The Han Imaginaire of Writing as Weaving: Intertextuality and the Huainanzi's Self-Fashioning as an Embodiment of the Way

Abstract

Writers in the Han dynasty (206 BCE–220 CE) apparently experienced writing in weaving terms. Such an imaginaire of writing as weaving was probably fully manifested in the first or second century BCE and crystallized in the coining of literary terminologies such as classics (jing), weft-writings (weishu), and literature/texts (wen). Situating the Huainanzi and its intertextual writing practice within this imaginaire enables us to reassess both the Huainanzi's widespread dismissal as a miscellaneous, encyclopedic behemoth in the first half of the twentieth century and its reappreciation over the last few decades. According to the Huainanzi's self-depictions, Liu An and his erudite courtiers apparently created the scripture in such an intertextual way in order to textually mimic the process of weaving. Since the Huainanzi commonly associates weaving with the Way's connective powers, the text's extraordinary design might be the result of a literary attempt to create an efficacious, textual artifact that embodies the Way by incorporating the act of weaving in its textual design.

Introduction

The Huainanzi 淮南子 (hereafter abbreviated as HNZ whenever I refer to specific chapters) is a comprehensive scripture from the early Western Han 西漢 dynasty (206 BCE–9 CE) that Liu An 劉安 (ca. 179–122 BCE, r. 164–122 BCE), the king of Huainan 淮南, presumably presented in 139 BCE at his inaugural visit to his nephew Emperor Wu 漢武帝 (born Liu Che 劉徹; 156–87 BCE, r. 141–87 BCE).¹ A prominent feature of this text is its extensive incorporation of passages from pre-Han writings that led to its categorization as a miscellaneous (za 雜) text by Ban Gu 班固 (32–92 CE) during the Eastern Han 東漢 dynasty (25–220 CE; Ban 1958, 4:30.897b). According to Charles Le Blanc (1985, 79, 83), more than a third of the Huainanzi consists of textual parallels to prior writings so that the extant scripture displays an extraordinarily intertextual form.² Moreover, Liu An's text contains various self-reflective passages that allow an evaluation of its design through its own words.³ Accordingly, the Huainanzi offers an exceptional opportunity to evaluate both its own textual structure and early Chinese understandings of (intertextual) writing practices.⁴

To visualize the scripture's textual fabric, I briefly showcase a typical passage from its tenth chapter, “Moucheng” 繆稱 (Profound precepts).⁵ Since I emphasize the Huainanzi's textuality throughout this article, I will present some passages merely from a visual point of view and without translations. The parts written in black refer here to portions that have no parallels with any other extant sources that originated before the Han dynasty. The varicolored sections, however, mark phrases the Huainanzi shares with the extant versions of other pre-Han texts.⁶

According to Sima Tan's 司馬談 (165–110 BCE) later categorization of early Chinese texts into the Hundred Schools (bai jia 百家), these phrases from HNZ 10.77–78 would be attributed to the Confucian or Ruist (Mengzi 孟子 and Xunzi 荀子), Daoist (Laozi 老子), and Miscellaneous (Lüshi chunqiu 呂氏春秋 [Mr. Lü's spring and autumn]) Schools (K. Smith 2003).⁸

In the past, this kind of intertextual writing practice led several reformist intellectuals such as Fung Yu-lan 馮友蘭 (1895–1990), Hou Wailu 後外盧 (1903–87) et al., and Wing-tsit Chan 陳榮捷 (1901–94), as well as Western sinologists such as Henri Maspero (1883–1945), to dismiss the Huainanzi. These scholars interpreted the “Liu clan's scripture” (Liushi zhi shu 劉氏之書; Zhang Shuangdi 1997, 2151), as the Huainanzi refers to itself at the end of the last chapter, “Yaolüe” 要略 (Summary of the essentials; HNZ 21.4),⁹ as a mere encyclopedia that lacks originality and a cohesive intellectual program (Chan 1963, 305; Fung 1983, 396; Hou et al. 1957, 78; Maspero 1924, 12).¹⁰ Over the last six decades, however, various scholars have increasingly argued against such earlier, dismissive readings of Liu An's text and have provided evidence for Liang Qichao's 梁啟超 (1873–1929) statement that the Huainanzi “is highly organized, and is anything but a loose and endless collection of quotations and allusions from a large number of books” (Liang 1971, 105; Le Blanc 1985, 34).¹¹ For example, the Huainanzi Translation Project's scholars recently excavated some coherence underneath the Huainanzi's miscellaneous design by illustrating how the image of root and branches (benmo 本末) manifests in the text's cosmogonic chapter organization and serves as a conceptual template that determines its doctrines and overall outlook on the world. In other words, they argued that the image of benmo is as much an underlying structure of the cosmos as it is an organizing model for Liu An's textual and intellectual project (Major et al. 2010, 13–22; Meyer 2014).

