Transcribing the Sacred in the Printing Era: A Study of Handwritten Buddhist Canon during the Northern Song Dynasty

Abstract

In an era marked by the advent of advanced printing technology during the Northern Song period, the tradition of transcribing the Buddhist canon endured rather than promptly fading away. This persistence is particularly noteworthy in the 10th to early 12th centuries when handwritten copies of the canon coexisted with the emerging prevalence of block-printed versions. Notably, the monasteries in the Jiangnan 江南 region remained prolific producers of handwritten copies. This article adopts a socio-political perspective to elucidate the coexistence of handwritten and block-printed copies. It delves into the dynamics of canon replication within a specific temporal and geographical context, unveiling a broader tableau of handwritten copies in the printing era. This exploration necessitates an examination of the intricate interplay between the state, the canon, and Jiangnan society. This study contends that the state’s printing policies before the Xining 熙寧 period (1068–1077), which rigorously regulated the production and dissemination of canonical texts, constituted a pivotal factor in sustaining the tradition of handwritten copies. Furthermore, it identifies a sustained societal demand for possessing the canon for constructing the zhuanlun zang 轉輪藏 (revolving wheel storage cabinet), which was most pronounced in the Jiangnan region. This persistent societal need fortified the enthusiasm for transcribing the canon within Jiangnan monastic communities.

1. Introduction

The Northern Song dynasty (960–1127) was pivotal in the history of the Chinese Buddhist canon as it ushered in a flourishing block-printing period, marking the creation of the first woodblock-printed canon, the Kaibao Canon 開寶藏, in 983. However, despite the availability of convenient and advanced printing technology, the tradition of transcribing the canon did not soon disappear. Instead, handwritten copies of the Buddhist canon continued to exist alongside the block-printed versions and remained influential. It was not until the early 12th century that block-printed editions eventually replaced handwritten copies and emerged as the primary method of producing and distributing the Chinese Buddhist canon (Fang 2006, p. 24). Therefore, there was a lengthy period from the 10th to the early 12th centuries when manuscript and block-printed copies coexisted, which is particularly significant and instrumental in revealing the transitional history from the manuscript era to the printing era and the manuscript culture of the Song period.

Nevertheless, this coexisting period has been given short shrift by scholars of both Buddhist canon and manuscript studies. Current scholarship on the Song Buddhist canon predominantly concentrates on the five woodblock versions, whereas manuscript scholars primarily delve into the Dunhuang cave manuscripts. Undoubtedly, the Dunhuang manuscripts mark a magnificent chapter in the history of handwritten Buddhist scriptures; however, they do not represent the complete history. Countless manuscript copies of the canon were produced during the Song, with only a few surviving and many others only retrievable in historical records. The Song manuscript canon, which represents the last splendor of the Chinese manuscript era, has been largely overlooked by scholars.

Recently, scholars have taken the initiative to comb through the remaining Song manuscript copies of the canon. Zhang Xu and Feng Guodong’s work offers an exceptional account, presenting a crucial description of the extant Song manuscripts from five different handwritten versions of the canon and a few unknown versions (Zhang and Feng 2021, pp. 170–75). Zhang and Feng have not delved into an in-depth analysis; however, their research paves the way for further investigation of the Song handwritten canon. In addition, Cai Qiuxia’s M.A. thesis focused on the Jinsu shan Canon to explore the variant Chinese character and the calligraphic styles of handwritten canon during the Song period, which is particularly informative in presenting all extant manuscripts of the Jinsu shan Canon preserved in the Shanghai Library (Cai 2010, pp. III–V). Cao Ganghua’s research on the circulation of the canon in the folk society during the Song Dynasty also includes a discussion on the handwritten copies, which sheds light on the printing policy in the early Song period and is particularly conducive to viewing canon production along with state power (Cao 2006, pp. 169–70).

Closely reading these remaining manuscripts and those documented in historical records might help decipher one of the most crucial inquiries regarding the Song period’s handwritten copies of the canon. Namely, why did copious handwritten copies continue to be produced, particularly in the Jiangnan 江南 region, in a period when printing techniques were mature and available? Cai speculates on the factors of cost and state control without digging into this (Cai 2010, p. 16). Cao argues that the state control over the canon was relaxed under Huizong 徽宗, which is subject to scrutiny and debate (Cao 2006, p. 170). Fang aptly periodizes the stage of the coexistence of manuscript and block-printed copies in the history of the Buddhist canon; however, he has not explained this phenomenon, only depicting that, in the early days of the wood-printed canon, circulation was not yet widespread; thus, handwritten copies still dominated (Fang 2006, p. 24). Given the preferable circumstances that the private book engraving business had been in full swing since the early Northern Song period, it is worth further considering why the circulation of the printed canon was not widespread and why the handwritten copies remained popular. This inquiry might help reveal a broader picture of handwritten copies in the printing era, which entails the exploration of the relationship between the canon and the state, and between the canon and the Jiangnan society.

In this article, I elucidate the stage of the coexistence of handwritten and block-printed copies from a socio-political perspective. I argue that the state printing policy before the Xining 熙寧 period (1068–1077) that rigorously controlled the printing and the purchase of imprints of canonical texts was the direct factor accounting for the continuous existence of handwritten copies. Moreover, the lasting social demand for possessing the canon to construct the zhuanlun zang 轉輪藏 (revolving wheel storage cabinet), especially in the Jiangnan region, is another driving force that supported the passion for transcribing the canon in the Jiangnan monasteries.

