A Study of the Format and Formation of the * Si gao 四告 ( Four Proclamations ) of the Tsinghua Bamboo Slips

The *Si gao 四告 (Four Proclamations) manuscript, published in the tenth volume of the Qinghua daxue cang Zhanguo zhujian 清華大學藏戰國竹簡 (Warring States Bamboo Slips in Collection of Tsinghua University), consists of four texts of prayers of Zhou Gong Dan 周公旦 (Dan, the Duke of Zhou), Zengsun Boqin 曾孫伯禽 (Great-Grandson Boqin; i.e., the Duke of Lu 魯公), Zengsun Man 曾孫滿 (Great-Grandson Man; i.e., King Mu of Zhou 周穆王), and Zengsun Shao Hu 曾孫召虎 (Great Grandson Shao Hu; i.e., Duke Mu of Shao 召穆公). They are not only important in terms of their content, but are also presented in a quite special format of considerable significance for understanding the process of book formation in pre-Qin times.

bamboo and silk 6 (2023) 1-21 blank except for numbers.3 These blank slips are different in nature, the blank slips between two sections used in the *Si gao for the purpose of distinguishing two parts of the main text and can therefore be called "separator slips," whereas the ones at the end of the manuscript roll containing the Xingcheng and Bingfang can be called "surplus slips." This reminds us of the blank bamboo slips that have often been found among previously excavated materials. For example, the excavation report of Tomb No. 1 of Guodian 郭店, Jingmen 荊門, says that there are 804 bamboo slips.4 However, only 731 slips have been published. According to the editors, the other slips bear no writing.5 Among these blank slips, the fragments may originally have been blank slips at the end of a paragraph or a chapter, while the complete ones should be the separator slips or surplus slips left over from the bamboo roll. With information concerning the size and shape of these blank slips, the carved lines on the back of the slips, and the binding notches, it might be possible to reconstruct the original state of the manuscript, which would help clarify the relationship between text and manuscript roll. A roll with a separator slip or surplus slip would have been bound before writing, which is reflected in the format of *Si gao. The most remarkable feature is that the text avoids the binding notches so that no graph would have been obscured by the binding cords throughout the whole roll. Moreover, the text is written from the top of the slip to the bottom, so it was necessary to avoid both the cords in the middle and lower parts of the slip. The places marked in Figures 1 and 2 are close to the middle and lower cords. The spacing between graphs is somewhat uneven at the places of the binding cords.
From the extant traces of the binding cords at the notches on slips 10 to 20 (see Figure 3), we can understand that this roll was once folded with slip 15 at the central axis. As shown by previous research,6 this roll was likely buried in a folded position. This provides an additional concrete example for the study of how ancient bamboo-slip manuscripts were rolled and folded up.  It is likely that the binding and writing of the *Si gao were done by different people, which can be seen by the analysis of the handwriting in the next section. However, this leaves the question of whether the slips were bound into a roll by the same person? I have also found some clues to answer this question.
According to the positioning of the carved lines on the backs of the slips, the bamboo used in this roll came from three different sections of bamboo: Slips 1 to 24 belong to one section, Slips 26 to 49 to another section, and Slip 50 to yet another section. Although all three sections are of the same length and width, there are some differences in how they were prepared. For example, the first section has rounded edges at the spots that were scraped away on the back of the slips for the writing of the slip number (see Figure 4), while the second section has flat edges there (see Figure 5), and the third rounded edges again. These different shapes suggest different methods of preparation, which must have come from different craftsmen or different batches. Since the bamboo was not all prepared together, the production of the roll probably also involved two separate procedures. The evidence in the *Si gao shows that there was a division of labor in the production of the roll, from the preparation of the bamboo material through to the writing of the text. This is very important for understanding the manufacture of bamboo-slip manuscripts at that time.

