some kind of brief description or key word, which may have been noted on each separate part of a submission if it consisted of several multi-piece manuscripts. Cf. Yuelu Lüling 105 (punctuation modified): 令曰：上事散書取急用者，上勿謂刺。 “The ordinances state: [The labels/tags] used [to note] selected key words whenever [one] official matter is submitted in scattered documents shall not be referred to as ‘cards’ (ci 刺) by the superiors/in the submission.” On the various meanings of ci 刺 in Han administrative documents see Li Junming 李均明, “Jiandu wenshu ‘ci’ kaoshu” 簡牘文書“刺”考述, Wenwu 文物 1992.9, 55–59. Qu ji 取急 in this closely related ordinance probably refers to the same thing as qu qi ji ci 取其急辤（辭） in the passage under discussion. Therefore, it may not be suitable to understand qu ji as “to take leave due to urgent matters (such as a funeral of a relative),” a meaning in which the formulation occurs in Han administrative documents. See, for example, Juyan 居延 52.57 or Juyan xinjian 居延新簡 E.P.T48:138. For transcriptions of the “old” Juyan manuscripts (Juyan) see Guihua, Xie 謝桂華, Junming, Li 李均明, and Guozhao, Zhu 朱國炤, eds., Juyan Han jian shiwen hejiao 居延漢簡釋文合校, 2 vols. (Beijing: Wenwu, 1987)Google Scholar; for the “new” Juyan manuscripts (Juyan xinjian) see Gansu sheng wenwu kaogu yanjiusuo, Gansu sheng bowuguan, Wenhua bu gu wenxian yanjiushi, and Zhongguo shehui kexue yuan lishi yanjiusuo, eds., Juyan xinjian: Jiaqu houguan yu di si sui 居延新簡——甲渠候官與第四燧 (Beijing: Wenwu, 1990)Google Scholar. This kind of leave appears to have been called gao gui 告歸 (Qinlü shiba zhong 46) or simply gui 歸 (Yuelu Lüling 285–87) during Qin times.

78. Cf. Han shu, 56.2514–15: 前所上對 … 辭不別白，指不分明 。 “In the reply earlier submitted [to You, Majesty, …] my words were not distinct and evident, my allusions not definite and clear.”

79. Instead of interpreting bi 比 as “precedent,” the Yuelu Academy editors read bi bian 比編 as “to let documents be close to each other by tying them consecutively.” See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 153n39.

80. Cf. the translation of this sentence in Lau and Staack, Legal Practice, 47n334.

81. The fragmented character following zhang 章 has not yet been identified. It is clear, however, that the unidentified character must stand for a transitive verb. With regard to the beginning of the sentence the meaning of this verb in all likelihood is “to separate” or “to mark,” for example, with the help of spaces or brush-written marks, but certainly a layout feature. Both spaces as well as large or small round marks were used to divide sections of criminal case records in the Zouyan shu from Zhangjiashan tomb no. 247 as well as in the Yuelu Academy Wei yu deng zhuang si zhong. See also n. 91 below.

82. Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 153n42.

83. It is unclear which word the character 繠 stands for. The Yuelu Academy editors interpret it as ju long 聚攏 “to gather together, accumulate.” See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 153n43. It is possible that the character is a graphic variant of 絫 used to write the word lei 纍/累 “to accumulate.” Cf. the character 羸 apparently used to write the same word in Wei yu deng zhuang si zhong 95, 102, 106 (cf. Lau and Staack, Legal Practice, 180–82, 184): 羸（纍/累）論 “pass judgement based on the accumulation [of sentences].” See also the formulation rui bian zhi 繠編之 on Yuelu Lüling 118 below.

84. On the opposition between bie bao 別報 and rui que 繠卻 see Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 153n44.

85. There is a space without writing in the lower third of slip 115, i.e., between 牘廣一寸九分寸八 and the following 四行者, which begins on slip 116. However, judging from the wording, the text at the beginning of slip 116 without doubt directly connects to the last part of text on slip 115. The Yuelu Academy editors have observed traces of scraping (Chen Songchang, Yuelu shuyuan cang Qin jian [wu], 153n45), so it seems possible that a superfluous or redundant part of text on slip 115 was erased, instead of completely replacing the slip. The reason for this may have been that the mistake was only discovered after a replacement had become too inefficient or impractical, for example, because writing had already been applied to the following slips as well.

86. According to the Liye excavation report, most of the more than 37,000 pieces of wood from well no. 1 measure between 1.4 and 5 cm in width. See Hunan sheng wenwu kaogu yanjiusuo, Liye fajue baogao, 179. A future investigation will have to show in how far the Liye pieces accord with the column-to-width ratios stipulated in this Qin ordinance.

87. The word huan 讙 “shout; joyous” would not make any sense in the present context. However, the form of the character closely resembles the character with which the word hu 護 is written in another Qin ordinance from the Yuelu Academy collection. See Yuelu Lüling 1449 (unpublished). The only difference is an additional 又 component at the bottom of the latter. The form 讙 then might be a graphic variant or a scribal mistake. Moreover, hu 護 in that other ordinance is likewise preceded by tiao 調. The formulation tiao hu 調護 occurs in received literature in the meaning “to take care of/protect somebody” (Shi ji, 55.2047, 122.3139), which could also fit the two passages in the Yuelu Academy ordinances, if tiao hu is understood in the sense “to take care of something” instead of referring to a person. The Yuelu Academy editors interpret tiao hu as “to handle appropriately.” See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 154n46.

