This book discusses what happens when multiple texts are brought together on the same manuscript. It examines how texts, whether originally in oral, remembered, or written form, were gathered and integrated on manuscripts, how collections were received, and how they influenced transmission and shaped access to knowledge and text. This study asks what changed when independent sayings, stories, verse, and prayers were brought together into a single collection. How did the physical format of the manuscripts shape the texts they carried? How did recipients approach and use collections, and what did they take away from them? What extra levels of meaning and potential for interpretation were unlocked? How did qualities such as morality, causality, genre, and technique, which transcend individual texts, emerge out of the collocation of similar texts?
I use the term “collection” as a broad term to refer to single manuscripts containing multiple (originally) distinct texts.1 These texts can be integrated as a composite or simply included side by side. Collections provided a selection of text considered useful and important by their producers and users, and transmitted these selections to later generations of users. I use “collection” to cover the whole process or life-cycle if you will, involving gathering, selecting, integrating, and, organizing, transmitting, and receiving text in manuscript form.
Two short examples from Chapter 1 show why I study such a broad range of features of gathering text through the lens of collections. The Guodian *Yucong
The text includes a number of such blocks of rhymed sayings, but what sets it apart, really, from a manuscript such as the Shanghai Museum Fan wu liu xing
(7) […] 祭祀奚升[*lhəŋ]， 吾如之何使飽[*prûʔ]? 順天之道[*lûʔ]， 吾奚以為首[*lhuʔ]? 吾欲得 (8) 百姓之和，吾奚事之[*tə]? 敬天之明奚得[*têk]? 鬼之神奚食[*m-lək]? 先王之智▂奚備[*brəkh]?
What rises up in sacrifices, and how does one satiate [the ancestors]? In following the way of heaven, what does one take as beginning? When one wants to obtain the compliance of the hundred surnames, how does one serve them? How does one respect heaven’s brilliance? How do the spirits of ghosts eat? How is the wisdom of the former kings procured?
聞之曰: 升 (9）高從埤[*pe]，至遠從邇[*neʔ] 。十圍之木[*môk]，其始生如蘖[*ŋat] 。足將至千里[*rəʔ]，從寸始[*lhəʔ] 。┗
I have heard it said that: “To rise high you have to start from below. To reach far you have to start from nearby. A tree of ten arm spans is like a sapling when first growing. When you are about to traverse a thousand miles on foot, you have to start with an inch.”3
A series of questions is answered with reference to a selection of sayings. This pattern is repeated throughout the text, and additionally, there is a topical focus on the “holding on to the one” (zhi yi
Certain collections drew on paratextual and visual forms of organization to bring multiple texts together in an almost rhetorical statement of completeness, a classic and highly advanced case can be found in what has been called the Four Lost Classics from Mawangdui.5 That silk manuscript collects several texts and carefully structured collections of sayings and narratives. It is famous for its association with the Yellow Emperor and parallels to the transmitted Laozi
Because levels of organization present a sliding scale from the “ragbag” sometimes referred to in the literature,7 all the way to the profound unifying meanings sometimes discovered in collections,8 it makes sense to treat them together. As material objects, they would have come to their users as a single unit, and these users would have a range of possible ways of finding common ground among the materials of the collection. Who is to say that the reader(s) or teacher(s) of *Yucong 4 likewise did not imbue it with more complex levels of organization? An added advantage of treating collections of different levels of integration together is that it allows us to look for shared patterns in the (re-)use of text and manuscript across genres. This way, we can explore larger questions on the role that collections in manuscript form played in (re-)distributing, (re-)integrating, and (re-)interpreting knowledge from the stream of tradition as opposed to “schools,” “traditions,” and “genres,” for example.
I use “collections” to cover both “multiple-text manuscripts” (MTM s) and composite texts. The former term focuses on the inclusion of multiple distinct texts on a single material carrier.9 The latter, while likewise composed at least in part of sections of previously independent text, integrates these into a single composition.10 Multiple composite texts were also sometimes gathered on a single manuscript, forming a collection of collections in its own right. I include both of these elements of manuscript and text production to acknowledge that sometimes the lines are hard to draw. Take the Shanghai Museum Zhuangwang ji cheng
My use of “collection” in this study is meant to reflect this broad range of functions enabled when manuscripts bring together a variety of types and forms of text onto a single carrier. A fair share of these collections may not have started out in manuscript form but emerged in oral traditions, and some of them continued to grow over time, and some became organically integrated to the extent that they came to be perceived as a single text.12 Many of the technical collections discussed in Chapter 4 are MTM s in that they contain a planned selection of multiple individually differentiated texts within the space of a single material carrier and were often expanded at later stages. Some of these feature a selection so varied that they are often referred to as miscellanies.13 Others, such as the historiographical collection Rongchengshi
Some of the manuscript texts discussed in this book will resemble composites and others will behave more like an MTM, and it is not clear to what extent early Chinese recipients cared about such distinctions.14 As such, while these terms are useful for clarifying the production of collections, they are less useful for understanding questions of reception. The way collections were approached by users was not solely determined by the origins of the texts or the way these were integrated in the production of the manuscript. The users and producers of collections likewise did not strictly adhere to the genre divisions that crystalized in the early empire, as evidenced by the collection of historical and divinatory, or calendrical and hemerological, materials on a single carrier, as discussed in Chapter 4.