With this article, I join their concerted effort to debunk the Huainanzi's previous dismissal as a hodgepodge of anecdotes and philosophical thoughts by focusing on the way the image of weaving (jingwei 經緯)—like benmo—has apparently been implemented in the scripture's composition.¹² However, I tentatively propose that the Huainanzi's highly constructed design may be read as the remnant of a textual formalization, pointing toward an undocumented yet potentially ritualistic function for the “Liu clan's scripture” at the imperial court in Chang'an 長安 (near modern-day Xi'an 西安).¹³

I develop this argument in the following three steps. First, I show that the Huainanzi repeatedly refers to itself as a textual fabric that connects various aspects of the multifarious universe. Second, I suggest that these self-references may be explained by what I have called an imaginaire of writing as weaving during the late Warring States 戰國 period (475–222 BCE) and Han dynasty.¹⁴ Apparently, contemporaries of Liu An perceived the practices of weaving and writing as parallel processes, a fact that famously crystallized in the coining of literary genres based on weaving terminology during the Han. Third, I suggest accordingly that situating Liu An's text and its intertextual writing practice within this imaginaire sheds new light on the Huainanzi's extraordinary design. In addition to reading the Huainanzi's self-depiction as a fabric in terms of a (lexicalized) metaphor for ordering processes, I consider this image to be an illustration of the scripture's intertextual design. Like the textual materialization of the benmo image in the chapter organization and content of Liu An's text, Liu An and his erudite courtiers also implemented the image of weaving, which the Huainanzi repeatedly uses to depict the sage and the Way (Dao 道), in order to create a homology between the Dao, the sage, and the “Liu clan's scripture.” In this article, I therefore further specify Michael Puett's (2014) groundbreaking assertion that the Huainanzi might have been meant to function as a kind of textual sage. I explicate one of the formalizations, that is, the implementation of the (cosmic) act of weaving via its intertextual writing practice, through which the Huainanzi—like the sage—divinizes and fashions itself as a textual artifact that embodies the Way (tidao 體道) and therewith actualizes its connective powers. In other words, in this article I reverse a common tendency in the humanities to interpret artisanal practices and products with the help of texts by observing the practice of weaving as a means to illuminate textual production in early China.¹⁵ So, let me begin this journey by analyzing some of the Huainanzi's self-depictions.

The Huainanzi's Self-Depiction as a Textual Fabric

The Huainanzi repeatedly utilizes in its self-illustrations an image that gained prominence during the Han dynasty, namely, that writings are textual fabrics. At the beginning of the “Yaolüe” chapter, for example, Liu An's text refers to itself as weaving together a warp-weft fabric (jingwei) and knotting a net (jigang 紀綱):

This first section of chapter 21, which functions like an appended preface (xu 序), explaining its structure, style, and purpose in detail, illustrates in this long passage a general purpose for the Huainanzi before it continues to elaborate on the titles and specific summaries of each chapter. It lays out several of the functions the author(s) of this chapter (presumably Liu An and his courtiers) attributed to the text (Roth 1992, 18–23). On the one hand, it apologizes for the Huainanzi's magnitude. It claims that it was necessary to accumulate that many words on such a wide range of topics in order to avoid any dissection, limitation, and digression from the Great Ancestor—either the Grand One (Taiyi 太一 or 太乙) or the Dao. Only if the text contains everything will it be able “to shift with the times” (yu shi fuchen 與世浮沉) and “roam and rest with the transformations” (yu hua youxi 與化遊息), meaning it will be able to respond to the various ages and the Myriad Beings (wan wu 萬物). On the other hand, it provides a rather explicit and vivid depiction of what the Huainanzi supposedly does; namely, it asserts that it serves as a scripture that “knot[s] a net of the Way and its Potency” (jigang daode 紀綱道德) and “weave[s] a fabric out of humankind and its affairs” (jingwei renshi 經緯人事). Thus, the passage presents the Huainanzi as an utterly comprehensive text that connects rather than dissects.

This vision of the Huainanzi as a weaver of a fabric, however, does not just appear in the “Yaolüe's” general depiction of the “Liu clan's scripture.” In the first part of the “Yaolüe's” summary of chapter 20, titled “Taizu” 泰族 (The exalted lineage), Liu An's text also refers to the same image. It states:

In the “Yaolüe,” Liu An and his workshop apparently illustrate the “Taizu” chapter as a text that connects and “weaves together the paths/teachings of the past and present” (jing gujin zhi dao 經古今之道) and “assembles the aims of the myriad methods/regions” (zong wan fang zhi zhi 總萬方之指). In so doing, it “weave[s] together the governing ways and knot[s] together the kingly affairs” (jingwei zhidao, jigang wangshi 經緯治道，紀綱王事). Consequently, it seems as if the Huainanzi does not use the compounds jingwei and the term jing 經 as nouns to depict the “Liu clan's scripture” as a canonical text as would be typical in later periods (Nylan 2001, 16–51; Tsai 1992, 87–118).¹⁸ Within this context, the “Yaolüe” rather uses them in a verbal sense, meaning a process of weaving that the text was apparently thought to perform.

These two examples raise questions regarding the “Yaolüe's” choice of imagery and vocabulary. Why does it make these claims about the Huainanzi's weaving together at such prominent positions of the last chapter that, according to Martin Kern (2014) and the team of the Huainanzi Translation Project, was meant to function as a summary performed during the text's presentation to Emperor Wu? And what does it mean for a text to weave together? In addition to reading this expression as a metaphor for ordering processes, it might also refer to the Huainanzi's textual fabric, which led to its widespread dismissal as an encyclopedic behemoth in the first half of the twentieth century. In other words, I propose that one may better and more sympathetically explain the Huainanzi's intertextual design and writing practice through the (material cultural) context of a perceived parallel between writing and weaving during the early imperial period to which the “Yaolüe's” remarks seem to allude.¹⁹

The Imaginaire of Writing as Weaving in the Warring States and Early Imperial Period

Images of weaving play a prominent role in early Chinese conceptualizations of writing. Regarding the use of the term “pattern” (wen 文 or wenzhang 文章), which would take on the lexicalized meaning of a written text sometime around the late first century BCE, Kern (2001, 54) remarks that:

As Kern mentions in this passage, the term “wenzhang” that later texts often use to depict writings apparently has its origins in textiles and their ornamentation. He claims that such fabrics displaying various patterns and insignia were commonly used within a ritualistic and/or political context in early China, demarcating the various ranks of the people involved (Hinsch 2003; Powers 2006, 1–159). Thus “wen,” a term that had been predominantly used after the Eastern Han dynasty to collectively depict writings and culturality as a whole, originally may not have connoted any form of text at all (Kern 2001, 46).