88. The translation of hao fu 好浮 largely depends on how fu is understood. From a more general interpretation of its usual meaning “to float” as “to move on a surface” one might arrive at the tentative translation “movement of the brush on the surface of the writing support,” which could in the present context refer to a pleasant (hao 好) appearance of the writing. The Yuelu Academy editors consider the possibility of reading the character as fu 桴 with the meaning “to tie together,” which they probably derived from an interpretation of its basic meaning “raft, small boat,” as “tied together pieces of wood.” See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 154n47. However, considering the fact that there are two other common words for “tying together” that are used in the context of manuscripts (i.e., bian 編 and chan 纏), as well as the fact that the preceding and following sections of the text discuss layout features, this interpretation seems less than convincing.

89. Following prescriptions for the number of columns in relation to the width of the du, the text here apparently contains further details on the way in which writing should be applied. Depending on whether the object pronoun zhi 之 in shu zhi 書之 refers to du 牘 or hang 行 one may expect further details on the layout in general or the columns as such. The proposed translation, which understands the described way of writing as “diligently taking care of a pleasant appearance of the characters,” is highly tentative. In general, scribes were probably expected not to write sloppily, for the sake of readability, but at the same time to write adequately swiftly, so as not to waste time or get caught up in calligraphic details. It seems also possible that the passage in question refers to spacing—between the characters of one column and/or between columns—or the alignment/orientation of the columns (exactly vertical, not slanted).

90. From the preceding sentences it is possible to calculate the maximum number of characters that was intended to be written on (one side of) a du with the specified width (or three, four, or five columns, respectively) and length. Du with a length of one foot would carry between 66 (with 3 columns and a width of c. 3.5 cm) and 110 characters (with 5 columns and a width of c. 4.4 cm); du with a length of one foot and two inches would carry between 78 and 130 characters, respectively. This roughly accords with the maximum number of 100 characters per fang stipulated in the Yi li (see above). Following these prescriptions, the maximum space per character was roughly 1 cm². The maximum height of an individual character varies slightly between 1.05 and 1.07 cm for du of 23 and 27.5 cm length. With regard to width (per character), the variance is more substantial with 1.16, 0.96, and 0.87 cm for du carrying 3, 4, and 5 columns, respectively.

91. This character is only partly visible. It appears between zhang 章 and zhi 之 in exactly the same way as the fragmented character on slip 114 (see n. 81 above). Although it is certainly possible from the context that the two characters stood for the same word with a meaning such as “to separate” or “to mark,” the character forms are quite different and the identification of each of the two character forms is problematic. The Yuelu Academy editors tentatively identify it as 次. See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 154n48.

92. The right part of this character looks remarkably similar to the component 坴 in forms of the character 陸 found in pottery, seal, and bronze inscriptions. See Linyi, He 何琳儀, Zhanguo guwen zidian: Zhanguo wenzi shengxi 戰國古文字典——戰國文字聲系, 2 vols. (Beijing: Zhonghua, 1998), 225–26Google Scholar. Furthermore, there is a Chu 楚-form of 陸 (occurring as part of the place name Anlu 安陸), which resembles the whole character including the 月/肉 component on the left. See Rensheng, Teng 滕壬生, Chuxi jianbo wenzibian (zengding ben) 楚系簡帛文字編（增訂本） (Wuhan: Hubei jiaoyu, 2008), 1191Google Scholar; cf. He Linyi, Zhanguo guwen zidian, 225. If the character can indeed be identified as 陸, the respective word lu 陸 with the basic meaning “spot/piece of land” possibly refers to applying a spot or blot (German “Klecks”) of ink, i.e., some kind of mark, for example, dots or hooks. Another possibility is that the word refers to the ancient equivalent of a “paragraph break.” The Yuelu Academy editors identify the character as 腏 and consider it to stand for either zhui 綴 “mark” or chuo 輟 “stop, suspend, interrupt.” See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 154n49.

93. It may be possible to assign the following wei gu 爲故 to the end of this sentence instead of the beginning of the next, as done by the Yuelu Academy editors. See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 106. This would lead to the subclause yi bie yi zhi wei gu 以別昜（易）智（知）爲故 “for the reason of distinguishing [separate parts of the text] and making them easy to understand/recognize.” However, the meaning of the sentence would not be significantly different without the final wei gu. Furthermore, yi zhi 昜（易）智（知） occurs at the very end of a sentence in two other passages on slips 112 and 113 as well, which is why wei gu is considered to be the beginning of the next sentence.

94. Jie 阶 may either be understood as jie 界 “to draw/set a boundary; demarcate” (Chen Songchang, Yuelu shuyuan cang Qin jian [wu], 107) or jie 介 “to insert.” The choice depends on how the following fang 方 is interpreted, and the resulting translation would be either “to demarcate the sections” or “to insert a ‘section-piece.’” See also n. 95 below.