To take an example from Chapter 4, the *Liang chen
The visual layout of the *Liang chen and *Zhu ci
Source: Li Xueqin 2012: 16, 18. Image courtesy of the Research and Conservation Center for Unearthed Texts, Tsinghua University, Beijing
For example, on slip two of the *Liang chen we find a list of ministers for Tang
唐有伊尹，▬有伊陟，▬有臣扈。█
Tang had Yi Yin, Yi Zhi, and Minister Hu.
A typical example from slip 1 of the *Zhu ci can be found in the following prayer against drowning:
恐溺，乃執幣▬以祝曰: “有上茫茫，有下湯湯，司湍滂滂，侯茲某也發揚。” ▬乃舍幣。█
When in fear of drowning, grab a bolt of silk, and say the following prayer: “Above it is brimming, below it is rushing, the manager of rapids is vast and flowing, let this person soar!” Then discard the bolt of silk.
The two texts are vastly different in genre and style, and while both present a list of collected items, knowledge of ministers of antiquity was arguably used in different contexts than prayers against drowning. As such, the reason for collecting these two collections on the same manuscript should probably not be sought primarily in the texts but rather in their material form. The use of two instead of three binding strings and the use of the full length of the slips including the fragile ends suggests that little care was taken to protect the writing on the slips. The extensive use of punctuation to divide the items in the list may point to uses for memorization or consultation as it would have helped a reader to parse the text and highlight individual items. Maybe the manuscript was used as an aide de memoire, and was meant to transfer a collection of text to the user’s memory. This brings us to the problem of orality and memory culture.
1 A Note on Orality
The nature of orality precludes any discussion of collections before their appearance in written form. We do not have concrete evidence of the oral practices that likely informed much of the production, teaching, transmission, and reception of text in early China. In all likelihood, advanced wielders of text that did not rely on written artifacts carried a “collection” of materials in their memories.16 Given the lack of evidence to substantiate the details and extent of these practices, this book analyses collections preserved in writing. To what extent and at which moments oral traditions fed into written collections is likewise unclear, but for the vast majority of texts, especially sayings, anecdotes, songs, and even spells and recipes, it seems clear that at some points these texts were generated and transmitted in settings that did not rely (at least extensively) on writing.
I also do not believe the evidence suggests that with the spread of writing, oral traditions simply disappeared; instead, as I have argued elsewhere, writing would need to prove its qualities and advantages over other forms of communication for an extended period.17 Where such phenomena have been documented in greater detail, oral traditions and the written word coexisted for long periods of time.18 Anecdotes and sayings were still remembered, retold, and changed in the telling. Songs were still sung and appreciated for their rhythm and musicality, and technical experts were still being hired—all while the manuscripts of the period increasingly circumscribed the texts that accompanied these practices as stand-alone, fixed, and organized, and with a clearer awareness of specific wording and meaning. In other words, lived traditions that continuously reshaped the cultural body of text did not just stop or become irrelevant with the use of written collections.
Nevertheless, collecting texts in manuscript form must have influenced the ways in which users interacted with the contents and communicated the materials over time and distance. Collecting text in manuscript form provides a platform to aid memorization, a central selection of texts around which oral teachings can be organized, or it ossifies in written form a snapshot of the range of texts that communities relied on in their still largely oral and memory-based text practices, all the while making a ritual and rhetorical claim to completeness of knowledge.19 In their written form, these manuscripts would have been better suited for shaping the reception by communities and preserving continuities of selection, organization, and interpretation over distance and time, thereby influencing the access to text by later generations.20 These selections, once put to writing, need not have been final, and Chapter 4 especially will highlight several examples in which texts were added to existing collections—collecting is an ongoing process, and in order to fully appreciate this process the whole life-cycle of a manuscript needs to be considered.21
In all likelihood, many of the hallmark collections of early China, say the Lunyu
2 Collecting Text in a Manuscript Culture
Writing preserved from before the Warring States period occurs predominantly on ritual objects of bronze, stone, and bone, but these weighty media tend to contain single texts and appear far removed from what we comfortably recognize as collections.24 Obviously, these media were not the sole or indeed the predominant method of transmitting text. Indeed, the transmission of ideas and cultural memory among the literate elite appears to have drawn heavily on oral, ritual, and other performative modes of communication.25
Written evidence for the collections central to this book appears first on bamboo, and to a lesser extent, silk and wooden, manuscripts from the Warring States period and early empires (roughly the fifth to second centuries BCE). As the evidence stands, it appears that during this period, a manuscript culture emerged, or at least gained in prominence, and there was an increase in the spread and reliance on lightweight carriers such as bamboo manuscripts for transmitting and communicating text and ideas.26 Before this period, there is no hard evidence for the widespread use of collections or lightweight manuscripts, and while it seems likely that these existed they cannot be shown to have had a widespread impact.27 The term manuscript culture used here denotes more than just the existence of manuscripts,28 rather it points to these exerting a broad and noticeable influence in the development of written culture, whether it be significant developments in administration, literature, historiography, thought, or technology.29
I am therefore not claiming that the lack of bamboo or other manuscripts in tombs or other sites dated before the Warring States period suggests that they were not used at all. For one, this would simply reflect a preservation bias, on which more below.30 I am claiming that as of yet there are no indications, archaeological or historical, that bamboo or other lightweight media played a significant role in communicating, let alone broadly disseminating, text and knowledge beyond small enclosed circles. Only from the Warring States period onward is physical evidence available for such an increase in reliance on manuscripts.31 This phenomenon coincides with a marked increase in reflections on the use of written media in transmitted literature, developments in the vocabulary for reading, and changes in the perception, production, use, and forms of text.32
Traditional narratives place the origins of the classical tradition in the early Western Zhou with texts such as the Documents (Shu
At this point, there are still no claims of a widespread circulation and use of physical texts. Neither is there any claim towards significant production or reproduction of text in writing. Indeed, references to the use of written text in transmitting thought appear only in texts traditionally dated to the Warring States period.34 If one were to subscribe to the position that an elevation in the status of written text started roughly with Confucius, is it so difficult to imagine that the widespread adoption and social effects of such practices occurred only slowly within the centuries that followed his death?