Not surprisingly, then, we find exactly such an explication of wen as textual patterns in Liu Xi's 劉熙 (second or third century CE) Shiming 釋名 (Explaining terms), a slightly later lexicographical work than the sources Kern discusses in his article.²⁰ In that sense, Liu Xi affirms Kern's claim that sometime in the first century BCE “wen” would take on the lexicalized meaning of a text. Despite this lexicalization of “wen” as a text, however, Liu Xi (1983, 399b) still employed images derived from the process of weaving in his definition of “wen,” evoking the term's previous connotations as a textile pattern when he wrote in his Shiming:

Liu Xi's definition of “wen” explicitly considers the process of accumulating and linking “many characters” (zhong zi 眾字) into “phrases and meanings” (ciyi 辭義) as being equivalent to linking many “varicolored silk threads” (zhong cai 眾綵) into warp-faced “brocades and embroidery” (jinxiu 錦繡). Like weavers collecting strands of silk to warp them in a mounted loom so they may weave in weft threads, authors would assemble and connect bamboo slips in order to write words on them.²¹ In that sense, Liu Xi explicitly correlates the processes of weaving and writing, a fact that is clearly reflected in the prevalence of intertextual writing practices during the Six Dynasties 六朝 period (220–589 CE; Swartz 2018; Williams 2015, 23–49).

These observations raise two important questions: what happened in between the production of Liu Xi's Shiming, in which “wen” clearly refers to both written and woven patterns, and the earlier accounts Kern discusses, in which “wen” evidently signifies an ornamented fabric? And why did Liu Xi explicitly parallel the processes of writing and weaving in his definition of “wen“? In the following pages, I will lay out five reasons behind this semantic transformation or paradigmatic shift from “wen” as a fabric to “wen” as a text. In particular, I will focus on the shared verticality of the practices of writing and weaving in early China, the increasing utilization of silk as a writing material in the Han, the shared cosmic origin of the two technologies, the attribution of ordering properties to fabrics and texts, and finally some further crystallizations of this imaginaire in intellectual and material responses to texts during the Han.

The Shared Verticality of Weaving and Writing in Han China

The common mode of weaving during the late Warring States and early imperial period was the warp-faced compound tabby, a weaving technique unique to the Middle Kingdoms (zhongguo 中國) at that time that preserved its predominance until the Tang 唐 dynasty (618–907; Sheng 1995).²² The tabby is sometimes also called a plain weave since its threads move alternately one up and one down. Dorothy K. Burnham (1911–2004) defined this Chinese weaving technique during the late Warring States period and early Han dynasty as:

According to Burnham's depiction, the weaver of a warp-faced fabric produces a silk by laterally inserting the weft (wei 緯) into the (often varicolored) warp threads (jing). Contrary to the weft-faced textiles that were prevalent in any other part of the world (see figure 1), however, the warp-faced compound tabby generated patterns by foregrounding the vertical threads (see figure 2).

This process of vertically creating a weaving pattern by accentuating singular threads of the warp that run lengthwise across the loom with the help of the weft threads may be seen in the “Silk Figured Plain Weave” from the Han dynasty housed in the Philadelphia Museum of Art (see figure 3). By foregrounding the vertical threads, the warp-faced compound tabby effectively hides the wefts as a side effect of the fabric's patterning, emphasizing the warp as the centerpiece of the weaving process.

The importance of the warp thread for the production of early Chinese fabrics is indubitably reflected in traditional explanations of the term “jing.” As Xu Shen's 許慎 (ca. 58–147 CE) Shuowen jiezi 說文解字 (Commenting patterns and dissecting characters) from the late first or early second century CE claims, “the warp (jing) is weaving” (經：織也; Xu 1965, 9a.271b). This utilization of the warp as a synecdochical image for the entire weaving process is seemingly based on the fact that the vertical warp threads built the backbone of fabrics and were the very ground on which patterns (wen) appeared in early China.

Writing also displayed a vertical orientation in early China (see figure 4). This material cultural fact already manifested on bronzes and oracle bones from the Shang 商 dynasty (ca. 1600–1046 BCE), yet it is commonly associated with the material constraints provided by bamboo slips (Tsien 2004, 31–35, 96). Although Chinese may be written in almost any direction without losing its meaning, since characters build a self-contained semantic unit, the preferred direction of writing throughout the Warring States and early imperial period remained from the top down and from right to left.

Considering the materiality of most excavated manuscripts, it seems as if a typical text in the late Warring States and early imperial period was written on a scroll of interlinked bamboo slips or wooden boards. Writers generally applied ink directly on the surface of prepared bamboo slips or wooden plates that they wove together with two or more threads similar to a rattan fence, creating written patterns on their vertical surfaces. Hence, both writing and weaving shared a vertical orientation in the Western Han, an important aspect that clearly materialized in Liu Xi's later definition of patterns as the insertion of words and silk threads on a vertical axis.