95. From several Han manuscripts containing lists of grave goods, we know that different categories of grave goods are separated in these lists by subheadings. A subheading is usually written on a separate piece following after (i.e., to the left of) the pieces describing grave goods of the category to which the subheading refers. Frequently found formulations for subheadings closely match the formulation ci yi you ruo zuo ruogan die 此以右若左若干牒 “to the right or left of this [piece] so and so many die” in the present Qin ordinance. See, for example, you fang si die zhu qi 右方四牒竹器 “to the right: four die with bamboo utensils/containers” or you fang qi hua mu qi ba die 右方漆畫木器八牒 “to the right: lacquer-painted wood utensils/containers, eight die” on Mawangdui yi hao mu zhujian qiance 馬王堆一號墓竹簡遣策 283 and 219, respectively. All references to the lists of grave goods from Mawangdui are according to Xigui, Qiu 裘錫圭, Hunan sheng bowuguan, and Fudan daxue chutu wenxian yu gu wenzi yanjiu zhongxin, eds., Changsha Mawangdui Han mu jianbo jicheng 長沙馬王堆漢墓簡帛集成 (Beijing: Zhonghua, 2014), vols. 2Google Scholar (photographs) and 6 (transcriptions). On first sight, one might tend to understand fang in the cited subheadings as “direction.” However, in the light of the present Qin ordinance it appears more likely that fang refers to a part of text between two subheadings. This also makes more sense than merely “to the right” because in the cited examples you fang does definitely not mean all the die that may be situated to the right of the piece with a certain subheading (including die with grave goods of other categories). However, if fang means “section of a text,” what is the difference to zhang 章? If the preceding character is read as jie 介 “to insert,” fang could also refer to the piece carrying the subheading itself. The lists of grave goods found in Mawangdui tombs no. 1 and 3 show that wider pieces with several columns of writing were also used for the subheadings. Apparently, the use of either a narrow or a wide piece depended on the number of characters to be written. For example, the list from tomb no. 3 contains both narrow and wide pieces (see, for example, Mawangdui san hao mu zhujian qiance 馬王堆三號墓竹簡遣策 396/154, 397/168). One may speculate that pieces with subheadings originally were comparatively wide, which could explain the use of the term fang (cf. the Yuelu Academy editors’ interpretation as ban 板 “board,” see Chen Songchang, Yuelu shuyuan cang Qin jian [wu], 154n52). As has been argued above, this term likely referred to pieces of writing support that were used as single-piece manuscripts, which were usually wider than the pieces used for multi-piece manuscripts. It is possible that du replaced fang at some point, because fang came to acquire a new meaning in the context of manuscript production. The meaning of fang may have shifted further between Qin and Han times from “piece with a subheading” to “section of text demarcated by a piece with a subheading.” This would at least explain the slightly different use of fang in very similar contexts in Qin administrative documents and Han lists of grave goods.

96. Although it is unclear exactly which word 江 is supposed to write in the present context, it is possible to make an educated guess on the basis of extant Qin and Han manuscripts. The most likely possibility is that the respective word refers to the process of filling the top of a piece of writing support with ink, resulting in a black rectangle or square or in a thin black line that might be seen as vaguely resembling a jiang 江, or “river.” In any case, this feature probably served as an orientation aid, and there are many examples of it on the previously mentioned pieces with subheadings in Han lists of grave goods. See, for example, Fenghuangshan ba hao mu qiance 鳳凰山八號墓遣策 88, 115, 125, 170 (Hubei sheng wenwu kaogu yanjiusuo, ed., Jiangling Fenghuangshan Xi-Han jiandu 江陵鳳凰山西漢簡牘 [Beijing: Zhonghua, 2012])Google Scholar; Mawangdui yi hao mu zhujian qiance 10, 18, 22, etc. For examples of this layout feature in the Liye documents see Momiyama, “‘Taba’ to hyōdai kan no kankei ni tsuite”; Zhang Jin 張今, “Liye Qin jian zhong de jie” 里耶秦簡中的楬, August 21, 2016 (http://www.bsm.org.cn/show_article.php?id=2609), accessed November 21, 2017. If the character 江 is read as kong 空, “leave empty” (Chen Songchang, Yuelu shuyuan cang Qin jian [wu], 107), the word would simply mean that no characters are to be written at the very top of the respective piece of writing support. This of course likewise implies the possibility that this blank space may have been filled with ink.

97. The difference between qian 前 and gu 故, “former, original” (cf. preceding slip 117 of this regulation), is as yet unclear. Both seem to refer to earlier submissions that were apparently re-submitted as part of a new submission. Possibly, gu was the “first,” qian the “previous” in a series of submissions related to the same official matter.

98. See n. 34 above. Shu 疏 probably implies the use of du rather than die, at least in the present context. Cf. the view of the Yuelu Academy editors, who suspect it to be a “type of document.” See Chen Songchang, Yuelu shuyuan cang Qin jian (wu), 154n55.

99. This means that the minimum thickness of die is exactly two-thirds of the minimum thickness of du. Note that the Shuowen jiezi commentary by Duan Yucai also suggests that die are thinner than du. See Shuowen jiezi zhu, 7.318: 厚者爲牘，薄者爲牒。 “Thick [pieces of wood] become du, thin ones become die.”

100. This maybe refers to concrete standard pieces or models to be created locally, because the abstract norm itself must have been created by the central government. Cf. the local production of standard weights during the Qin period; on this see Li, Kin Sum (Sammy), “To Rule by Manufacture: Measurement Regulation and Metal Weight Production in the Qin Empire,” T’oung Pao 103.1–3 (2017), 24CrossRefGoogle Scholar.