At the same time, an expansion in the user base of the manuscripts occurred. Writing moved out of the confines of the central and local courts, and as it did, it was increasingly used by persuaders, thinkers, and teachers—often grouped together as the shi
3 Collections in Manuscript Form
At a certain moment—we only have material evidence from the mid-Warring States period onwards—it appears that a critical mass was reached. Texts that were previously circulating independently, whether in oral, remembered, or written form, were brought together in the space of a single manuscript, forming collections. Archaeologically excavated tombs, whether of lowly clerks or aristocrats, and recovered looted caches of texts have revealed rather eclectic selections of manuscripts spanning the spectrum of traditional schools of thought and divisions of literature.38 They tend to be shorter compositions, often without a title, and without any indication of authorship. Some (parts of) classical texts known from tradition have been unearthed too, but even these familiar texts tend not to be of the same length, order, or even wording as their received counterparts.39 Comparing the unearthed to the transmitted record thus reveals significant differences in both the kind and distribution of text.
Collections occupy a surprisingly central place within this newly discovered realm of manuscript and text. Whether in Warring States tombs containing predominantly literary materials such as Guodian tomb 1; retrieved caches of looted manuscripts spanning the range of technical and literary manuscripts such as those held by Tsinghua University and the Shanghai Museum; or the early imperial tombs predominantly containing technical materials such as Zhoujiatai, they all contain several collections alongside single-text manuscripts.40 Why are collections so prominent in the early textual landscape?
I suggest that one of the main forces behind the formation of collections is exactly the proliferation of written text during the period. Remembered text appears to have been increasingly written down in manuscript form, and new compositions were likewise increasingly written out on bamboo. An increase in the overall presence of text presents a twofold problem: how to gather as many texts as possible in a context of limited access and how to select relevant texts from an increasing range of options.41
In all likelihood, there was no indexed library or reference system before Liu Xiang
Relevance presents the second problem encountered with increased availability of text. Unearthed evidence has revealed the presence of multiple copies of the same manuscript text, sometimes in different versions, even within the same cache.43 Sections of text, whether individual sayings, songs, or stories, likewise occurred in varying versions within the same corpus and across the textual landscape as a whole.44 These different takes on the same subject present a problem of choice. As will become clear in the chapters below, collections had different strategies for dealing with this problem. Some, such as the technical collections discussed in Chapter 4, likely in response to problems of access mentioned above, chose to embrace variance, and often collected different versions of the same recipe or spell. Others, such as the anecdote collections discussed in Chapter 2 made a selection, and chose a particular version most suited to the argument of the collection even though we know there were different versions potentially available.45 As Bausi notes for Ethiopian manuscripts, the collection often works as a corpus organizer in that it purposefully selects a relevant subset of texts from the larger realm of texts in the ether, whether in song, memory, or written form.46 This observation tallies well with early Chinese collections, which drew from a vast range of songs, anecdotes, sayings, and techniques preserved in the cultural memory of early China as well as more recent texts reflecting current concerns.47
4 The Form and Use of Collections
Most of the collections familiar to us in Chinese literary history were state-sponsored megaprojects, such as the Zhong shu
On the material side, collections drew on the visual and physical potential of the carrier to organize texts on the manuscript. This codicological toolkit included, among others, the size, binding, and form of the manuscript, the use of punctuation, titles, and subtitles, or simply the layout and order of the texts.51 These material characteristics of the manuscript can be fruitfully analyzed as paratextual devices mediating the reader’s engagement with the materials and reveal how they were likely meant to be used.52 Punctuation and titles, for example, guide the reading process by dividing individual texts, signaling a reader where to pause, pay attention, or differentiate sections, and in turn favor certain usages of its texts. In terms of the general production quality, some manuscripts are easily identified as rough copies or the work of scribes in training, whereas others were produced with more care, including binding and formatting that favored increased durability in use, or more legible and carefully executed writing. Given that information on the collectors and recipients themselves is largely absent, as is true for most early texts, an analysis of manuscript codicology allows us to make more reasoned predictions about the envisioned use of the manuscript.