2. A Sketch of the Song Handwritten Copies of the Canon

The current scholarship recognizes five known versions of the canon that were handwritten during the Song Dynasty: the Jinsu shan Canon 金粟山藏經, the Faxi si Canon 法喜寺藏經, the Haihui yuan Canon 海惠院藏經, the Chongming si Canon 崇明寺藏經, and the Jingde si Canon 景德寺藏經.1 Additionally, a few manuscripts from unknown versions have also been discovered. In

Table 1. Song handwritten versions of the canon.

Version	Title of Manuscript	Date	Place of Production	Place of Preservation
Jinsu shan Canon	Jietuo dao lun解脫道論 (vol. 1), marked with the character bei 背	1068	Haiyan 海鹽縣, in today’s Zhejiang 浙江	Shanghai Library
Faxi si Canon	Da bore boluomiduo jing 大般若波羅蜜多經 (vol. 89)	1077	Haiyan 海鹽縣, in today’s Zhejiang 浙江	National Library of China
Haihui yuan Canon	Fayuan zhulin 法苑珠林 (vol. 65), marked with the character pei 陪	1064	Huating 華亭縣, in today’s Shanghai 上海	Shanghai Library
Chongming si Canon	Shuo yiqie youbu pinlei zu lun 説一切有部品類足論 (vol. 11), marked with the character tou 投	1090	Jurong 句容縣, in today’s Jiangsu江蘇	Shanghai Library
Jingde si Canon	Mohe bore boluomiduo jing摩訶般若波羅蜜多經 (vol. 34), marked with the character he 河	1067	Kunshan 昆山縣, in today’s Jiangsu江蘇	Tianjin Museum
Unknown	Za api tanxin lun 雜阿毗曇心論 (vol. 11)	1055	Unknown	Shanxi Museum
Unknown	Wenshu jing文殊經	1069	Unknown	Museum of Fine Arts, Boston

, I provide essential information concerning the Song handwritten versions of the canon, with each version represented by a manuscript with a known date.2

The date and place of the production of the Song handwritten copies are noteworthy. Almost all existing manuscripts were produced before or during the Xining reign (1068–1077) of the Northern Song, except for those from the Chongming si Canon. Furthermore, the five known versions were exclusively produced by monasteries in the Jiangnan area, with three of them originating from the same prefecture—Xiuzhou 秀州 (located in today’s Zhejiang and Shanghai). It is also noteworthy that many unknown versions with no surviving manuscripts that are yet documented in historical records mostly follow the same pattern of date and place.

Zhang Yanchang 張燕昌 (1738–1814), a Qing scholar who specialized in collecting and studying the Song handwritten canon, recorded several manuscripts of the Da bore jing 大般若經 from the canon produced in the Jingyan monastery 精嚴寺 in Xiuzhou in 1055, which he saw from his friend’s collections.3 Song scholar Chen Shunyu 陳舜俞 (960–1127) alone authored three commemorative records for three handwritten canon collections produced in monasteries in Xiuzhou. In addition to the Haihui yuan jingzang ji 海惠院經藏記 (Record of the Haihui yuan Canon), he composed a record in 1054 for the canon produced by the Zisheng monastery 資圣禅院 and another in 1061 for the version made by the Bujin monastery 布金院, indicating that these two versions probably predate the Haihui yuan Canon.4 Another Song scholar, Lin Lu 林露, detailed the start and end of the handwritten Yongming si Canon 永明寺藏經 in Cixi 慈溪 (in today’s Zhejiang): “The canon copying started during the Jiayou 嘉佑 reign (1056–1063) and finished in the years of Zhiping 治平 (1064–1067) 經起於嘉佑而畢于治平.”5 There are many other accounts of the handwritten copies of the canon scattered throughout stone inscriptions, monastic and local gazetteers, and literati’s anthologies, revealing that handwritten copies of the canon boomed in large numbers in the Northern Song period despite the scarcity of manuscripts that survived to today. Moreover, these accounts evince that the starting date of most handwritten copies was before or during the Xining reign and organized mainly by monasteries in Jiangnan or nearby regions, such as Jiangxi 江西 and Anhui 安徽.

The surge of copying the canon in a specific period and region is intriguing. What accounts for this phenomenon? It behooves us to investigate beyond the manuscripts to explore the context of manuscript production—the historical circumstances under which these handwritten copies of the canon were produced, disseminated, and used.

3. State Printing Policy and the Production of the Canon

We should first ponder state politics’ influence on the canon’s production. Throughout Chinese history, state power has exerted tremendous control over the Buddhist canon, and such intervention has only increased over time (Fang 2006, p. 12). The coexistence of handwritten and block-printed copies for over a century directly resulted from the Song imperial government’s intervention in producing and disseminating the canon, embodied by the printing policy of the time, which aimed at strengthening the state regulation and control of canonical texts.

During the early Northern Song period, a state policy that favored literature (wen文) over martial (wu 武) virtues led to a surge in the large-scale, official-sponsored printing of canonical literature. The creation of the Kaibao Canon was part of the endeavor. In addition to preserving cultural traditions and ensuring social and cultural consistency, early Song emperors regarded printing as a significant vehicle for establishing government authority over the canonical texts, particularly Confucian classics utilized in the civil service exams (Qian 1992, pp. 234–38). As Cherniack aptly points out, the Guozi jian 國子監 (Directorate of Education) became the hub for controlling printing and undergirding state textual authority in the early Song period (Cherniack 1994, pp. 40–42). Correspondingly, strict regulation was imposed to protect the central government’s monopoly over printing canonical texts. According to the Song huiyao jigao 宋會要輯稿 (Song Government Manuscript Compendium), an edict issued in 998 even banned all unauthorized private reprinting of classics.6 Furthermore, this policy was implemented scrupulously by successive emperors. The Song scholar Luo Bi’s 羅璧 miscellany reveals that unauthorized printing was still strictly prohibited before the Zhiping 治平 reign (1064–1067).7 Therefore, unless bestowed by the Song court, the government imprints could be acquired only by application to the Guozi jian at a high cost.