101. With regard to the fact that “imperial du”—probably du that were sent by and/or to the Qin emperor—were supposed to have a length of one foot and one inch, it seems very likely that “normal” du had a shorter standard length of only one foot (c. 23 cm), cf. the examples of du from Liye well no. 1 in Table 1. In fact, this would correspond to the standard lengths for pieces of writing support used for normal official documents vs. imperial documents (more precisely imperial decrees) in the Han period. See Itaru, Tomiya 冨谷至, ed., Kan kan goi kōshō 漢簡語彙考証 (Tōkyō: Iwanami, 2015), 17Google Scholar.

102. The Yuelu Academy editors placed Yuelu Lüling 122 at the very end of this ordinance, which would add a final die 牒 to the last sentence. The whole sentence would read geng chi yi cun du die 更尺一寸牘牒 “Change [the standard length of imperial] du and die to one foot and one inch.” This has not been accepted here for the following reason: The title on Yuelu Lüling 122 reads zuling bing si 卒令丙四 “Distributed ordinances, C4.” In fact the exact same title also occurs on Yuelu Lüling 1160, where, however, it follows an ordinance with an entirely different content. See the transcription published in Zhou Haifeng 周海鋒, “Qin lüling zhi liubu ji suizang lüling xingzhi wenti” 秦律令之流布及隨葬律令性質問題, Huadong zhengfa daxue xuebao 華東政法大學學報 2016.4, 51–52. Furthermore, there is a duplicate of that ordinance, which is included in volume 5 of the Yuelu Academy manuscripts (see Yuelu Lüling 111). Interestingly, only the beginning of that duplicate is found on a fragmented slip, while the rest is missing. I suspect that Yuelu Lüling 122 carries the end of the duplicate that begins on Yuelu Lüling 111. As to the die 牒, it does admittedly not occur on Yuelu Lüling 1160, which likewise carries the title zuling bing si 卒令丙四. However, there are a few examples, where the main text of an ordinance is followed by a note on the number of die on which it was written. See, for example, Yuelu Lüling 145 as well as its duplicate Yuelu Lüling 0476 (unpublished). With this in mind, the text before Yuelu Lüling 122 might be tentatively emended as ling yi 令一, leading to the note ling yi die 令一牒 “ordinance on one die” at the end of the ordinance presumably consisting of Yuelu Lüling 111 and 122.

103. See n. 95 above.

104. Most single-piece manuscripts found at Liye well no. 1 have at least three columns of writing. See Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 115. In fact, Sumiya noted cases in which the writing seems to have been deliberately stretched into three columns although it would easily have fit into two columns (e.g., Liye 8-1559). This also points towards the possibility that single-piece manuscripts were supposed to have at least three columns of writing—which would match the description of du in the Qin ordinance.

105. Further evidence for this is the occurrence of expressions such as jiao quan yi die 校券一牒, “checking tally, one die,” on Liye 8-135 and 8-677.

106. An analysis of the length to width ratio of pieces referred to as du or die is indeed revealing. With regard to the pieces collected in Table 1, the ratio for pieces named du (all values rounded to whole numbers) is between 6 (Liye 8-1517) and 13 (Liye 8-1566). If we calculate the ratio for the standard measurements stipulated in the Qin ordinance discussed above, the numbers are even lower, between 5 (length of c. 23 cm/width of c. 4.4 cm) and 8 (length of c. 27.5 cm/width of c. 3.5 cm). The situation is much different for pieces named die. The measurements in Table 1 give ratios around 15 for the two complete/completely reconstructed pieces (Liye 8-1069+8-1434+8-1520, 9-2352). If the tallies (quan)—also referred to as die and collected in Appendix B—are considered, the numbers are even higher. The lowest ratio is 17 (Liye 8-900+8-1583), the highest even 34 (Liye 8-2246), with the length to width ratio of most pieces ranging between 20 and 26. Although no width is stipulated for die in the cited Qin ordinance, it is possible to infer the intended width of pieces with only one or two columns of writing from the measurements prescribed for pieces with three, four, or five columns. The latter would correspond to widths of 1.16, 0.96 and 0.87 cm per column. This would mean that pieces for one column of writing were probably supposed to have a width between 0.87 and 1.16 cm (roughly 1 cm), while pieces for two columns were probably supposed to have a width between 1.74 and 2.32 cm (roughly 2 cm). This would suggest length to width ratios between 12 (length of c. 23 cm/width of c. 2 cm) and 28 (length of c. 27.5 cm/width of c. 1 cm). To sum up, the length to width ratio of pieces designated as du appears to range between 5 and 8 (stipulated) or 6 and 13 (actual examples); the ratio of pieces designated as die ranges between 12 and 28 (stipulated, hypothetic) or 15 and 34 (actual examples). Depending on whether one takes a normative or descriptive approach, the threshold between du and die—based on length to width ratio—was roughly at 10 or 14, respectively.

107. Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian (yi), preface, 2.

108. See Table 1 in Zhang Chunlong 張春龍, Ōkawa Toshitaka 大川俊隆, and Momiyama Akira 籾山明, “Liye Qin jian kechi jian yanjiu: jian lun Yuelu Qin jian Shu zhong de wei jiedu jian” 里耶秦簡刻齒簡研究——兼論嶽麓秦簡《數》中的未解讀簡, Wenwu 文物 2015.3, 53–69+96.

109. For these see Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian (yi) (layer 8) as well as Liye Qin jian bowuguan et al., Liye Qin jian bowuguan cang Qin jian (layers 9 and 10).