Six Types of Handwriting and Four Scribes in the *Si gao
One purpose of studying the handwriting on bamboo slips is to distinguish different scribes and to understand as much as possible how a manuscript was written, how different batches were produced, and the relationships between different texts. If multiple types of handwriting appear in the same text (or manuscript roll), then it is also possible to discuss the relationship between the various parts of the manuscript. This type of study usually has a tacit premise that different types of handwriting represent different scribes. Nevertheless, a trained calligrapher is able to write in different styles or to imitate a variety of handwriting. Also, the same scribe may also write differently due to different times of writing, different materials used, or different orthography in the base text. Thus, to understand the writing on the slips, we should also distinguish between different handwriting styles of the same scribe. The differences in the structure of the graphs in *Si gao are quite subtle, but a close look reveals six different types of handwriting belonging to four different scribes. For the sake of presentation, I will number these types of handwriting A to F according to their order of occurrence in the text. Types B and E are likely to have come from the same hand, and D and F were also written by one scribe. The relationship between the six types of handwriting and scribes is shown below.
Scribe I -Type A Scribe II -Types B and E Scribe III -Type C Scribe IV -Types D and F The following discussion is based on five calligraphic features to analyze the similarities and differences between these different types of handwriting. For the sake of analysis, I will mainly compare common graphs (or components).
(1) Type C (Scribe III) is relatively unique, and is easily distinguished by the two graphs tian de 天悳 written in a smaller size between the graphs de 悳 and yong 用 just below the middle binding cord of slip 11 (see Figure 3). They clearly represent the addition of missing graphs. In particular, the "heart" 心 signific of the graph 悳 is written in a different style from that of the other graphs with it, and it can be inferred that Scribe III was a proofreader. (2) The main differences between the three scribes responsible for Type A (Scribe I), Type B \ E (Scribe II) and Type D \ F (Scribe III) can be seen in such frequently used graphs as 氒, 之, 命, 敢, 典, 先, 辟, 我/義, 亡, etc.  (3) Type A (Scribe I) calligraphy uses thick, bold strokes, with the brush pressed hard. The four graphs 拜= =者魯 at the beginning of "Proclamation I" all show the same characteristics, especially the graph 者, which shares the same characteristics as the graph 者 in "Proclamation II" (Slip 16) but differs from the others.
bamboo and silk 6 (2023) 1-21 It should be noted that, given the harder pressing of the brush of Type A, what the editor has transcribed as (任)and 壬(任)on slip 16 should be read as 功 and 工. Graphs like 王, 皇, 惠, and 周 are also characterized by a thickening in the middle of a vertical stroke. A similar phenomenon can also be observed in the Shanghai Museum slips. For example, the graphs 厇(宅)不工(空) on slip 18 and the graph 工 in the phrase 乃立 (禹)以爲司工 on slip 23 of the Rongchengshi 容成氏 also have similar vertical strokes. (4) Regarding the similarities and differences between Types B and E, the graphs 氒, 之, 命, 先, and 亡 listed in section (2) are sufficient to show that these two types are from the same hand. However, I distinguish them because there are still some obvious differences in the way some graphs are written, such as 我/義. Consider the four graphs 事, 邦, 民, and 弗.  (2), we can see from the graphs 氒, 之, and 我 that Handwriting D and F belong to the same hand. The similarities between D and F can be seen in the graphs 母 and 若. However, the reason for distinguishing these two hands is that many graphs are written differently, bamboo and silk 6 (2023) 1-21 such as the graph 辟 mentioned above. Another four examples can be seen below: 余, 其, 畢, and 朕. In addition, in Type F the ends are more rounded than those of Type D. The common features between Types D and F: The main differences between types D and F: As for where exactly Type D starts, it is easy to see that the strokes of the graphs 其 at the end of the third slip of "Proclamation III" (Slip 18) are more slender than those in the preceding text. If we compare the upper and lower parts of the graph with the graphs 之, 先, and written by Scribe IV, we can further understand that they are likely to have been written by the same person. Jia bamboo and silk 6 (2023) 1-21 Table 9 Scribe IV's transition to Type D script It is also worth mentioning that the handwriting of Scribe IV is also found in the text of Feng Xu zhi ming 封許之命.8 The same characteristics occur in the following eight graphs in both texts: 之, 晨, 雚, 光, 於, 乍, , and 永. Of course, due to the influence of the base texts, there are also many differences in the structures (or components) of the graphs in these two texts.
8 Jia Lianxiang 賈連翔, "Feng Xu zhi ming zhuibu ji xiangguan wenti tanyan" 《封許之命》 綴補及相關問題探研, Chutu wenxian 出土文獻 2020.3, 13-20. Apart from the main text, there is little material with which to compare the calligraphy of the sequence numbers on the backs of the slips, so it is impossible to judge the relationship between it and the handwriting of the main text. Thus, for the time being, I list it as yet another type of handwriting.