From a literary point of view, the collections under discussion offer varying degrees of topical grouping and other forms of textual organization including, for example, chronological, narrative, or rhythmic arrangements. These structures embed the texts in the collection under a larger context or even intellectual program and in doing so reveal the function, if not the underlying rationale, of gathering a particular set of texts together. Structures such as rhyme guide the reading process, signaling what text belongs together, and frame narratives focus the envisioned interpretation of the songs or stories they embed, for example.
As a result of material and literary organization, the collection is more than just the sum of its parts. Whether planned out by a collector or perceived by a recipient, multiple texts gathered in close physical proximity will always have a greater inherent potential for meaning than a single unit simply by virtue of the interplay between the component parts.53 The materiality of the manuscript collections that this book discusses is crucial. Given the high level of intertextuality of many early Chinese texts and the weight of remembered tradition brought to any reception of text, the act of gathering a selection of texts within a single volume focuses the reception experience and narrows the horizon of expectations.
Put simply, texts bound together are more easily perceived as related, and as such connections are more easily perceived within the limited space of the collection. As a result, functions or qualities emerge from perceived relations between the texts. Some of these qualities are prefigured in the text, for example in the form of genre labels or recurring structuring devices highlighting causality.54 In other cases, the relations between the materials emerge from the recipient’s discovery of areas of commonality or difference among the materials. Most recipients would recognize a topical grouping and would observe similarities among materials without these necessarily being explicitly marked, and a more daring recipient might extend such similarities to draw broader conclusions about, say, the moral, historiographical, or technological potential of a particular arrangement of texts.55 In that sense, the collection on the one hand draws on a recipient’s horizon of expectations (i.e., their existing assumptions of similarity and difference), while on the other hand it actively constructs the literary microcosm that shapes those very expectations. This brings us to the potential influence of collections on the language habits and textual preferences of recipients.
5 The Role of Collections in Early China
How did early collections contribute to the development of early Chinese literature as a whole? Consider for example the influence of the following classic collections. These examples all ended up with clear bibliographic categorizations, if only through the craft of the Han librarians, and exerted their influence over writing in the genres they shaped.
The sayings collected in texts now known as the Laozi and Lunyu were fixed in form, or, in the case of the Laozi, given a specific authorship attribution, only later in their life.56 These collections influenced discursive habits for years to come, and texts such as the Wenzi
The study of collections, therefore, is fundamental to a revaluation of how text and manuscript worked and were used during the Warring States period and how they shaped the texts that followed in their wake. Arguably, their contribution lay not so much in presenting new knowledge, but rather in recombining, organizing, and transmitting knowledge in ways optimized for various usage contexts and making it accessible to a wide range of audiences. Whether as school text, manual, handbook, or as aide de memoire, the collection was a popular and effective method of gathering the knowledge of the time and redistributing it among an increasingly growing readership. Through this process, collections shaped the access to textual knowledge of future readers and writers, and in the process, molded their perceptions and writing habits.
A recognition of the role of collections therefore provides a corrective to views on text formation and transmission overly focused on institutional forces of canonization, or narratives tracing authorship and literary influence as the driving force in the development of early Chinese literature. Accordingly, the study of collections allows us to sidestep traditional divisions by text type to find common ground between prayer and song, anecdote and saying, and appreciate fully the broad interests of Warring States and early imperial collectors, a fact often hidden by traditional understandings categorizing text and user by school and master.62
The role of collections in shaping the literary world of the Warring States and beyond should also not be overstated, however, as the legacy of collections in shaping the literary habits and preferences of their users is far from absolute. The song collections discussed in Chapter 3 make this abundantly clear. Many of the topical groupings or even selections of material advanced by these collections did not leave a clear imprint on the shape of later collections. As with many other recently retrieved texts, many of these texts were not transmitted. In other words, the unearthed evidence reveals the range of possibilities faced by Warring States and early imperial text users but certainly does not indicate a linear path of textual formation.
Likewise, while collections were potentially open to continuous recompilation, and some of the examples discussed in this book indeed reveal additions over time, it is also clear that the collection’s re-use of existing text does not fit the image of the ring binder, wherein users would freely slot in or take out materials.63 As the evidence stands, there are no clear signs of unwanted text being removed other than the deliberate destruction characteristic of preparing goods for the grave. Likewise, adding materials often followed a program of sorts even if only by leaving space for future materials.64 The evidence for continuous rebinding of materials is not as forthcoming, although undoubtedly such instances also occurred, but it seems likely that this process of reorganization may have occurred in the formation of a new collection.