110. In the above ordinance the length of “imperial du” was set to about 25 cm. As pointed out to me by Tang Junfeng, if this length was in fact reserved for certain du supposed to exhibit the emperor’s authority and status, it would seem unlikely that longer pieces—even if they would theoretically have to be classified as du—were actually referred to as du. See also n. 101 above. It is actually possible that all pieces that exceeded the standard (or “imperial”) length of du, probably 23 up to 25 cm, were referred to as die.

111. In the Zhuangzi, du occurs in one of the so-called miscellaneous or mixed chapters, which includes some material said to be authored by Zhou, Zhuang 莊周 (4th c. b.c.e.). See Zhuangzi jijie 莊子集解, ed. Xianqian, Wang 王先謙 (Beijing: Zhonghua, 1987), 8.281Google Scholar; Watson, Burton, The Complete Works of Chuang Tzu (New York: Columbia University Press, 1968), 356Google Scholar. However, it is difficult to precisely date the passage in question as the extant recension is that of Guo Xiang 郭象 (d. 312 c.e.). See Roth, Harold D., “Chuang tzu,” in Early Chinese Texts: A Bibliographical Guide, ed. Loewe, Michael (Berkeley: Society for the Study of Early China, Institute of East Asian Studies, 1993), 56–57Google Scholar. As to the Zhanguo ce, the alleged compiler Liu Xiang (first century b.c.e.) stated that it is based on sources that existed already during the Warring States period. See Tsuen-Hsuin, Tsien, “Chan kuo ts’e,” in Early Chinese Texts: A Bibliographical Guide, ed. Loewe, Michael (Berkeley: Society for the Study of Early China, Institute of East Asian Studies, 1993), 5Google Scholar. If this is true for the passage in question (Zhanguo ce 13.473)—and provided it remained unaltered by Liu—then it would prove that du was already used to refer to writing support in pre-imperial China. At least the events narrated in the passage took place in the early second half of the third century b.c.e., as they focus on the death of Queen (Mother) Junwang of Qi 齊君王后 in 249 b.c.e. (Shi ji, 46.1902).

112. Xu Xiqi, Xi-Zhou (Gonghe) zhi Xi-Han lipu, 1250.

113. 8-461 is the number according to the edition in Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian (yi). In earlier articles the board was usually referred to by its excavation number (8-455). The piece has received considerable scholarly attention. For two of the earliest works see Chunlong, Zhang 張春龍 and Jingsha, Zhang 張京沙, “Xiangxi Liye Qin jian 8-455 hao” 湘西里耶秦簡8-455號, Jianbo 簡帛 4 (2009), 11–15Google Scholar; Pingsheng, Hu 胡平生, “Liye Qin jian 8-455 hao mufang xingzhi chuyi” 里耶秦簡8-455號木方性質芻議, Jianbo 簡帛 4 (2009), 17–25Google Scholar.

114. Jian, Chen 陳劍, “Du Qin Han jian zhaji san pian” 讀秦漢簡札記三篇, Chutu wenxian yu gu wenzi yanjiu 出土文獻與古文字研究 4 (2011), 370–75Google Scholar.

115. Xing Yitian 邢義田, “‘Shou,’ ‘ban,’ ‘yue Wu yue Jing’ yu ‘Qianling gong’: Liye Qin jian chudu zhi yi” “手”、“半”、“曰啎曰荊”與“遷陵公”——里耶秦簡初讀之一, May 7, 2012 (http://www.bsm.org.cn/show_article.php?id=1685), accessed November 21, 2017.

116. Xu Xiqi, Xi-Zhou (Gonghe) zhi Xi-Han lipu, 1250.

117. Wei, Chen 陳偉 et al. , Qin jiandu zhengli yu yanjiu 秦簡牘整理與研究 (Beijing: Jingji kexue, 2017), 10–11Google Scholar. Chen Wei has pointed out additional roughly contemporary lexical changes, for example, the change from the standard reference to the first month (in administrative documents) from duan yue 端月 to zheng yue 正月 between 220 and 218 b.c.e. and the change of the terms chen qie 臣妾 or nu qie 奴妾 for private slaves to nu bi 奴婢 between 219 and 215 b.c.e. See his “Qin bihui ‘zheng’ zi wenti zai kaocha” 秦避諱“正”字問題再考察, August 27, 2014 (http://www.bsm.org.cn/show_article.php?id=2062), accessed November 21, 2017; “Qin bihui ‘zheng’ zi wenti zai kaocha buzheng” 秦避諱“正”字問題再考察補證, September 1, 2014 (http://www.bsm.org.cn/show_article.php?id=2067), accessed November 21, 2017; as well as “Cong ‘chenqie,’ ‘nuqie’ dao ‘nubi’” 從 “臣妾”、“奴妾” 到 “奴婢”, January 27, 2017 (http://www.bsm.org.cn/show_article.php?id=2715), accessed November 21, 2017.

118. The graphical similarity of the two forms in fact led to misidentification of ban 半 as shou 手 by the Liye editors in certain cases. See Chen Jian, “Du Qin Han jian zhaji san pian,” 370–75.

119. The earliest such tally, dating exactly to the first day of the thirty-first year, is Liye 8-1545.

120. Hand measurements of the photographic reproduction of Liye 8-1551 (in Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian [yi]) rather suggest a length of c. 35.2 cm. However, even this would be closer to “one foot and five inches” than to “one foot and six inches.” On an interesting side note, a manuscript excavated from Zhangjiashan tomb no. 247 describes a mathematical problem that might speak in favor of this hypothetic change of standard length. It asks how many slips (jian 簡) with a length of one foot and six inches could be produced from a bamboo culm segment that would yield exactly 366 slips with a length of one foot and five inches. See Suanshu shu 71. Admittedly, quan seem to have been produced from wood rather than bamboo, but the occurrence of exactly these two measures of length in a mathematical problem related to the production of writing support is still noteworthy.