Reflect the Formation of the Manuscript
The distribution of Type A is most unique. It is found in the first four characters of the "Proclamation I" (Slip 1) and on the first three slips of "Proclamation II" (Slips 16-18). Hand B covers the majority of "Proclamation II" (Slips 1-14), with Hand C (Slip 11) appearing once as a proofreader's note. Hand D begins with the last two graphs of the third slip of "Proclamation II" and ends with the end of "Proclamation II" (Slips 18-24). Hand E is found in "Proclamation III" (Slips 26-37), while Hand F is found in "Proclamation IV" (Slips 38-50). The general distribution of the different hands is shown in Figure 6, with the details given in Figure 7. Based on the distribution of the calligraphy, we can clearly see that the *Si gao manuscript was not created in one go. Moreover, we can also make further inferences about the actual process by which the four parts were written. "Proclamation I" was completed by Scribe I (Type A) and Scribe II (Type B) together. From the order of their handwriting, it would seem that Scribe I started the copying, and then Scribe II finished it. Later, Scribe III (Type C) joined to correct the two graphs 天悳 on slip 11. bamboo and silk 6 (2023) 1-21 "Proclamation II" was copied by Scribe I and Scribe IV (Type D) working together. From the order of their handwriting, it would seem that Scribe I started with the first three slips (slips 16-18) but then was interrupted for some reason. Beginning with the last two graphs of the third slip (slip 18), Scribe IV continued to finish the remaining part.
Since the handwriting of Scribe I in "Proclamation I" and "Proclamation II" is largely consistent, it is likely that he produced these two passages at the same time. It is thus likely that he was the one who planned the manuscript's layout. Of course, there is also another possibility, which is that "Proclamation I" and "Proclamation II" were first copied by Scribes II and IV, and then Scribe I made changes to their text. There were not many changes in "Proclamation I," but the changes made to "Proclamation II" were more substantial. However, since there is no evidence of any changes in Type A, there is not much support for the latter conjecture.
"Proclamation III" was copied by Scribe II (Type E) again. This time, he seems to have written it with great care, and there is no obvious trace of proofreading changes throughout the text.
"Proclamation IV" was again copied by Scribe IV. He must have made an estimate of the number of words of "Proclamation IV" and the number of "surplus slips" before he started copying. He probably found that there were not enough bamboo slips left, so he discontinued the previous format of having "spacer slips." Instead, he started writing directly from the next slip (Slip 38) following "Proclamation III." Based on the results, the adjustment he made was necessary, since the two graphs 祜福 at the end of this part did indeed reach the last slip of the roll (Slip 50).
It is worth noting that not only did Scribe II and Scribe IV write different graphs in their two batches of work, their strokes also look somewhat different. I have pointed out that the end of the strokes of Type F are clearly more rounded than those of Type D. For those who are experienced in calligraphy, it is easy to tell that this is often caused by the use of a different brush. In a sense, the change of scribes and writing tools may suggest a certain interval between the production of the four parts of this manuscript.

The Formation of *Si gao Indicated by the Process of Manuscript Production
The process of copying *Si gao by different scribes in several batches also reflects that the four proclamations must have been four separate texts, each with a "base text" for copying. Although it is almost impossible for us to discern bamboo and silk 6 (2023) 1-21 I have pointed out elsewhere that the three parts of *Yue ming were originally compiled in a single manuscript roll.10 Now, we can see even more clearly that they were produced in the same way as *Si gao. Although the four parts of *Si gao were not transcribed by a single person at one time, the contents follow a strict chronological order, suggesting that the production of the manuscript was guided by an overarching plan. This might suggest that *Si gao was produced in the context of an editorial project of similar literature. It was not simple copying, but a project with specific purposes, either for pedagogical purposes or in the service of a ruler. In either case, this edition has a degree of "originality." It could have been used as a new "base text" to influence the production of following manuscripts, or it could have disappeared completely after being buried with the tomb owner, so it might also be a "temporary" edition. As mentioned above, "remaining faithful to the base text" was an important principle for the scribes of ancient texts. However, was the compiler who conceived the manuscript necessarily "faithful to the base text"? The Gushi bian 古史辨 (Debates on Ancient History) group represented by Gu Jiegang 顧頡剛 (1893-1980) has referred to "stratigraphy" in ancient texts. Every time a text is copied or recompiled it provides an opportunity for the transformation of the "base text." If we consider the *Si gao as a record of the prayers of the Western Zhou period, the historical figures recorded in the four parts spanned over two hundred years. However, the four parts have all adopted a highly consistent format. Zhao Ping'an points out that the word hufu 祜福 at the end of each part is a term used after the Spring and Autumn period, which means that "these four proclamations were edited collectively during the Spring and Autumn period and thus infused with some elements of that time."11 This is a transformation that took place even before the formation of the Tsinghua manuscript. The four graphs 拜 者魯 written in Proclamation I by Scribe I are also part of the unified format, so we cannot rule out the possibility that this also reflects a transformation of the "base text." The process discussed above involves the writing of the textual contents on bamboo slips as well as how the material form of the manuscript took shape. This is not the same process of how historical materials are compiled into 10 Jia Lianxiang 賈連翔, "Zhanguo zhushu zhengli de yi dian fansi -cong Tianxia zhi dao, Ba qi wu wei wu si wu xing zhi shu, Yuxia Yin Zhou zhi zhi san pian de bianlian tanqi" 戰國 竹書整理的一點反思-從《天下之道》《八氣五味五祀五行之屬》《虞夏殷 周之治》三篇的編聯談起, Chutu wenxian 出土文獻 13 (2018), 142-152.

11
Zhao Ping'an, "Qinghua jian Si gao de wenben xingtai ji qi yiyi." bamboo and silk 6 (2023) 1-21 written documents, but rather how a written document is transformed into another written document. Specifically, I have used the calligraphic characteristics in the Tsinghua manuscripts to describe how the four parts of *Si gao were compiled. The new written document produced by this editing could mark either the advent of a new "base text" or the end of previous "base texts." Based on this, we can assume that each instance of copying could exert a similar impact on the transmission of ancient texts, which helps us to better understand the formation of pre-Qin texts more broadly.