Scholars such as Qian and Cherniack have shed light on the state’s control over the printing and reprinting of the Confucian classics. It is crucial to note that, during the early Song period, the production of the Buddhist canon was situated within the same policy and legal framework, and, thus, subject to state control and scrutiny. During this era, the imperial government wielded exclusive textual authority over the Buddhist canon, meticulously overseeing every stage of its production from translation to printing. As Sen observes, a defining feature of the Song Buddhist translation projects was the complete centralization of the procedure under state guidance. In the evaluation of Buddhist text authenticity, the state played a more substantial role than influential members of the Buddhist community, deviating from traditional norms wherein religious authorities held primary influence (Sen 2002, p. 37). This increased state intervention is evidenced by the appointment of high officials as the yijing runwen guan 譯經潤文官 (Officials of Translation and Embellishment), tasked with overseeing the translation project to the minor details (Wu and Chia 2015, p. 150). In addition to translation, the printing of the Buddhist canon was under the firm control of the state. A record in the Fozu tongji 佛祖統紀 (Chronicle of the Buddhas and Patriarchs) indicates that, in 983, the imperial government established the yinjing yuan 印經院 (Institute of Printing the Buddhist Scriptures) at the Taiping xingguo monastery 太平興國寺 to house “wood block plates of the Buddhist canon produced in Chengdu by order of Taizu” 成都先奉太祖勅造大藏經板,8 which referred to the Kaibao Canon. The yinjing yuan thus served as a government agency responsible for preserving the Kaibao Canon plates and managing the printing and circulation of the canon. Therefore, monasteries during this period were deprived of the right to reprint the canon unless they obtained imperial authorization, let alone to produce a new version of the woodblock canon. Yang Yi’s 楊億 (974–1020) memorial writing composed for the newly constructed Great Canon Tower of the Kaiyuan monastery 開元寺 in Wuzhou 婺州 (in today’s Zhejiang) corroborates the imperial government’s stringent control over printing the canon:

“Monk Wenjing from the Kaiyuan Monastery shared a common goal with Cao Weixu, a military officer and commander of the prefecture. Together, they worked towards their success in reprinting the canon. During the Chunhua reign (990–995), they went to the imperial court and beat the dengwen drum (drum for petitioning the court), requesting to borrow the original plates to make copies. When the emperor learned of their request through a memorial, he granted it. It was only at the beginning of the Zhidao era (995–998) that Weixu and others brought the payment for paper and ink by carriage to the capital. The emperor exempted them from paying tolls and bestowed them with a deed. After the printing was completed, they escorted the woodblock canon back home.

乃有本寺僧文靖，與本州 都知兵馬使曹維旭，同發誌誠，共營勝利。爰以淳化中相率詣闕，擊登聞鼔，求方借版摹印真文。奏牘上聞，帝俞其請。逮至道初，維旭等始共輦置楮墨之直，聿來京都。詔免關市之征，授以要券，繕造既畢，護持而歸。”9

Yang Yi’s account is remarkable in revealing the emperor’s direct control over authorizing the reprinting of the Buddhist canon. The intricate and strict process of petitioning in the capital and receiving imperial approval resembled a display of etiquette or the enactment of canon-related rituals, underscoring the sanctity of the Buddhist canon. On the other hand, it demonstrated that the canon’s sacred status was ultimately solidified only through the emperor’s endorsement. It is conceivable that the demanding procedure of obtaining imperial authorization may have prevented most monasteries from pursuing a printed canon. Consequently, only a handful of monasteries persevered through the arduous procedure to procure the printed canon (Cao 2006, p. 170). For the majority of monastic institutions, producing a handwritten copy of the canon emerged as a more practical and feasible alternative.

However, this situation eventually changed. While the exact year is unknown, a significant shift occurred in the printing control regime of the Song Dynasty during the Xining period (1068–1077), during which the imperial government relinquished its monopoly over the printing of canonical texts and eased its strict control over the publishing market. Luo Bi recorded that only during the Xining period was this ban completely relaxed, and, therefore, scholars born later were fortunate indeed 熙宁后方尽弛此禁。然则士生于后者，何其幸也.10 Luo’s account referred primarily to the Confucian classics. Nevertheless, the Xining period was also a turning point for the printing of the Buddhist canon, as recorded in the Southern Song encyclopedia Yuhai 玉海 (Sea of Jade), which shows that, in the third month of the fourth year of the Xining era (1071), an edict was issued ordering the Institute of Printing the Buddhist Scriptures to be rescinded.11

Moreover, the Song huiyao jigao recorded that, after the institute was abolished, the woodblock plates of the Kaibao Canon were entrusted to the Xiansheng monastery 顯聖寺 in 1071.12 This significant move streamlined the process for monasteries to acquire a printed copy of the canon, eliminating the need for complex procedures, such as petitioning the emperor. Instead, monasteries could conveniently obtain a copy from the Xiansheng monastery, which subsequently became a central hub for printing and distributing the woodblock canon. Surprisingly, despite the availability of the printed canon, there was no substantial surge in purchases by monasteries. The expense of purchasing a printed canon might not have been enticing enough to attract monasteries, especially those in southern China, to undertake the journey solely for a printed canon set. Some historical accounts indicate that the expenses of procuring a woodblock-printed canon from the Xiansheng monastery encompassed not just printing costs but also various additional fees, interaction costs, and shipping expenses, as evident in Japanese Tendai monk Jōjin’s 成尋 (1011–1081) travel diary, San Tendai Godai san ki 參天臺五臺山記 (A Record of a Pilgrimage to Mt. Tiantai and Mt. Wutai):

“On the thirteenth day of the fourth month of the sixth year of the Xining reign (1073), the canon arrived and was counted by scroll number. Money was paid to officials: three hundred wen for one of them, two hundred for another, and one hundred for each of the remaining four people. Besides, each of the eight porters was paid fifty wen. However, all of them claimed that it was not enough.