121. He in fact considers the three as representing the same word, which he reconstructs as *lêp. See Schuessler, ABC Etymological Dictionary of Old Chinese, 213.

122. If not otherwise noted, Old Chinese reconstructions follow William H. Baxter and Laurent Sagart, “BaxterSagart Old Chinese reconstruction, version 1.1 (20 September 2014),” (http://ocbaxtersagart.lsait.lsa.umich.edu/BaxterSagartOCbyMandarinMC2014-09-20.pdf), accessed November 21, 2017.

123. As of now, the evidence for the claim that du 牘 and du 獨 are definitely etymologically related is not sufficient. Therefore, this is no more than a hypothesis based on phonological similarity between two words. And even the seemingly most up-to-date Old Chinese reconstruction by Baxter and Sagart, on which I base this, is not uncontested. See, for example, the critique in Harbsmeier, Christoph, “Irrefutable Conjectures: A Review of William H. Baxter and Laurent Sagart, Old Chinese. A New Reconstruction (Oxford: Oxford University Press, 2014),” Monumenta Serica 64.2 (2016), 445–504CrossRefGoogle Scholar. Still, at least according to the reconstruction, the phonological similarity between the two words is rather striking. Among the nearly 5,000 words for which Baxter and Sagart offer Old Chinese reconstructions, there are overall 27 words ending in OC * ˤok. As far as the initials are concerned, there is of course quite a variety. The two words under discussion also seem to have different initials, “*l” in the case of du 牘 vs. “*[d]” in the case of du 獨, although the square brackets show that the latter reconstruction is less secure. However, apart from du 牘 and some other words written with a character indeed sharing the same phonophoric (such as du 讀 “to read,” du 櫝 “coffer; box,” and du 犢 “calf”), du 獨 is the only one among said 5,000 words that evolved to Middle Chinese duwk. Phonological similarity between du 獨 and words written with the phonophoric 賣 may also be gleaned from the fact that there is at least one example where the character 獨 was used to write du 犢 “calf.” See Feng Qiyong and Deng Ansheng, Tongjia zi huishi, 382.

124. At least this is true with regard to all submissions to superior authorities that fall into the categories “request” (qing), “answer” (dui) or “memorial” (zou)—without doubt at least a significant part of all official correspondence.

125. However, the expression die du 牒牘 may also serve as a generic term for manuscripts made of wood (or possibly bamboo). See, for example, Lun heng jiaoshi, 13.613: 豈徒用其才力，游文於牒牘哉？ “How could [governors, ministers, and high functionaries] merely use their intellectual faculties for scribbling on die and du?” The translation was adapted from Forke, Alfred, Lun-Hêng, vol. 2 (New York: Paragon Book Gallery, 1962 [1911]), 301Google Scholar.

126. The word occurs only once on Jianshui Jinguan 肩水金關 73EJT21:213 with only rudimentary context. For the manuscripts excavated at Jianshui Jinguan see Gansu sheng jiandu baohu yanjiu zhongxin, Gansu sheng wenwu kaogu yanjiusuo, Gansu sheng bowuguan, Zhongguo wenhua yichan yanjiuyuan gu wenxian yanjiushi, and Zhongguo shehui kexue yuan jianbo yanjiu zhongxin, eds., Jianshui Jinguan Han jian (yi–wu) 肩水金關漢簡（壹—伍）, 5 vols. (Shanghai: Zhongxi, 2011–16)Google Scholar. Nor can it be found in the “old” or “new” Juyan manuscripts, nor the manuscripts from Dunhuang 敦煌 or Changsha Wuyi guangchang 長沙五一廣場 published so far.

127. Tomiya, Kan kan goi kōshō, 361–67 (including a discussion of the two multi-piece manuscripts Juyan xinjian E.P.F22:56–60 and E.P.F22:80–82); cf. Kyōto daigaku jinbun kagaku kenkyūjo kantoku kenkyūhan 京都大学人文科學研究所簡牘研究班, ed., Kan kan goi: Chūgoku kodai mokkan jiten 漢簡語彙——中国古代木簡辞典 (Tōkyō: Iwanami, 2015), 260Google Scholar. In Han documents, the piece with the cover letter normally followed the attached pieces, at least in case these contained a kind of register (buji 簿籍). See Hou Xudong 侯旭東, “Xibei suo chu Handai buji ceshu jian de pailie yu fuyuan: cong Dong-Han Yongyuan bingwu bu shuoqi” 西北所出漢代簿籍冊書簡的排列與復原——從東漢永元兵物簿說起, Shixue jikan 史學集刊 2014.1, 58–73.

128. Enno Giele originally suggested this interpretation of die in Han documents to me and also proposed the English translation “attachment.”

129. I am grateful to Tang Junfeng for the proposal that the creation of the terms zha and liang hang itself might point to certain changes in the way the term die was understood in the Han period.

130. Ernian lüling 256. For this manuscript see Zhangjiashan ersiqi hao Han mu zhujian zhengli xiaozu, Zhangjiashan Han mu zhujian [ersiqi hao mu]. Translation adapted from Barbieri-Low and Yates, Law, State, and Society in Early Imperial China, 703.