The Increased Utilization of Silk as a Writing Material during the Han

In addition to a shared verticality, the increasing use of silk as a writing surface in the third and second centuries BCE contributed to the parallel construction of writing and weaving. During the pre-Han period, writers began to utilize silk fabrics to produce texts. However, it seems as if the production of silk was still so expensive up until the late Warring States period that fabrics were reserved for sacred imagetexts (Mitchell 1994, 83–110) such as the Chu boshu 楚帛書 (Chu silk manuscript; Li and Cook 1999; Tsien 2004, 130). This rare use of silk for texts, however, might have changed during the late Warring States period in the third and second centuries BCE when the Middle Kingdoms possibly developed advanced weaving technologies such as complex looms (see figure 5) that might have increased the productivity and in consequence probably decreased production costs (Kuhn 2012, 11).²⁵

According to Tsuen-Hsuin Tsien 錢存訓 (1910–2015), these developments might have led to a wider dissemination of silk as a medium for writings, as reflected in the Mawangdui boshu 馬王堆帛書 (Mawangdui silk manuscripts) from the early second century BCE that contain medical and historical writings, among others (Tsien 2004, 126–31). Consequently, Hans van Ess (2005–6, 255) assumes that “[a] more widespread use of silk or precursors of paper … must have taken place during the first half of the second century B.C.”

Even though there is very little material cultural evidence to substantiate Tsien's (2004, 130) claim of a significant increase in the utilization of silk as a writing material during the Warring States and early imperial period, a look in the imperial “Yiwen zhi” 藝文志 (Record of the literature [of artistic writings]) in Ban Gu's Han shu 漢書 (History of the [former] Han) supports his reading. In fact, it seems as if the wider dissemination of silk as a writing surface crystallized in the terminology that bibliographers at the imperial library used to catalog texts. While the “Yiwen zhi” reflects a clear separation between bamboo chapters (pian 篇) that appear 515 times and silk or paper scrolls (juan 卷) that appear 192 times (Ban 1958, 4:30.874a–915a), the “Jingji zhi” 經籍志 (Treatise on canonical and other texts) in the Sui shu 隋書 (History of the Sui) mainly uses the term “scroll” as a measure word for texts (pian = 98x and juan = 3,847x; Wei 1958, 21:32.468a–539b). This suggests that either the Sui 隋 dynasty (581–618) largely housed silk and/or paper scrolls in its imperial library or that the occurrence of such scrolls had become such a common phenomenon that there was less need to separate strictly these media in the bibliographical treatise anymore. In either case, the increasing appearance of the character “scrolls” (juan) rather than “chapters” (pian) as a measure word for texts in bibliographical sources during and after the Han dynasty supports Tsien's claim for a transformation of writing technologies in the early imperial period.

This apparent change from bamboo to silk as a writing medium continued until the third or fourth century CE, far beyond the early development of paper traditionally attributed to Cai Lun 蔡倫 (50–121 CE).²⁶ Hence, the Han seems to mark an important period in the development of writing materials in which the dominance of bamboo and other firm surfaces slowly waned and writers increasingly switched to softer, woven fabrics as their material of choice. As a result, writers apparently perceived the two materials as closely related media, as the wide usage of expressions such as “the bamboo and silk of writings” (shu zhi zhubo 書之竹帛), “transmit ordering principles on bamboo and silk” (jili yu zhubo 寄理於竹帛), and other forms of the synecdoche zhubo in texts such as the Mozi 墨子, Han Feizi 韓非子, Huainanzi, and Wang Chong's 王充 (ca. 27–100 CE) Lunheng 論衡 (Discourses weighed in balance) suggest.²⁷ In other words, writers of the Warring States period and Han dynasty associated and grouped together the practice of writing with bamboo slips and woven fabrics—an association that already establishes a firm bridge between the seemingly disparate aspects of wen as “fabric” and as “text” (see figures 6 and 7).

The Shared Cosmic Origin of Writing and Weaving in Early Chinese Texts

The existence of such an imaginaire of writing as weaving during the Han dynasty may be further substantiated by the fact that texts from the Warring States and early imperial period clearly construed writing and weaving as homological processes beyond their shared utilization as writing materials and vertical orientation. First of all, both were perceived to have cosmic origins. As Edward Schafer (1913–91) observed: “[t]he traditional Chinese vocabulary of basic features of the cosmos (and also of the political and social order of this world, which derives from it), is full of images from the language of threads, textiles, weaving, cords, and nets” (Schafer 1977, 262).²⁹ Such a cosmological and astronomical understanding of weaving, which seems to be mirrored in the mythological stories about the Weaving Maid (Zhinü 織女; Pankenier 2013, 364–82; 2015), appears prominently in Wang Yi's 王逸 (ca. 89–158 CE) “Jifu fu” 機婦賦 (Rhapsody on weaving women). This poetic piece, which Wang Yi—a native of the famous silk city Yicheng 宜城 in Hubei 湖北 Province who is known for his commentary to the Chu ci 楚辭 (Elegies of Chu)—might have written after visiting female workers at a weaving workshop, explicates the process of weaving as well as the creation and set up of a loom in cosmic terms (Kuhn 1995, 97–102; Wang Yi 1993; Zürn 2016, 248–51).³⁰ The rhapsody narrates a journey from the origin of the universe to the realm of the Myriad Beings in order to illustrate weaving and its accompanied tools as cosmic products that had been revealed to Fu Xi 伏羲 and subsequently his court ladies.³¹ In other words, Wang Yi seems to associate the production of fabrics and the setup and working of the drawloom, whose individual parts are depicted in zoomorphic terms,³² directly with the genesis and running of the entire universe.³³