煕寧六年四月十三日三藏來坐，共依字號計卷軸預納了。官人等祿錢，一人三百文，次一人二百文，次四人各一百文，擔夫八人各五十文。各皆稱不足由。”13

Wealthy monasteries in the prosperous Jiangnan region may have found it practical to create their own plates of the canon, given the long distance and unexpected costs involved in purchasing a printed copy. Furthermore, with the imperial government relinquishing its monopoly on printing the canon, the manufacture of plates and the subsequent reprinting became a highly profitable business and one of the most significant sources of income for the monastery. Thus, a few capable monasteries in the Jiangnan region came to the fore, organizing the creation of new versions of the woodblock canon one after another. In 1080, only three years after the conclusion of the Xining reign, the Chongning Canon 崇寧藏 began to be carved at the Dongchan monastery 東禪寺 in Fuzhou 福州, marking the beginning of privately produced woodblock canon in China and ushering in a wave of monastery-produced woodblock canon. The emergence and flourishing of privately produced woodblock canon happened almost a century after the completion of the first woodblock copy, the Kaibao Canon. This time interval is also roughly equivalent to the coexistence of handwritten and block-printed copies, indicating that the state’s control over the production of canonical texts before the Xining reign directly contributed to the continued existence of handwritten copies. The Xining shift in printing policy profoundly affected the production and dissemination of the canon, paving the way for the complete replacement of the manuscript canon by block-printed editions.

The aforementioned analysis sheds light on the notion that the merit associated with handwriting sutras may not have been the primary driving force behind the continued transcription of the canon during the Northern Song period. Some scholars, including Cai, draw parallels between the merit-oriented, sutra-copying trend observed in Dunhuang Buddhist manuscripts and the Northern Song manuscript canon, suggesting that the fervor for transcribing the canon in the early Northern Song era was primarily fueled by devotion and the pursuit of merit (Cai 2010, pp. 15–16). Undoubtedly, the abundance of handwritten sutras found in Dunhuang resulted from the desire to accumulate merit, as printing technology was not yet available or becoming commonplace. However, as the early Northern Song dynasty witnessed significant advancements and the widespread adoption of printing techniques, the motivation for accumulating merit alone does not suffice to explain the sustained transcription of the canon. If it was the primary reason, then the enthusiasm for copying the canon should have persisted even after the introduction of woodblock versions. Instead, following the relaxation of the state policy, the latter flooded the market and eventually replaced handwritten copies. With the increasing prevalence of printing technology, “copying the sutras” and “printing the sutras” came to be seen as equally meritorious in the Song Dynasty. This was due to the belief that both practices constituted offerings to the Dharma 法供養, thus carrying an equal amount of merit. When monasteries were allowed to take advantage of more efficient and cost-effective printing methods, the printed copies of the canon soon became their preference, and the handwritten copies correspondingly shrank. Therefore, while the merit of copying sutras was undeniably valued in this process, it was not the main reason. Then, what other factors played a role in sustaining the transcription of the canon? Besides the state printing policy, the social aspect is another indispensable perspective.

4. Society and the Production of the Canon

The state printing policy directly caused the continuity of the handwritten copies; however, without significant societal demand for the canon, no state policies—whether prohibitive or permissive—would have affected its production. The fact that, when printing was not allowed, monasteries persisted in the arduous and time-consuming process of producing handwritten copies underscores the indispensability of the canon in Song society. Therefore, social factors can be seen as the driving force behind canon production. As Wu points out, creating the Chinese Buddhist canon was “a complex social practice” that depended on multiple layers of social relations and various conditions. It is essential to examine the social dynamics involved in the canon’s production and the community it emerged from, as the two are inseparable (Wu 2014, pp. 374–76).

4.1. Canon and Monastery

Upon examining the remaining manuscripts and historical records, it becomes evident that almost all of the Northern Song’s canon-copying projects were initiated and executed by monasteries. Additionally, it is worth noting that, except for the Kaibao Canon, all other woodblock canon versions created during the Song period were coordinated and manufactured by four monasteries in Fujian 福建, Zhejiang, and Jiangsu. As the canon represented the primary asset and symbolic capital for monasteries, they became the principal producers of the canon. The interaction between the Dharma and the Saṃgha, two of the three treasures, was rooted in the connection between the canon and the monastery in a local community.

On the one hand, the persistent thriving of the Dharma embodied by the Buddhist canon primarily hinged on the Saṃgha’s effort of preservation and dissemination. Therefore, producing the canon in an appropriate form allowed by the conditions became the most fundamental religious mission of the monastery. As the head of the monastery and the spiritual leader of the local community, the abbot bore this mission. In a record composed for the handwritten canon produced by the Chongshou chanyuan 崇壽禪院, Qisong 契嵩 (1007–1072) elaborated on the motive of the project:

“Once, the Buddhist monk Yihuai preached here and attracted many followers. However, there was a shortage of scriptures, causing a problem for those who sought to study them. In response, Zhang Gonggan from Qinghe gathered his fellow townspeople and pooled their resources to commission the copying of the canon, the collection of the Buddha’s teachings.