6 Sources and Caveats
Using Warring States manuscripts, while avoiding Han and later imperial constructions, comes with a set of problems of its own. To prepare a manuscript for reading in modern orthography, scholars rely on methods from paleography, phonology, and codicology to reconstruct a text. Just as the Han librarians, modern scholars come with a set of assumptions about standards of transcription, what constitutes a valid reading of a graph, and how texts relate to the manuscript for instance. Improvements in publishing standards, an emerging methodological discussion, and most importantly access to (images of) the manuscripts themselves, however, open these assumptions to critical appraisal and necessary revision.
The question to what degree manuscript finds are representative of larger trends is significant. The majority of Warring States finds comes from tombs in the ancient Chu
Preserving the Tsinghua University slips in trays of distilled water
Image courtesy of the Research and Conservation Center for Unearthed Texts, Tsinghua University, Beijing
Ordering the Tsinghua University slips
Image courtesy of the Research and Conservation Center for Unearthed Texts, Tsinghua University, Beijing
Another explanation is that placing texts in tombs was an aspect of Chu mortuary culture, which slowly spread across early China during the early empires.67 Some of the manuscripts, such as tomb inventories and underworld narratives, were specifically produced for the grave, and others show signs of deliberate destruction in the form of erasing of characters and breaking slips, effectively rendering them into burial artifacts.68 While this aspect needs to be taken into account when using manuscripts as source material, the lion’s share of the material does not appear to have been specifically prepared for the grave and can be considered to have had a lifetime before entombment, accordingly the material can be used as a source for broader intellectual and cultural developments at the time.
This study draws from a wide range of manuscript caches from the Warring States period, including Baoshan
The material from the Shanghai Museum and Tsinghua University collections, for example, stems from tomb robberies and was purchased on the Hong Kong antique market in 1994 and 2008 respectively. As such, many vital clues from the burial context are unknown, including the status of the tomb owner, the way in which the manuscripts were buried, and the extent to which the manuscripts within the collections are complete or miss certain slips or sections.72 The collections were dated on the basis of characteristics of the calligraphy, and a calibrated carbon dating of unwritten slips. The Shanghai collection materials were dated to 260 BCE ± 65 years, and the Tsinghua materials to 305 BCE ± 30 years.73 In the process, paleographers, scientists, and archaeologists from several competing institutions have been involved in the process of authenticating, organizing, and transcribing the materials. The consensus is that the materials are authentic artifacts from the mid to late Warring States period.
Some concerns have been raised regarding the authenticity of looted manuscripts. With famous forgeries such as the Zhejiang University manuscripts in mind,74 isolated scholars have asserted that some of the manuscripts are forgeries, and by extension, that the status of the sources as a whole is compromised.75 A closer look at the arguments advanced for this doubt reveals that the main source of concern lies in (a) a perceived discrepancy between the received textual tradition and the excavated manuscripts, and (b) a lack of understanding of the scribal processes that govern manuscript cultures. For example, texts from the Tsinghua manuscripts that relate narratives about the early Western Zhou are accused of containing Warring States grammatical and lexical phenomena, and as such should not be considered authentic Western Zhou texts.76 This position quickly fails in light of arguments showing that texts such as the Zhou Wuwang youji Zhougong suo zi yi dai wang zhi zhi
Criteria in the identification of forgeries have been advanced by Hu Pingsheng and Xing Wen.81 By now, controlled excavated discoveries have confirmed supposedly suspicious readings of graphs in looted manuscripts, while at the same time new phenomena discovered in acquired manuscripts such as the numbers and lines on the back of bamboo slips found in the Beida and Tsinghua collections have been retraced in earlier excavated collections such as Guodian or Baoshan.82 With many of these original questions dispelled, a majority of experts is now confident in using properly tested looted manuscripts as valid source material for the period. Nevertheless, I have tried to balance discussions of retrieved manuscripts with contemporary materials as much as possible and am confident that the core argument of this book on the role and use of collections in this period does not hinge on the status of a single cache or source.
7 Outline of the Study
This study consists of four chapters, each focusing on one genre of writing that made up the lion’s share of unearthed texts for the period: philosophical sayings, historical narrative, song and poetry, and techniques to control health and fortune. Because these genres of writing are vastly different, they each serve to highlight specific characteristics resulting from gathering multiple texts together into single collections. Taken together, the chapters of this book sketch out different phases in the life-cycle of collections, from their intertextual source materials, their inclusion or integration into collections, and the influence that collections exerted on their users, to the ways in which they shaped the future understanding and availability of their texts.
In the first chapter, “Manuscript Materiality: Organizing Sayings in a Collection,” I examine the role of manuscripts in the formation of text, and the structuring and reception of knowledge. The chapter explores the role of materiality in the formation and use of collections and discusses how collections of sayings inform intertextuality, influence language habits, and instill ideas on morality. An increase in the use of visual and rhythmic structuring was formative to new ways of arranging texts into building blocks, influencing reading strategies and the reception of knowledge. These collections selected and organized the vastly increasing amounts of knowledge in circulation, and as such, could serve as bases for memorization and education. Manuscripts gained in importance in teaching practices and enabled the reproduction of proper language habits, while at the same time they provided a handy digest of apt phrases and arguments, to be used in persuasion or argumentative settings.