131. The expression shu shu also occurs in connection with other single-piece manuscripts; see n. 34 above.

132. Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 121–22.

133. See in detail Barbieri-Low and Yates, Law, State, and Society in Early Imperial China, 70–72, 698–99 (on Ernian lüling 246–48 and a textual parallel on a wood tablet dated to 309 b.c.e.).

134. On the dating of the Ernian lüling see Barbieri-Low and Yates, Law, State, and Society in Early Imperial China, 62–64.

135. See Changsha jiandu bowuguan, Changsha shi wenwu kaogu yanjiusuo lianhe fajuezu, “2003 nian Changsha Zoumalou Xi-Han jiandu zhongda kaogu faxian” 2003 年長沙走馬樓西漢簡牘重大考古發現, Chutu wenxian yanjiu 出土文獻研究 7 (2005), 57–64Google Scholar; cf. Shaohua, Song, “Changsha chutu de jiandu ji xiangguan kaocha” 長沙出土的簡牘及相關考察, Jianbo yanjiu 簡帛研究 2006 (2008), 249–62Google Scholar.

136. See Changsha shi wenwu kaogu yanjiusuo, “Hunan Changsha Wuyi guang-chang Dong-Han jiandu fajue jianbao”; Changsha shi wenwu kaogu yanjiusuo, Qinghua daxue chutu wenxian yanjiu yu baohu zhongxin, Zhongguo wenhua yichan yanjiuyuan, and Hunan daxue Yuelu shuyuan, eds., Changsha Wuyi guangchang Dong-Han jiandu xuanshi 長沙五一廣場東漢簡牘選釋 (Shanghai: Zhongxi, 2015)Google Scholar.

137. See Changsha shi wenwu kaogu yanjiusuo, and Zhongguo wenwu yanjiusuo, eds., Changsha Dongpailou Dong-Han jiandu 長沙東牌樓東漢簡牘 (Beijing: Wenwu, 2006)Google Scholar.

138. See Changsha shi wenwu kaogu yanjiusuo, Zhongguo wenwu yanjiusuo, Beijing daxue lishixue xi, and Zoumalou jiandu zhengli zu, eds., Changsha Zoumalou Sanguo Wu jian: Jiahe limin tianjia bie 長沙走馬樓三國吳簡——嘉禾吏民田家莂, vol. 1 (Beijing: Wenwu, 1999), 30–35Google Scholar as well as Figure 41 (for J22-2695).

139. Tsuneko, Sumiya 角谷常子, “Kantoku no keijō ni okeru imi” 𥳑牘の形狀における意味, in Henkyō shutsudo mokkan no kenkyū 邊境出土木𥳑の研究, ed. Itaru, Tomiya 冨谷至 (Kyōto: Hōyū, 2003), 89–118Google Scholar; cf. Takamura, “Shin Kan jidai no toku ni tsuite,” 59, 63–66. On private letters from early imperial China see Giele, Enno, “Private Letter Manuscripts from Early Imperial China,” in A History of Chinese Letters and Epistolary Culture, ed. Richter, Antje (Leiden: Brill, 2015), 403–74Google Scholar. The so-called “greeting tablets” (ci 刺 or ye 謁) that have been found in ancient wells and tombs in central China would have to be added to the common types of single-piece manuscripts during the Han period. For an in-depth study of the greeting tablets see Korolkov, Maxim, “‘Greeting Tablets’ in Early China: Some Traits of the Communicative Etiquette of Officialdom in Light of Newly Excavated Inscriptions,” T’oung Pao 98 (2012), 295–348CrossRefGoogle Scholar.

140. Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 109–13.

141. Takamura, “Shin Kan jidai no toku ni tsuite,” 62, 66–68. It has to be noted that the distinction made by Takamura is between toku 牘 and sakusho 冊書. He defines the former as “pieces that were planned to act as writing support and which carry at least three columns of writing” (Takamura, “Shin Kan jidai no toku ni tsuite,” 58). This definition alone would not equate the respective pieces with single-piece manuscripts. However, the fact that toku are contrasted with multi-piece manuscripts (sakusho) and that Takamura considers cases in which the former became pieces of multi-piece manuscripts as exceptional (Takamura, “Shin Kan jidai no toku ni tsuite,” 62) suggests that he also generally understands toku as pieces used for what have been termed “single-piece manuscripts” in the present paper.

142. One of course has to consider that to date the only actual examples of Qin administrative documents are the finds from Liye. Whether the widespread use of single-piece manuscripts may to some extent reflect a regional peculiarity will only become clear as soon as additional Qin documents from other regions are discovered.

143. Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 121–24.

144. Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 126–27.

145. Takamura, “Shin Kan jidai no toku ni tsuite,” 60. Cf. Sumiya’s contrary view that the amount of characters to be written does not seem to have been the reason for the choice of single- or multi-piece manuscripts. See Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 123.

146. For a multi-piece manuscript consisting of several formerly independent multi-piece manuscripts see, for example, Juyan 128.1. In principle, there is no limit to the length of a text that can be written on one multi-piece manuscript. In actual practice, multi-piece manuscripts of course get too unwieldy at a certain point. It was proposed years ago that the maximum number of pieces in administrative multi-piece manuscripts might have been around 100. See Yitian, Xing 邢義田, “Handai jiandu de tiji, zhongliang he shiyong: yi Zhongyanyuan shiyusuo cang Juyan Han jian wei li” 漢代簡牘的體積、重量和使用——以中研院史語所藏居延漢簡爲例, in Di bu ai bao: Handai de jiandu 地不愛寳——漢代的簡牘 (Beijing: Zhonghua, 2011), 22–23Google Scholar. This assumption has now received excellent backup from the Qin ordinance discussed above, which stipulates exactly 100 die as the maximum.