昔沙門義懷於此說法聚徒。懷之徒既眾且盛。而經教缺然。患其無所視覽。而清河張拱感是遽相率其邑人出財鳩工。謀寫先佛三藏之說。”14

In this record, the monk Yihuai refers to the prominent Song Chan monk Tianyi Yihuai 天衣義怀, who, like Qisong, belonged to the Yunmen school 雲門宗. Even Chan Buddhism, which boasted of tearing the sutras under the slogan of “no setting up words,” still took the canon as an indispensable substance in real monastic life, and Chan monks also took an active role in preserving the Buddha’s words. In addition, the success of canon-copying projects hinged on the moral leadership of the chief monk, their influence within the community, and their ability to collaborate effectively with local secular authorities. For instance, the initiators of the Haihui yuan Canon, Faxi si Canon, and the Jingyan si Canon—Fengying 奉英, Sigong 思恭, and Liaocheng 了乘—were all bestowed with the prestigious title of ciziseng 賜紫僧 (monk bestowed with a purple robe), denoting their elevated status within the monkhood. The eminent monks’ great aspiration and outstanding ability to transmit the dharma was instrumental in the continuous transcription of the canon.

On the other hand, it is essential to note that the monks and monasteries also depended on the canon. The canon was essential to the Buddhist community, and, thus, the monastery’s ability to organize large-scale canon-copying projects and possess a complete canon set symbolized its local power and spiritual authority. This, in turn, helped the monastery secure patronage and prestige within and beyond the local society. In the highly competitive environments where many monasteries vied for eminent monks, owning a canon collection became a strategic advantage. Huang Tingjian’s 黃庭堅 (1045–1105) record of how the Donglin monastery 東林寺 pursued a canon to keep Donglin Changzong 東林常總 (1025–1091) as abbot serves as a good example. According to his account, Changzong, a renowned Chan master with hundreds or thousands of disciples, refused to take the abbotship. Then, the monks and the laities strived to build him a revolving wheel storage cabinet housing the Lotus Sutra Canon 輪轉蓮花經藏, to which local people dedicated their skills, wealth, and efforts within a few years.15 Though the canon they contributed was a simplified and printed version, it poignantly demonstrates both the symbolic significance and the realistic value of the canon for a monastery. Huang also recorded another canon-related story, revealing that his one comment propelled the genesis of the canon in the Yunyan monastery 雲巖禪院, located in Jiangxi: “The abbot qing (Faqing 法清) heard that Shangu (Huang Tingjian) once said that the Yunyan monastery did not have a collection of the canon, so he eagerly wished to accomplish this task. 清聞山谷嘗道雲巖初無藏經, 慨然欲辨此緣.” Notably, a literati’s comment carried so much weight for the monastery, but more illuminating is that the monastery did not have a canon collection that drew the attention of a literati. Huang’s comment might represent many of the Song literati’s propositions: a canon collection was an essential component of a monastery. Another Song scholar, Zhu Chuyue’s 朱處約, asserted that “Today, all renowned mountains and grand monasteries must have a set of canon, revered as the fundamental basis of solemnity and observance 今天下名山劇寺必有《大藏經》, 奉為偉觀秩守之宗.” (Fu 1974)16 For the literati of the Song period, having a collection of canons was a crucial criterion for the prestige of a monastery. Thus, housing a canon served as a cornerstone for cultivating relationships with the literati class, who were among the most influential patrons of Buddhist institutions. Numerous records composed by Song literati for monasteries’ completion of a canon collection and construction of the revolving wheel storage cabinet attested to the canon–bond rapport between monasteries and literati. Besides the eminent monks and the well-known literati, a canon collection also attracted the general public, the monastery’s most consistent and regular benefactors, to come and give alms.

The sacred mission of preserving and spreading the Buddha’s teachings and the real benefits of possessing a canon collection triggered a “canon complex” in monasteries, especially those in the affluent Jiangnan area where Buddhist textual tradition had a long history and the Buddhist belief penetrated even to the very small locales. The “canon complex” aroused an intense canon-copying competition in the Jiangnan region. The aforementioned Yongming si Canon, produced in Zhejiang, was a direct result of such a regional competition, as recorded in Lin Lu’s writing:

“The words of Buddha have been spread in over 5000 fascicles in China. His disciples categorized and collected them, known as the “Buddhist Canon.” In the Zhejiang region, people generally have a great affinity for Buddhism; thus, the Buddhist canon collections are particularly abundant in all districts. However, the Yongming monastery in Cixi alone does not possess a collection. Puzheng, a senior monk of the Yongming monastery, said, “Without access to the Buddhist canon, those who study and practice Buddhism cannot understand the Buddha’s teachings.” Therefore, the monastery poured money and sought land to start copying the canon and building a canon house.

佛之言流於中國五千餘卷，其徒傳錄，類聚而藏之，世謂之“藏.” 浙中大率喜奉佛，所謂藏者尤多於諸道，獨慈溪永明寺未始有之。寺之僧普證大師知簡以謂：“學佛者不得其書以觀，則無以知佛之意.” 故傾財求地，首為寫經造院之謀。”17

The intense competition for canons in the area is a conceivable possibility. Chen Lin’s 陳林composition for a canon collection in the Longping monastery 隆平寺 also attests to such intensity, indicating that the Qinglong county 青龍鎮 (in present-day Shanghai) had three monasteries in total, and all possessed a canon collection (Chai and Pan 2004, p. 22). Given this environment, it is unsurprising that six known canon versions were produced in Xiuzhou alone. During the Song dynasty, monasteries in Jiangnan generally harbored a “canon complex,” which propelled them to devote all efforts to having a canon collection, thus possessing the primary symbolic asset for monasteries and potentially securing more resources and prestige. Some raised money to purchase printed versions, while others organized copying the canon—the coexistence of handwritten and block-printed copies of canons, for a century, owes itself to the all-encompassing “canon complex” that pervaded monasteries across Jiangnan, from bustling cities to small counties and villages.