In the second chapter, “Collecting Stories: The Reformation and Integration of the Past,” I analyze the formative role of collections in showing how anecdotes were integrated into a single narrative, and what this meant for the structuring of the past, the writing of causality, and the shaping of cultural memory. These narratives developed from isolated and local accounts predominant in the Western Zhou through the Spring and Autumn periods, which relied on contextual and memorized information. From the mid to late Warring States we see the emergence of comprehensive, universally oriented, and causally structured accounts. Having juxtaposed these two ideal types of representing the past, I analyze the development between them on the basis of a number of excavated historiographical mini-collections.
In Chapter 3, “Collection and Canon: The Formation of a Genre,” I focus on the changes to the use, understanding, and genre classification of verse resulting from collections. It describes how collections shaped access and thus influenced future understandings of verse. The chapter traces the Warring States development in conceptualization of verse from contextually situated song to poetry interpreted for its wording. The role of the collection in wedding selections of verse to interpretive frames shaped perceptions of genre, authorship, and interpretation. Their use as teaching material would be instrumental in spreading these perceptions of the material to shape the way verse was understood in later periods. In this way, written collections laid the foundations for many developments that emerged in the canonization of verse in the early empires.
The final chapter, “Collecting and Disseminating: Using Technical Knowledge,” analyses the use of collections by focusing on the technical miscellanea from the early empires. It describes how expertise shifted from specialists to texts, and how collections were used didactically as handbooks, manuals, and aides de memoire. These collections reveal a change from the recording of proceedings by experts towards the collection of technical knowledge for a broader audience. Technical knowledge, whether of medicine, magic, divination, or hemerology, moved away from being the sole prerogative of the expert, and was presented more and more as a set of instructions to follow. Whether in materials prepared for education, handbooks, or as aide de memoire, collections of technical knowledge provided the tools for regular users to use technical knowledge to help them navigate the problems of daily life. Technical collections, being concerned with concrete problems, reveal intended usage and the concerns of their user-producers much more clearly than their literary counterparts. Accordingly, they provide good evidence for the access to knowledge of the literate strata of society, but even more, they offer an insight into the motivations, needs, and reasoning behind their selections of knowledge.
Other, more specific terms such as “compilation,” “compendium,” “anthology,” and “florilegium,” are predominantly used to describe products of mature text cultures, involving highly conscious selection and editorial procedures, and seem less applicable to the wider, and less strictly circumscribed, range of options of gathering text that I discuss in this book.
The edition and translation follows Cook 2012: 910–11. Phonological reconstructions follow Schuessler 2009 where possible, but rely on Baxter and Sagart 2014 when reconstructions are absent in Schuessler.
The edition and translation are my own, for a discussion, see Chapter 1 below.
Both of these collections pale in light of the concerted levels of organization seen in the Han Feizi
Vankeerberghen 2015. Note especially her discussion on pp. 314–15 on the processes necessary for the creation of the manuscript and the collection of its texts, to form what she calls an “inclusive text.”
See You 2021 for a study analyzing the paratextual organization of the Shiliu jing
De Reu 2015: 243–96.
For example, the “one thread” tying together the teachings of Confucius in the Lunyu; see Makeham 2003 and Hunter and Kern 2018: 1–2.
See the introduction to Friedrich and Schwarke (2016: 15–16) and the definition in the follow-up volume, Bausi, Friedrich, and Mariachi (2019: vii) where MTM s are considered “made up of more than one text and have been planned and realised for a single project with one consistent intention; as a result, they are usually made of a single production unit.”
For a discussion on composite texts, see Boltz 2005.
For an example of these approaches, see Krijgsman and Vogt 2019.
The line is not easily drawn between composite texts and MTM s: when does an arrangement of previously existing texts become a single unit? In cases, it has been shown that MTM s were instrumental in the formation of new, singular textual units, effectively forming a new text out of multiple parts. See, for instance, Xiao 2013.
On technical miscellanies, see for example Harper 2018.
It should be noted that for much of the Warring States period, a single manuscript carrier appears to refer to a single roll of bamboo, silk, or bundle of wood. But this general observation comes with two major caveats. For one, the poor state of preservation of many of the manuscripts and the state of disarray that saw them emerge from the tomb or antique market—and as such often compounded with a lack of archaeological context—makes it hard to establish the existence of manuscripts composed of multiple rolls. Analysis by Staack 2018 shows that at least from the Qin dynasty onwards, such distinctions were made for administrative documents, and it is entirely likely that material forms of gathering multiple manuscript rolls (whether of similar material features or not) likewise existed during earlier periods.
As an example, the Guodian *Laozi manuscripts discussed in Chapter One were composed of three rolls with different material characteristics and may be considered three different production units. Nevertheless, it is entirely possible that these were considered to form a single unit by a recipient, regardless of these material differences. Judging from the way texts were organized in the Han imperial library, it seems that similar concerns governed the understanding of manuscript and text; for an overview, see Fölster and Staack 2021.