Similar narratives about the cosmic origin of writings appear in several contemporaneous texts. The invention of the script and the trigrams were likewise perceived to be discoveries revealed to culture heroes. In fact, several texts mention that the sages Fu Xi and Cang Jie 倉頡 derived the trigrams and the writing system from cosmic patterns (Chow 1979). Such stories that connect the etiological myths of Fu Xi's discovery of the trigrams with Cang Jie's derivation of Chinese characters from tracks of birds and beasts appear in various texts, including the Huainanzi.³⁴ In these narratives, the sages developed images (xiang 象) and characters (zi 字) by closely observing and emulating the workings of the cosmos (Birrell 1993, 44–47; Boltz 1994, 130–38). In other words, these mythical narratives did not depict the two sages as inventors in the sense of a creatio ex nihilo that plays such an important role in a Christian and European cultural historical context (May 1978). Rather they created a scenario—similar to the case of weaving in Wang Yi's rhapsody—in which the Dao revealed the script through natural processes in Heaven and on Earth (tiandi 天地), “reducing” the sages to the role of transmitters of the Way's call (Mersmann 2011, 200).³⁵ Hence, the technology of writing was also thought to be a cosmic revelation to the sages, a vision of scriptures’ sacrality that clearly reappears in later Daoist movements such as the Heavenly Masters (Tianshi 天師), Highest Clarity (Shangqing 上清), and Numinous Treasures (Lingbao 靈寶; Bokenkamp 1997, 29–148, 373–438; Hsieh 2005; Raz 2012, 127–76).

The Ordering Powers Attributed to Writing and Weaving in Early Chinese Texts

In addition to the shared cosmic origin of writing and weaving, early Chinese texts attributed ordering powers to both practices as Chow Tse-tsung 周策縱 (1916–2007), Martin Kern, Mark Edward Lewis, David Pankenier, and Helwig Schmidt-Glintzer have already emphasized in previous publications. For example, it is apparent that several writings of the late Warring States and early imperial period construed the warp as the stabilizing element of a fabric, an idea derived from the central role of the warp in the weaving process as laid out above. As Lewis (1999, 297–98) remarks in his Writing and Authority in Early China:

In Lewis's understanding, early Chinese usages of the term “jing” apparently refer to a fabric's construction of well-ordered patterns via the warp's dividing lines.³⁶ Such an understanding of “jing” as an ordering procedure, for example, crystallized in the literary trope of weaving together or setting up camps (jingying 經營) in the songs “Beishan” 北山 (Northern hills; Kong 2001b, 444b; Waley 1996, 189); “He cao bu huang” 何草不黃 (What plant is not faded?; Kong 2001b, 527b; Waley 1996, 221); and “Jiang Han” 江漢 (The Jiang and the Han [rivers]; Kong 2001b, 685b; Waley 1996, 280) from the Shi jing 詩經 (Classic of songs). In fact, early Chinese texts clearly projected the warp's order-giving properties on the organization of the social and human body as exemplified by medical terminology such as vertical vessels (jingmai 經脈; Unschuld 2011, 1:14–15) or in Ji Jingjiang 季敬姜 of Lu's 魯 ekphrastic utilization of a loom to explain political organization in a narrative from Liu Xiang's 劉向 (77–6 BCE) Lienü zhuan 列女傳 (Biographies of exemplary women; Wang Zhaoyuan 1995–99, 669b; Kinney 2014, 13). Hence, the warp and its verticality became commonly associated with the process of ordering in early China, leading to an association of weaving's product—the (textual) pattern (wen)—with culturality and well-ordered cosmic, social, and physical bodies, in return. Consequently, this vision of weaving was also projected on spatial organization as reflected in HNZ 4.8's depiction of “the terrestrial forms as East/West being the weft (wei) and [the royal] South/North being the warp (jing)” (凡地形，東西為緯，南北為經; Zhang Shuangdi 1997, 451).³⁷

The same may be said for writings. As Helwig Schmidt-Glintzer summarized this tendency of attributing ordering powers to texts:

Schmidt-Glintzer claims that writers of the Middle Kingdoms perceived texts to be correspondents to natural orders since writings originated from signs and patterns through which the world expresses its own workings (J. Liu 1975, 16–62). Writing and literature, in that sense, were not just a way to communicate information to another person. Given writing and literature's origin in the orderly functioning of the world, they were thought to be imbued with the power to “impact the orderliness of the world” as also expressed in the “Da xu” 大序 (Great preface) to the Shi jing (Kong 2001b, 12b; Owen 1992, 39) or the medicinal use of writings in early China.³⁸ Accordingly, they were considered to be elementary aspects of governance in early China as reflected in Cao Pi's 曹丕 (187–262 CE) expression that “writings are the great business of weaving together a kingdom” (蓋文章經國之大業; Liu, Chen, and Ho 2000, 68). In summary, early Chinese texts apparently construed writing and weaving's etiologies, functions, and forms in homological ways.