4.2. Canon and Jiangnan Community

Beyond its essential role within the monastic community, the canon held profound significance for the local populace, carrying a weighty and profound meaning. While Jiangnan monasteries were the initiators of canon-copying projects, none of the Northern Song’s canon-copying projects were completed without the support of local communities. Take the case of the Haihui yuan canon; in a stone inscription, Chen Shunyu rendered a succinct account of its genesis:

“The monk bestowed with the purple robe, Fengying, was a person of superior intelligence and ability, and he was appointed as the leader. He then recruited people to transcribe the canon that had been passed down. The handwritten canon consisted of 800 cases and 5048 fascicles. There were local people surnamed Wu, very benevolent and generous, who were deemed as the heads of lay believers. He raised funds and hired workers to build a revolving wheel cabinet to store the canon.

賜紫僧奉英，智力膚敏，傑為主者，乃募人書所傳之經，其函八百，其巻五千四十有八，而居人吳氏子仁義施，號為長者，為之募財僝工作轉輪而蔵之.”18

Moreover, the production procedure for the Haihui yuan Canon is extensively documented in three surviving manuscripts, each bearing an almost identical weiti 尾題 (end titles), as shown in the manuscript of volume sixty-five of Fayuan zhulin 法苑珠林 (Forest of Gems in the Garden of the Dharma):

“This text was initiated on the 15th day of the fourth lunar month in the first year of the Song Dynasty’s Zhiping era (1064). Feng Yu from Siming transcribed it with reverence. Monk Chu Yan from Jingyan Monastery proofread. Shouying, the abbot bestowed with the purple robe, gathered donations from the public to create the canon. The chief donors who contributed to the construction of the canon were Wu Yanliang and Yan You from Bohai. The virtuous senior monk Zhaoyi preached the Sūtra, vinaya, and śāstra. The monks Shourong and Shouning were responsible for categorizing matters.

聖宋治平元年歲次甲辰四月十五日起首，四明馮預敬寫；精嚴寺沙門楚顏校證；住持募眾緣寫造大藏賜紫沙門守英；造藏檀越渤海吳延亮、延宥；耆宿講經律論沙門昭益；法屬沙門守榮、守寧.”19

Additionally, an identical note right after the end titles on the three manuscripts states that “Xu Hui, Weng Xiu, and Chen Yi donated the payment for ink 徐惠, 翁秀, 陈义舍墨.” Thus, a comprehensive view of a network centered around canon copying is presented, eloquently attesting to the deep involvement of local communities in the production of handwritten copies of the canon. Notably, the colophons and end titles of the other four known extant canon collections reveal a similar network and procedure. The handwriting canon project was a collaborative effort between monasteries and local communities, requiring the mobilization of diverse monastic and social resources. The responsibilities were divided among the participants, with the abbot as the chief fundraiser and organizer, the local gentry acting as chief sponsors, other locals as donors, senior monks as preachers, collators, and proofreaders, and locally renowned calligraphers as scribes.

Like the formal procedures of copying scriptures during the Tang period, canon copying during the Song period also featured a clear division of labor, usually recorded in the end titles of the manuscript. Furthermore, individuals or families could fund the copying and offering of a single text from the canon, with the names of donors and a prayer (yuanwen 願文) typically included in the colophon. A manuscript of the Jinsu shan Canon Sotheby auctioned in 2022 is marked with the character gui 貴, which contains a colophon revealing the donor’s motive: “Disciples Zhu Jiu’ao and his younger brother Jiulin from Xiashi donated a scripture, urgently praying for a swift recovery of their mother, Lady Song, from illness 硤石市弟子朱九翱弟九璘捨經一。亟保母親宋氏二娘妳疾痊除者。”20 As in the Tang Dynasty, copying or donating copies of the scriptures constituted a vital channel to accumulate merits. The canon-copying project allowed locals to pray for various worldly benefits.

Despite some similarities, copying canons during the Song period had distinctive features. In the Tang period, high-quality manuscripts were typically commissioned by the court, which demanded specific standards for the paper, calligraphy, and dyeing technique of the huangbo 黃檗 (Phellodendron) solution. Locally produced manuscripts, on the other hand, often fell short in terms of quality (Rong and Galambos 2013, pp. 484–85). However, during the Song period, locally produced canons showed a remarkable quality that was not inferior to court versions, demonstrating the local community’s active and capable role in producing handwritten copies of the canon, particularly in the Jiangnan area. Notably, wealthy local elites, known as yiren 邑人, juren 居人, or xiangren 鄉人 in historical records, were the main benefactors of these canon-copying projects. These elites were closely connected to their local communities, further empowering them with a great mobilizing ability. The significant involvement of the Jiangnan local society in producing the canon is attributed to the transformation of the social structure starting from the ninth century. As scholars have argued, this social transformation led to the rise of the gentry in local society, particularly in wealthy South China. Thus, it is intriguing to see that, when the state reinforced its monopoly over textual authority, attempting to nationalize classical knowledge, the local society adopted the handwritten form to publicize the Buddhist classics, at least locally. While this paper does not investigate the power dynamics between the state and Jiangnan society regarding control over the Buddhist canon, it is worth considering the large-scale, canon-copying project’s role in undergirding the local gentry’s authority and rallying the local society.