Li Xueqin 2012: 3.156–65. For a recent overview of the manuscript and the script, including a discussion on Jin script influences on the writing, see Li Songru 2020.
Li Ling 2004: 202.
Krijgsman 2019: 104–08.
For example, Clanchy 2013: 255, 295; Thomas 1992: 12–14.
Sanft (2019: 54–73) notes how access to text and interpretive frameworks was often mediated by communities and did not necessarily take written form.
For comparative cases, see Johnson 2010 and Stock 1983.
See, for example, Johnston and Van Dussen 2015: 8–10.
Denecke 2010: 25, 66.
For discussions on oral literature and collections arranged by list and corpora, see Vansina 1985 and Finnegan 1988 respectively. For theorization in relation to early text cultures, see Jan Assmann 2011.
While there is evidence of brush writing and the existence of bamboo manuscripts from the Shang dynasty onwards, as of yet there is no hard evidence for their widespread use (i.e., beyond ritual centers), or a broad reliance of society on these materials. For the Shang evidence, see Bagley 2004: 218.
The use of written text on ritually significant objects gains in prominence and functionality during the Western Zhou period, but text and object still tend to be wedded into artifacts of ritually encoded meaning, whether they were boundary markers, contracts, oaths, or simply ritual vessels used in sacrifices; see for instance Kern 2009: 143–200; Li Feng 2011: 271–301.
Based on ritual texts and sacrificial bronzes, von Falkenhausen (1993, 139–226) argues that bamboo manuscripts played a role in many of the ritual proceedings that accompanied the bestowal of gifts and sacrifices, but he likewise notes the prominence of oral performance over text in the transmission of ideas and memory.
See especially Meyer 2011, Richter 2013, Nylan 2011, Galambos 2006. For developments in early medieval China, see Nugent, 2011, Tian Xiaofei 2013, Zhang Yongquan 2013.
Pace Li Feng 2012. For a critique, see Krijgsman 2012: 136–38.
The term manuscript culture was popularized by Marshall McLuhan (1962), who used it to distinguish a phase in media development prior to print culture.
Currently the term “manuscript culture” is often used in a configuration of new approaches to pre-print materials. These approaches include, for example, the “New Philology” as described by Stephen G. Nichols (1990: 1–10), a focus on “scribal culture” represented by the work of Karel van der Toorn (2007), and the notion of textual anthropology coined by Markus Hilgert (2010); or in the different forms of cultural exchange enabled and foreclosed by manuscripts as compared to print; see Johnston and Van Dussen 2015: 1. All these approaches share a sensitivity to issues of authorship, the importance of paratextual and other material features of manuscripts, the problem of critical editions, issues of intertextuality, and a consideration of larger cultural developments centering around actor network theory and other models describing the interaction between (groups of) people and artifacts, for instance.
Sobering are the numbers of manuscripts produced per capita in medieval to early modern Europe extrapolated in Eltjo Buringh and Jan Luiten Van Zanden (2009: 409–45). The British Isles during the sixth century produced roughly 0.9 manuscripts per million inhabitants, and front-runner Italy about 25.5 (p. 421, Table 3). What this suggests is that we should not expect to see vast numbers of manuscripts from the early stages of a manuscript culture, even when noting that early China likely had a much larger population than early European regions. See for example, Nishijima 1986: 595–96.
For overviews, see for instance, von Falkenhausen 2003: 439–526, Thote 2016: 11–56, and Tsien Tsuen Hsiun 2013.
Meyer 2014a: 21–38, Lewis 1999 (especially Chapter 2), Behr and Führer 2005: 1–42, Krijgsman 2019: 75–110.
It should be noted that for all these steps in the traditional narrative, we rely on accounts from texts hundreds of years later, and these do not make clear whether or not the use of texts in these circles relied on written text at all. Indeed, record-keeping scribes aside, it is entirely likely that the education of aristocrats in the central states took the form of oral teachings. For a balanced discussion including a review of the traditional account, see Nylan 2001: 16–31.
Krijgsman 2019.
See Lewis 1999 (especially pp. 63–83) and Hsu 1965.
Meyer 2014a.
For an overview in Chinese, see for example Pian Yuqian and Duan Shu’an 2006. For an early overview in English see Giele 1998–1999: 247–337. For indications of spread across states, see Zhou Bo 2012 and Venture 2009: 943–58.
Kalinowski 2003; Li Ling 2004.
See for example Shaughnessy 2021, and Kern 2002.
For a study and introduction of the Guodian materials, see Cook 2012. For the Tsinghua manuscripts, see Liu Guozhong 2011, and the English translation by Foster and French 2016. For a series introducing the Shanghai Museum collection, including translations in modern Chinese, see Ji Xusheng 2004–2017; Richter (2013) also has a clear introduction on the scope of the Shanghai Museum manuscripts. For the Zhoujiatai, and indeed most of the Qin dynasty manuscript finds, the introductions and annotations in Chen Wei ed. 2014 are the standard. Harper 2018 provides a good introduction to the Zhoujiatai technical materials in English.