Further Crystallizations of the Imaginaire of Writing as Weaving in Early China

These parallel constructions of writing and weaving in Han texts culminated in the specific coinages of terminologies for textual genres during this period. Terms such as the aforementioned textual patterns (wen); canonical texts (jing, lit. warp) and their commentaries (wei 緯, lit. weft); “weft-writings” (weishu 緯書 or chenwei 讖緯); and many more reflect the existence of such an imaginaire of writing as weaving.³⁹ In addition, the imperial court commissioned the first large-scale projects to recover seemingly lost texts during the Western Han dynasty. As Tsuen-Hsuin Tsien (2004, 14) remarked:

These editorial projects commissioned by the imperial court apparently faced the problem of sometimes literally weaving together comprehensive versions of texts out of remnants of possibly disjointed bamboo slips due to the vast destruction during the Warring States and the beginning of the imperial period. This interest in collecting and controlling the textual past during Emperor Wu's reign further manifested in the construction of imperial libraries and the later creation of the imperial “Yiwen zhi” under Ban Gu (Fölster 2018; Wu Kuang-Ts'ing 1937). In that sense, it is quite possible that the process of recollecting and assembling texts from the Zhou 周 dynasty (ca. 1046–256 BCE), as well as the practical weaving together of bamboo scrolls out of individual slips, further evoked associations to the process of weaving. All these points suggest that there existed an imaginaire and practice of writing as weaving during the second or first century BCE that informed the transition from wen as a fabric to its lexicalized meaning as a text, as observed by Kern—albeit I push the date a bit earlier—and as later explicated by Liu Xi. In other words, it seems as if weaving and writing were perceived to be homological processes in the early imperial period.⁴⁰

The Huainanzi's Design as a Textual Fabric

After this little detour via the imaginaire of writing as weaving, I will now return to the Huainanzi's self-depictions with which I started this article. As I have shown above, the “Yaolüe” clearly illustrated the Huainanzi as a weaver and/or a woven text. In the following, I provide further evidence that the “Liu clan's scripture” in fact resembles a textual fabric by visualizing its technique to array and weave in textual fragments from the past and present. By observing the Huainanzi's design through the lens of the perceived parallel between the practices of weaving and writing during the Han, I propose that the “Yaolüe's” self-depictions—not unlike Liu Xi's definition of patterned writings as brocades of words—in fact refer to the text's intertextual writing practice, which I present now in the vertical orientation characteristic of premodern China. To do so, I will showcase some eye-opening passages from Liu An's text.

As demonstrated in the section on the Huainanzi's self-illustrations above, the “Taizu” (The exalted lineage) chapter apparently “weaves together the paths/teachings of the past and present” in order to “weave … together the governing ways and knot … together the kingly affairs.”⁴¹ As Queen and Major mention in their introduction to chapter 20, “Taizu” literally lines up a “number of exemplary rulers of the recent and remote past who embodied these [i.e., the Huainanzi's] ideals and thereby brought order and harmony to the wider world” (Major et al. 2010, 789). In other words, chapter 20 “restates and brings into sharper focus the lessons of the preceding nineteen chapters” (790) that—according to Liu An and his courtiers—are necessary to order all under Heaven (tianxia 天下) by construing such an “exalted lineage.” However, note that Liu An and his retainers did not deliver this “message” simply in plain speech. Instead, they chose to create a textual fabric that presents this “exalted lineage” by weaving together passages from a variety of pre-Han classics.

In a passage (HNZ 20.4–5) from the beginning of chapter 20 (see figure 8), which directly addresses how the Way impacts the universe via nonintentional actions (wuwei 無為), there is an interesting combination of snippets from writings that during the Han dynasty would be categorized as canonical (jing) texts—the Shi jing and Yi jing 易經 (Classic of changes)—and two Warring States master texts (zishu 子書)—the Zhuangzi 莊子 and Han Feizi (Denecke 2010, 1–31). While the textual parallels with the Zhuangzi and Han Feizi remain unmarked in the Huainanzi (the marking of a parallel with the Liezi 列子 is already part of the Han Feizi passage),⁴³ the “Liu clan's scripture” signifies the quotations from the Shi jing with the help of a “The [Classic of] Songs says” (Shi yun 詩云), clearly treating the various textual parallels in distinct and seemingly conscious ways.⁴⁴

This technique of arraying passages from pre-Han classics permeates the entire chapter. However, “Taizu” does not simply utilize pre-Han texts. It literally “weaves together the teachings (dao) of the past and present” by inserting passages from both the classics and the Huainanzi's previous nineteen chapters as displayed in figure 9 (HNZ 20.10).⁴⁵ Clearly, the Huainanzi includes here both self-references to several chapters (HNZ 9.11, 13.2, 15.1, 17.225, and 18.6) and passages shared with the Liji 禮記 (Record of rites), Shenzi 慎子, and Lüshi chunqiu. Examples like this suggest that Liu An and his erudite courtiers consciously assembled chapter 20 out of pre-Han classics and the Huainanzi's own writing (Major et al. 2010, 792). In other words, the “Taizu” chapter's design seems to reflect its self-depiction as a textual fabric in the “Yaolüe.” Rather than solely referring to the text's conceptual tapestry, as suggested by Queen and Major (Major et al. 2010, 794), it apparently also illustrates the Huainanzi's intertextual writing strategy that assembles various teachings from the Han and pre-Han classics and “returns them home to a single root [i.e., the Huainanzi].”⁴⁷