Additionally, the prevalence of handwritten copies of the canon in the Jiangnan region can be attributed to the material culture of the Jiangnan society. One crucial factor was the unique techniques used in paper production and preservation. The surviving Song manuscripts of the canon, particularly the Jinsu shan Canon, are renowned for their exceptional paper quality. In the Northern Song period, a specific type of paper called zangjing zhi 藏經紙 (paper for copying the canon) was invented in the Jiangnan region. For example, each sheet of paper used for the copies of the Jinsu shan Canon was marked with a small rectangular seal stamped in red ink, which reads Jinsu shan zangjing zhi 金粟山藏經紙 (paper for copying the canon from Jinsu Mountain). Similarly, the Faxi si Canon and the Haihui yuan Canon also bear a mark denoting the use of zangjing zhi paper. This type of paper was exclusively produced in the early years of the Song Dynasty in the Jiangnan area, known for its fine quality that yielded excellent writing results and maintained durability and beauty over time (Wu 2008, pp. 1–3). The paper was said to be made from silk cocoons; however, in reality, it was made from a combination of mulberry and hemp. Both sides of the paper were treated with wax, resulting in a smooth and glossy surface (Wu and Chia 2015, p. 33). Given that the above three versions were all produced in Xiuzhou, and the amount of paper used to copy the canon was enormous, there might be paper mills in the area that specialized in producing zangjing zhi.

We can further assume that there was a well-developed canon-copying business chain in the Jiangnan area during the Northern Song dynasty. The end titles of the manuscript of the Haihui yuan Canon, as mentioned earlier, provide evidence supporting this speculation. Although the Haihui monastery was located in Huating, the social network indicated in the end titles extended beyond the boundaries of Huating County to a broader region. Both Chuyan, the proofreader, and Feng Yu, the calligrapher of this manuscript, came from neighboring areas. Chuyan was from the Jingyan monastery 精嚴寺 located in the nearby Jiaxing 嘉興 county, and Feng Yu was from farther Siming 四明 in Mingzhou 明州. As discussed earlier, the Jingyan monastery also produced a canon collection, although no manuscript remains. It is likely that, when the Haihui monastery embarked on canon copying, they recruited Chuyan, an experienced monk involved in the production of the Jingyan si Canon. Moreover, the scribe Feng Yu also transcribed a manuscript of the Fo shuo liaozhi jing 佛說了知經 for an unknown monastery’s canon collection, which is now preserved in the Guangdong Museum. It suggests that, as a locally renowned calligrapher, he was involved in transcribing more than one canon version and was probably a professional chaojingsheng 抄經生 (Buddhist scripture scribe). Thus, from paper producers who manufactured paper specifically for transcribing the canon to calligraphers renowned in the Jiangnan region, who were usually recruited by monasteries to transcribe the canon, the canon-copying business chain during the Northern Song dynasty demonstrates a highly developed and sophisticatedly organized social practice.

Finally, to fully comprehend the fervor for canon copying in the Jiangnan region, examining the practical use of the handwritten copies of the canon within society is crucial. Jiang Wu highlights the significance of the “Cult of the Canon,” emphasizing that the canon’s transformation into an object of devotion is a vital aspect that cannot be overlooked (Wu and Chia 2015, p. 46). In this light, the Northern Song period’s handwritten copies were not only used for Buddhist learning and monastic purposes but as “an object of devotion” that was used by a wider community. During the Song dynasty, the canon’s most common form of use was directly associated with its container—the zhuanlun zang 轉輪藏 (revolving wheel storage cabinet), a devotional device designed to house the canon and rotated by followers for divine blessings. This practice originated in the Tang period yet flourished in the Song. Jiang Wu posits that the demand for a canon collection to fill the revolving wheel storage cabinet may have been a driving force behind printing privately produced editions after the Kaibao Canon (Wu and Chia 2015, p. 54). This theory is valid and might apply to the intense canon copying during the Northern Song period. Surviving records indicate that most handwritten copies were made in connection with constructing a revolving repository, which was already popular in the mid-eleventh century. Therefore, it is reasonable to assume that the need to fill the revolving repository with a canon collection may have also stimulated the demand for handwritten copies of the canon. In this context, the “Cult of the Canon” in the Song Dynasty manifested through the fervor for rotating the canon, which was particularly pronounced in the Jiangnan region. It is noteworthy that the inventor of the revolving wheel storage cabinet was said to be Fu Xi 傅翕 (497–569), known as Fu Dashi 傅大士, Shuanglin Dashi 雙林大士, or Dongyang Dashi 東陽大士, a native of Zhejiang. It seemed that this tradition of revolving wheel storage cabinet was thus passed down and became increasingly influential around this area, as Shi Yuanzhao 釋元照 (1048–1116) recorded that “Shuanglin initiated the revolving wheel storage cabinet, and later other monasteries followed suit. The ones in Eastern Zhejiang were especially prosperous in this regard 故輪藏自雙林為始厥後他寺皆倣之而浙東尤盛.”21