147. Possible drawbacks concerning revision, (re-)organization and storage with regard to single-piece manuscripts were also suggested by Sumiya and Takamura. See Sumiya, “Riya Shin kan ni okeru tandoku kan ni tsuite,” 126–27; Takamura, “Shin Kan jidai no toku ni tsuite,” 60n15.

148. The lack of sufficiently large pieces of wood may indeed have prevented the frequent use of single-piece manuscripts in Juyan and Dunhuang. See Takamura, “Shin Kan jidai no toku ni tsuite,” 61. The frequent reuse of writing support, which can be observed in the northwestern border regions of the Han Empire, also hints at the relative scarcity of raw materials. On reuse see in detail Takeyuki, Takamura 高村武幸, “Kantoku no sai riyō: Kyoen Kan kan wo chūshin ni” 簡牘の再利用——居延漢簡を中心に, in Bunken to ibutsu no kyōkai: Chūgoku shutsudo kantoku shiryō no seitaiteki kenkyū 文献と遺物の境界——中国出土簡牘史料の生態的研究, eds. Momiyama Akira 籾山明 and Satō Makoto 佐藤信 (Tōkyō: Rokuichi, 2011), 163–84Google Scholar.

149. Although wider pieces were sometimes made from bamboo as well, it was much more suitable for the production of comparatively narrow pieces, convenient to apply only one column of writing. Enno Giele also pointed out that pieces of bamboo are much less suitable for two-sided use. See Giele, “Private Letter Manuscripts from Early Imperial China,” 411n16.

150. See the Qin ordinance discussed above. In Han documents recording the transfer and disbursal of materials used for manuscript production, zha and liang hang occur frequently but are only exceptionally described in further detail. See the examples in Annuo, Ji 紀安諾 (Enno Giele), “Handai biansai beiyong shuxie cailiao ji qi shehui shi yiyi” 漢代邊塞備用書寫材料及其社會史意義, Jianbo 簡帛 2 (2007), 475–500Google Scholar. This also suggests that zha and liang hang were normally produced in standard formats.

151. Tomiya, Kan kan goi kōshō, 15–19.

152. It has also been argued that during the Han period pieces for only one column of writing (zha) were mostly used for drafts, whereas clean copies were usually written on wider pieces for two columns (liang hang). See Sumiya, “Kantoku no keijō ni okeru imi,” 98. This would support the assumption that the latter were generally considered more valuable.

153. See n. 139 above. Of course, the comparatively low amount of text usually written in private letters or on greeting tablets probably was another reason for the use of single-piece manuscripts.

154. The decision on the most probable reference object of die of course depends on the question of whether die occurs in an “original document” or whether the respective part of text was merely copied into a subsequent document (e.g., an answer to the original document). In the former case, the interpretation “same piece” or “attached piece(s)” is most likely, whereas in the latter die most probably refers to some other document. See also n. 32 above. The table only includes die-references that are likely to be part of “original documents.” A piece is assumed to represent an original document if there was no evidence suggesting that it might be a copy (as, for example, in Liye 8-135 or 8-653, where additional parts of text with more recent dates written by the same hand can be found on the same piece of writing support). In case a piece is so fragmentary that it provides too little evidence to make any judgement on whether die most likely refers to the same piece, an attached piece or to an entirely different document, it has not been included in the table. This is the case for the following ten pieces: Liye 8-5, 8-225, 8-234, 8-235, 8-528+8-532+8-674, 8-804, 8-1715, 8-2003, 8-2035, and 8-2543. Furthermore, Liye 8-1041+8-1043 was likewise not included, as it constitutes an unfinished or partly deleted document used for writing exercises. The measurements of Liye 9-1869 and 9-2352 were taken from Liye Qin jian bowuguan et al., Liye Qin jian bowuguan cang Qin jian, Appendix 2. All other measurements provided in the table are based on hand measurement of the photographs in Hunan sheng wenwu kaogu yanjiusuo, Liye Qin jian (yi).

155. The measurements of Liye 8-762, 8-763, 9-762, and 10-1157 were taken from Liye Qin jian bowuguan et al., Liye Qin jian bowuguan cang Qin jian, Appendix 2. Note that Liye 8-762 and 8-763 must be sought under their excavation numbers (8-763 and 8-764) in that volume. The other measurements, if not noted otherwise, follow Zhang Chunlong et al., “Liye Qin jian kechi jian yanjiu,” Table 1. For the reconstruction of Liye 8-886+8-1220 see Zhang Chi 張馳, “Liye Qin jian quan lei wenshu zhuihe san ze” 里耶秦簡券類文書綴合三則, July 31, 2015 (http://www.bsm.org.cn/show_article.php?id=2276), accessed November 21, 2017. For Liye 8-1590+8-1839 see Yao Lei 姚磊, “Liye Qin jiandu zhuihe zhaji (er)” 里耶秦簡牘綴合札記（二）, June 7, 2015 (http://www.bsm.org.cn/show_article.php?id=2254), accessed November 21, 2017.