This pattern of weaving textual threads from prior sources into the fabric of the Huainanzi is a staple of the entire scripture. As Michael Nylan (2014, 260) rightfully mentions about the “Jingshen” 精神 (Essence and spirits) chapter, “long—occasionally very long—swaths of text are inserted into the chapter with a bare minimum of particles.”⁴⁸ As figure 10 shows, the beginning of the “Jingshen” chapter (HNZ 7.1–2) includes textual parallels or, to use Liu Xi's terminology from above, it “assembles and collects many varicolored [textual] silk threads” (huiji zhongcai 會集眾綵) that belong to a variety of different traditions.⁴⁹ According to Sima Tan's model of the Hundred Schools, these phrases would be attributed to the Confucian or Ruist (Xunzi); Daoist (Laozi, Zhuangzi, and Liezi); Legalist (Guanzi 管子); and Miscellaneous (Lüshi chunqiu) schools, as well as historical writings such as the Guo yu 國語 (Discourses of the states). In that sense, the “Liu clan's scripture” does indeed connect and relate teachings from various writings of the pre-Han period, weaving their words together into a unitary whole. In fact, Liu An and his courtiers might have even tampered with the elaborate parallel to the Liezi.⁵⁰ In other words, it becomes increasingly apparent that the text literally weaves textual threads from the past of varying complexity and length into the Huainanzi, seemingly reflecting and materializing the “Yaolüe's” self-illustration as a textual fabric.

Based on the examples provided by Le Blanc, the Huainanzi Translation Project, and this article, I may, therefore, tentatively propose that such an intertextual writing strategy is a truly prominent feature of the text.⁵² The writings of Liu An and his courtiers frequently incorporate textual threads from earlier sources without marking them with any particles or explanations, replicating the hiding of weft threads in a warp-faced compound tabby that is so typical of early Chinese woven textiles (see figures 1 and 3). If one considers that “[o]ver 80 per cent of the [“Yiwen zhi”] catalogue's 631 entries are only known by their title” (Fölster 2018, 206), indicating that a lot of pre-Han texts are lost, it seems very likely that the Huainanzi weaves even more textual threads of the past in its textual fabric than one may discern today (Kern 2005, 295). Even though one can only retrace an admittedly reconstructed speckle of the Huainanzi's intertextual design, such a philological analysis of the Huainanzi and its parallels with the extant versions of pre-Han texts nonetheless provides at least a taste of how rigidly intertextual Liu An and his courtiers might have designed and woven together the “Liu clan's scripture.”

Conclusion: Is The Huainanzi a Powerful Scripture That Is in the Image of and Embodies the Way?

As I have shown in this article, the intertextual writing procedure utilized in the Huainanzi apparently resembles or even mimics the process of weaving in early China. It seems as if Liu An and his workshop integrated textual threads, which recorded the teachings (dao), words (yan 言), and deeds (shi 事) from the pre-Han period in the Huainanzi. In so doing, the “Liu clan's scripture” fashions itself as a complex, textual fabric, as explicitly mentioned in the “Yaolüe” chapter. In that sense, the Huainanzi apparently enshrines or at least reflects in its intertextual writing practice and design what I have called an imaginaire of writing as weaving that was prominent during the Han. Hence, it is likely that the producers of the Huainanzi did not just perceive weaving to be a metaphor of writing. The image of weaving rather materialized and was embodied in the text's design, transcending the scope of figurative speech.⁵³

This observation raises an important question: why would Liu An be interested in commissioning such a textual fabric that weaves in the textual threads from the past and present? In the remainder of this article, I will provide a brief glimpse into a potential reason behind Liu An and his courtiers’ production of such an extraordinarily intertextual text. A short analysis of the contexts in which the Huainanzi usually utilizes the images of weaving (jingwei) may shed some light on the potential reasoning behind the text's implementation of intertextual references.

Throughout the Huainanzi, there are several passages that employ an imagery of weaving to depict the world and the Dao. For example, the second chapter states:

Apparently, HNZ 2.4 correlates the woven fabric of the cosmos with the basic organization of human society. In the first part, it illustrates the Way as the one who holds and links the warp and weft of the universe together. In other words, it illustrates the Way as a cosmic weaver that creates Celestial Patterns (tianwen 天文) and Terrestrial Forms (dixing 地形) while connecting and “link[ing] the thousand branches and myriad leaves.”⁵⁵ In the second part, the text then projects this vision of the universe as being woven together by the cosmic Way onto the social structure of the human world. Within this vision, each silk thread would stand for a social rank—the noble ones (gui 貴), the lowly ones (jian 賤), the poor ones (pin 貧), and the hard-pressed ones (kun 困)—that, woven together as a whole, would make up the social fabric. It seems as if HNZ 2.4 suggests that as much as a fabric necessitates varicolored threads to spawn a pattern (wen), a well-ordered society needs each societal element to fulfill its role (fen 分) in order to function harmoniously (he 和). In fact, such a vision of social organization based on the utilization of the various people (yong zhong ren 用眾人) is prevalent throughout the Huainanzi (Ames 1994, 56–65, 142–52; Goldin 2005, 96). Hence, this passage explicitly depicts the Way as a weaver of cosmic order while construing social organization and hierarchies as manifestations of this cosmic fabric in the human world.