On the one hand, the revolving repository’s construction was such a luxurious endeavor that it was only within the financial reach of the affluent society of Jiangnan to make it a widespread phenomenon. The maturity and sophistication of wood construction techniques in the Jiangnan region also contributed to the surge in this device’s construction. On the other hand, the fervor for constructing and rotating the revolving repository sheds light on the relationship between the canon and society during the Song, revealing the characteristics of a “canon for devotion” based on the concepts of merit and relying on others’ power for salvation (Fang 2006, p. 540). In the Song period, the container took on a sacred status, akin to that of a canon, and was revered as an object of devotion. Its convenient and accessible nature made it popular among a wider community, as rotating the revolving repository held the same merits as reading through an entire collection of canons. The “Cult of the Canon” in the Song period led to the emergence of novel worship practices. In addition to the conventional five practices of upholding, reading, reciting, explaining, and copying, rotating the canon became a quintessential aspect of Song-era worship. Moreover, the monks and literati of the Song Dynasty spared no effort in justifying this new practice. For example, Chen Shunyu, in his writing for the Bujin Canon in Xiuzhou, illuminated the significance of revolving the wheel:

“Without sentient beings, there would be no Buddha. Without the Buddha, sentient beings would not be saved. However, when the Buddha saves sentient beings, he never leaves our wheel. He only points them in the right direction from the wheel’s position, and thus can the Buddha’s teachings spread universally without limits. Therefore, anything that can rotate can help end suffering and propagate the Dharma. In this sense, it is similar to the Buddha.

有衆生乃有佛,非佛不能度衆生,然佛之度衆生也,未尝脱吾轮而载之, 葢即其所乘而指其所向,故能方轨同辙, 而出乎无穷之域焉耳, 然则凢所谓轮者皆可以推止诸苦,令法流转,亦几于佛矣.”22

The monk Shi Yuanzhao also defended this invention, expounding its benefits for all:

“Some only saw the revolving wheel storage cabinet turning and gained faith, while others realized that it had never turned upon seeing it turning. Some dwelled in perfect stillness while following the turning all day. Whether the wise or the ignorant, everyone benefited according to their capacities.

或徒見其轉而生信者, 或因其所轉而知其未嘗轉者, 或安住無轉而從其終日轉者, 上智下愚. 隨分得益.”23

The endorsement of the construction of the revolving repository by both monks and literati reflected society’s yearning for merit. The spread of miraculous stories about the repository’s protective and fortune-eliciting functions further fueled this fervor. From these stories, we can see that a cult devoted to the revolving repository emerged during the Song period, with the appearance of a deity—zangshen 藏神 (Deity of the Canon). A record in the Yijian zhizhi 夷堅支誌 recounts a story in Taizhou 台州 about this deity:

“At Shangtingbao in Linhai County, Taizhou, there was a small monastery called Zhenru Monastery. In the east wing of the monastery was a revolving repository with the image of a deity there. This deity had always been efficacious. Prayers and offerings made by merchants traveling by sea never ceased. During the Shaoxing period (1131–1162), a novice monk trainee, Jin Fajing, was diligent and respectful in taking charge of the monastery’s incense and was promoted to the chief novice. Once, a barber was trimming his nose hair, and a monk named Zhiquan happened to pass by and accidentally bumped into Jin Fajing’s head. As a result, the scissors broke off in Jin Fajing’s nose and could not be pulled out. Blood gushed out suddenly, and Jin Fajing immediately fainted. In a trance, he seemed to see the “Deity of the Canon,” who raised his hand and pulled out the scissors. He immediately felt much relief from the pain; the scissors fell to the side, and his wound healed within ten days. This miraculous story soon spread, and as a result, more and more believers came to the monastery to make offerings.

台州臨海縣上亭保，有小剎曰真如院。東廡置輪藏，其神一軀，素着靈驗。海商去來，祈禱供施無虛日。紹興中，童行金法靜主香火之事甚敬，爲寺參頭，因令剃工繳鼻，爲僧智全從旁過，誤觸其首。刀中斷，牢不可取，出血至數昇，悶僕不醒。恍惚間見藏神至，舉手拔之，覺痛少止。刀墜於側，旬日瘡愈。自是遠近傳説，檀信益衆。”24

This story adds nuance to our understanding of the ‘Cult of the Canon’ during the Song period, which involved the materialization of textual worship and the subsequent anthropomorphization of material objects. The origins of the “Deity of the Canon” are shrouded in mystery, as it is unclear when and how this deity came to be. However, it is evident that the presence of this deity is intimately linked to the cult of the revolving repository. This cult gained momentum in the mid-eleventh century and reached its zenith in the twelfth century, with even “a small monastery” in Taizhou boasting a revolving repository complete with its deity. It is also seen that this deity performed the same function as other deities to protect the local people. The revolving wheel storage cabinet was vital for the local community’s spiritual and secular life, constituting a tangible object of devotion and a sacred spring of benefits with its intricate machinery, gold adornments, and precious manuscripts. Therefore, as historical records demonstrate, whenever a monastery in Jiangnan planned to construct a revolving repository, people from all strata of local society would donate generously. Once completed, the repository attracted a steady stream of worshippers worshipping the canon, rotating the cabinet, and seeking worldly interests. In light of the community’s strong demand for the revolving repository, one possible explanation for the intensity of copying canon projects in the Jiangnan area is that people were copying the canon to build revolving repositories to meet this demand.

5. Conclusions

The century-long coexistence of handwritten and block-printed copies of the Buddhist canon marked a significant period in the history of the Chinese Buddhist canon. This period represented the last era of handwritten copying and the dawn of printing. A socio-political perspective is needed to understand the proliferation of canon copying in the Jiangnan region during the Northern Song dynasty, given the similar pattern in the time and space of extant manuscripts. The state’s printing policy regarding canon production was the direct cause of this phenomenon. Additionally, the significant demand for the canon from Buddhist institutions and local society in the Jiangnan region, particularly for completing the revolving wheel storage cabinet, drove the intense canon-copying projects. It is important to note that the rich economic environment and sophisticated material culture of the Jiangnan region also played a vital role in supporting the copying of the canon and the construction of the revolving repository.