These two aspects might appear contradictory, but it should be noted that availability in the broad sense does not necessarily mean knowledge of, or access to, this range of options. For a stimulating discussion on how access to knowledge and taste through syllabi informs access to the canon, see Guillory 1993.
Networks and communities can be effective ways of increasing access to text, particularly revealing is a letter from Oxyrhynchus, see Thomas (1992: 1). Nevertheless, even communities would have struggled to gain knowledge of materials outside of their regional sphere or beyond the demographic make-up of their members. How would a local Chu community gain access to a Jin manuscript, for instance?
For studies, see for example Richter 2009; Morgan 2011; Krijgsman 2019: 83–91.
For examples from Guodian and the Shanghai Museum corpus, see Krijgsman 2014.
As shown for instance in the existence of two different versions of the *Wuwang jianzuo
Bausi 2010.
Li Ling 2004: 202, Boltz 2005: 63, Fischer 2009, and Krijgsman 2014 all discuss the practice of early texts to draw from a corpus of readily available (inter)textual material and adapt it to their specific arguments.
For a recent collection of relevant studies on especially later collections, see Jack W. Chen 2021.
For a study see Kurz 2007.
Hunter 2018.
For an overview, see Cheng Pengwan 2017; for studies of manuscript codicology to analyze use, see Richter 2011, and Krijgsman 2018.
Genette 1997.
My understanding draws on reception theory and the study of the book, for example, Jauss 1970; Iser 1978; Fish 1982; Chartier 1994; Genette 1997; McKenzie 1999.
These are paratextual devices in Genette’s sense in that they mediate the relationship between reader and text.
See Krijgsman and Vogt 2019, for an application of these ideas to historiographical MTM s.
Henricks 2000: 7–10; Hunter 2017: 8–12.
Nylan 2013; van Els 2018.
For an excellent discussion of the problems of size and weight involved in writing out a text such as the Shiji on bamboo, see Hsing I-tien 2007.
Durrant 1995: 32–34.
Hunter 2019.
Elman 2007: 143–46.
Csikszentmihalyi and Nylan 2003; Petersen 1995.
Pace Maeder 1992. See Richter 2018 for a detailed analysis of this problem.
Harper 2017: 112–16.
For script analyses and categorizations, see Zhou Bo 2012, Venture 2009, and Feng Shengjun 2007: 250–506. Note also a contemporaneous level of awareness of potential problems in understanding script from different areas. Particularly revealing is the presence of the graph jie
Note also the massive increase in modern development in the region, resulting in a disproportionate number of sites and finds in the region compared to the rest of China. See von Falkenhausen 2003, who also discusses the championing of local identity in Chinese archaeology as partly to blame for this discrepancy in excavations. See also Giele 1998–1999, for a general overview of finds. For the Qin and Han periods, manuscripts have been uncovered from modern Sichuan to the Shandong peninsula.
See especially von Falkenhausen 2003.
For a discussion on the status of manuscripts as burial items, see Giele 2003.
For Baoshan, see Hubei sheng Jingsha tielu kaogudui 1991 and Chen Wei 1996. For Geling, see Wuhan daxue jianbo yanjiu zhongxin and Henan sheng wenwu kaogu yanjiusuo, eds. 2013 and Song Huaqiang 2010.
For Guodian, see Jingmenshi Bowuguan 1998 and Cook 2012. For the Shanghai Museum corpus, see Ma Chengyuan ed. 2001–2012, and Zhu Yuanqing 2002. For the Tsinghua University manuscripts, see Li Xueqin ed. 2010–2018, Huang Dekuan 2019–2022, and Foster and French 2016. For the Anhui University slips see Anhui daxue hanzi fazhan yu yingyong yanjiu zhongxin, eds. 2019 and the studies collected in a special issue of Bamboo and Silk 4.1 (2021).
The collections can be found in Beijing daxue chutu wenxian yanjiusuo, eds. 2014. For up to date text editions and overviews of the scholarship of both Zhoujiatai and Yueshan, see vol. 3 of Chen Wei 2014.
For a discussion on the relevance of archaeological context, see Giele 2003.
See Zhu Yuanqing 2002 and Foster and French 2016.
Cao Jinyan 2011.
For example, Zhu Qixiang 2015. For an early discussion focusing on provenance and provenience, see Goldin 2013. For an overview in English on the problem of forged manuscripts, see Friedrich 2020. For arguments based on the Beida manuscripts, see Foster 2017.
For the *Baoxun in particular see Jiang Guanghui 2009, and the rebuttal in Wang Lianlong 2009.
Meyer 2014b.
See, for example, Baxter 1992.
See, for example, the *Wuwang jianzuo manuscript discussed in Krijgsman 2018: 83–90.
For early observations, see Chen Mengjia 1980.
Hu Pingsheng 2008 and Xing Wen 2012.
See for example, Shan Yuchen 2012; for the lines on the back of the slips, see Sun Peiyang 2011.