The same weaving imagery is repeated several times throughout Liu An's text to depict the Way and the sage. In fact, it seems as if HNZ 2.10, 3.16, 4.1, and 8.7 deliberately utilize this imagery to parallel the organizing powers behind the universe and a sage empire (Zhang Shuangdi 1997, 198, 295, 417, 849; Major et al. 2010, 99, 122–23, 154, 277). According to this vision, the sage—like the Way—establishes herself or himself in the center and actualizes through her or his corpus hidden powers (yinde 陰德) that may “magically” attune the phenomenal world, organize all under Heaven, and connect all Myriad Beings (Zürn 2018, 321–28). In other words, it seems as if Liu An's text commonly associates the processes of weaving with the concept and practice of wuwei and the connective powers of the Way and the sage via resonating correspondences (Zürn 2016, 207–24).

This paralleling between the Way and sages’ weaving, a homological relationship also replicated in the example of the tree (benmo), as shown by the Huainanzi Translation Project and the example from HNZ 2.4 above, might have significant consequences for our understanding of the Huainanzi‘s self-referential depictions and construction as a textual fabric. In addition to the image of weaving and the tree, the “Liu clan's scripture” construes the three entities in homological terms in several other instances and, therewith, creates a strong link between the Way, the sage, and the Huainanzi. For example, it also utilizes the images of a chariot wheel's hub (gu 轂) or axle (zhu 軸) and the forge (ye 冶), common illustrations of the Way and its powers, to create such homological relationships (Zürn 2016, 142–45, 158–206, 342–51). Accordingly, the Huainanzi's intertextual writing strategy was apparently not a failed attempt to create a philosophical text, as suggested by Wing-tsit Chan and the early Fung Yu-lan. Rather, the text connects the various threads of pre-Han teachings, words, and deeds into a comprehensive Way and, therefore, could be read as a literary attempt to fashion the “Liu clan's scripture” in the image of (xiang [yu] 象[於]) and as embodying the Dao (tidao). Or simply put in terms similar to those of Puett (2014), who has argued that the Huainanzi “serves as a final sage” (287): Liu An and his workshop might have written the Huainanzi in such an intertextual fashion in order to create a divinized scripture (Puett 2002, 80–121, 259–86). Like the sage who performs wuwei, the “Liu clan's scripture” would function as a textual embodiment of the Way that would be able “to resonate inexhaustibly with [all] beings” (yingwu wuqiong 應物無窮; HNZ 9.10 in Zhang Shuangdi 1997, 912; Major et al. 2010, 304) and their various categories (lei 類) by connecting and relating them exemplarily within the confines of the text (Le Blanc 1985, 191–206).⁵⁶

Since the sage ruler is supposed to mainly organize the universe without acting (wuwei), speaking (bu yan 不言), and teaching (bu jiao 不教), I therefore propose that the “Liu clan's scripture” might have been created not just to educate young Emperor Wu about how to rule the empire.⁵⁷ According to my analysis, it is possible that the Huainanzi was also meant to function as an efficacious, textual artifact in the tradition of the mystical Hetu 河圖 (River chart) and Luoshu 洛書 (Luo script) that could be utilized by the imperial Liu clan to ritualistically unite and order all under Heaven.⁵⁸ Placed in the imperial library at Chang'an, the alleged religiopolitical center of the universe during the Western Han, the Huainanzi—like the sage—might have been produced to create resonating correspondences with the entire world by performing wuwei.⁵⁹ To use a musical image popular in the Huainanzi and other early Chinese texts, the “Liu clan's scripture” would function like a universal string. If played and activated, it would have the effect that all strings in the universe would resound harmoniously (Rom 2017, 145–51).

If my speculation is reasonable, then we might also want to reconsider our approach to intertextual writing practices in early China. So far, scholars in the humanities tend to emphasize discursive functions for texts and their intertextual references, focusing on their argumentative, rhetorical, authoritative, and epistemic properties.⁶⁰ The Huainanzi's extraordinary design, textual formalization, and self-references, however, seem to point toward function(s) beyond such discursive readings. They suggest that intertextual writing practices might also be related to the production of ritualized, powerful scriptures in the Han dynasty that textually mimic the ordering faculties of the sages and the force behind the motions of the universe. In other words, my alternative interpretation of the Huainanzi as a wuwei-performing, textual embodiment of the Way suggests that visions and productions of efficacious writings might have already existed in the second century BCE—long before the consolidation of Daoism and the arrival of the Buddhist cult of the book during the late Han and early Six Dynasties.⁶¹

Acknowledgments

I would like to thank in alphabetical order the following scholars and institutions for their feedback and support that made this project possible: Cai Yuqian, Mark Csikszentmihalyi, Du Heng, Markus Greulich, Anne Hansen, Natasha Heller, Rania Huntington, Li Yuhang, Mark Meulenbeld, Julia K. Murray, Michael Naparstek, William Nienhauser Jr., Michael Puett, Gil Raz, Dennis Schilling, Michael Stanley-Baker, Mark Valeri, Nathan Vedal, Wong Tsung Kei, Stuart Young, the four anonymous reviewers of the Journal of Asian Studies, and the audiences at the workshop “Empire and the Media of Religion” (UCLA, 2015) and at the annual conferences of the American Oriental Society in Boston (2016) and the American Academy of Religion in Denver (2018). Lastly, I want to express my gratitude to the Religious Studies Program and the Department of East Asian Languages and Cultures at Washington University in St. Louis for their financial and intellectual support.

Notes

List of References