The Discipline of Writing: Scribes and Purity in Eighth-Century Japan

Abstract

This article focuses on ritualized scribal practices in eighth-century Japan. It uses colophons, scriptorium documents, and narrative tales to explore how sutra copyists upheld vegetarian diets, performed ablutions, wore ritual garments, and avoided contact with pollutants stemming from death and illness. Such practices, often described in terms of purity, spread widely on the Asian continent in the seventh century and reached Japan by the eighth century. This article argues that upholding purity was deeply connected to notions of ritual efficacy but also enabled pious lay scribes to train for monastic careers. The evidence is used to reassess historiographical debates on Nara Buddhism with particular attention to the well-known "theory of state Buddhism" (kokka Bukkyo ron). [PUBLICATION ABSTRACT]

Headnote

This article focuses on ritualized scribal practices in eighth-century Japan. It uses colophons, scriptorium documents, and narrative tales to explore how sutra copyists upheld vegetarian diets, performed ablutions, wore ritual garments, and avoided contact with pollutants stemming from death and illness. Such practices, often described in terms of purity, spread widely on the Asian continent in the seventh century and reached Japan by the eighth century. This article argues that upholding purity was deeply connected to notions of ritual efficacy but also enabled pious lay scribes to train for monastic careers. The evidence is used to reassess historiographical debates on Nara Buddhism with particular attention to the well-known "theory of state Buddhism" (kokka Bukkyo ron).

KEYWORDS: purity-sutra copying-scribes-Nara Buddhism-state Buddhism- Shosoin

(ProQuest: ... denotes non-US-ASCII text omitted.)

Transcribing Buddhist texts in early Japan required more than a fine hand and a careful eye. Before even picking up a brush, scribes upheld vegetarian diets, performed ablutions, donned ritual garments, and avoided contact with pollutants stemming from death and illness. Various sources from early Japan and the continent describe these practices in terms of "purity." But why would sutra copyists bother to copy texts in highly ritualized ways? What meaning did these practices hold for scribes and patrons?

In this article, I will use an interpretive framework that views scribal practices as disciplinary regimes. The word "regime," which now commonly implies an unjust system of government, originally referred to dietary and meditative practices that were intended to improve a person's welfare-a meaning that is preserved in the modern English "regimen." The term "discipline" similarly has a passive and active sense. First, discipline can be imposed from above, as in being disciplined by one's teacher. In the second sense, which may arise as a result of the first, one becomes self-disciplined and is able to do what is required independent of external force.1 More concretely, patrons and scriptoria administrators demanded that scribes uphold ritual protocol tied to diet and dress to ensure that the manuscript would become empowered to efficaciously answer the patron's prayers. These same practices prepared religiously motivated scribes for ordination. In this way, I hope to insert a sense of agency into our under-

This dual notion of discipline allows us to ask new questions about religious practice in eighth-century Japan. Most scholarship over the last half century fits within the framework of the "theory of state Buddhism" (kokka Bukkyo ron ...), a model most clearly and influentially articulated by Inoue Mitsusada.3 In this model, the state functions as a regulatory body that controlled (tosei ...) Buddhism through administrative and legal reform (Inoue 1971, 31-52). A new generation of scholars, led by Yoshida Kazuhiko, has challenged Inoue's conclusions. Yoshida's research suggests that the legal measures instituted by the state were for the most part dysfunctional, and he highlights the importance of figures operating independently of the state-a phenomenon he refers to as "the Buddhism of the masses" (minshu Bukkyo ...; Yoshida 1995, 30-97; 2006a, 25-28 and 36; and 2006b, 148-49 and 153-54).

Although both the state Buddhism model and the more recent revisions by Yoshida have shed much-needed light on early Japanese Buddhism, each approach presents its own set of problems. On the one hand, models that focus on regulation tend to see social structures as determinative and consciously instituted by elites as a means of exerting control. On the other hand, scholars who emphasize resistance against norms and religious freedom frequently impose Western, liberal values of autonomy and self-expression on an eighthcentury culture, where such values may in fact be nonsensical. Furthermore, an overemphasis on "the Buddhism of the masses" simply moves attention away from the state without ever assessing its purpose and limits. It is undeniable that the state did play a key role in both promoting and institutionalizing Buddhism, so a more complete study of the period must take into account the relationship between the state and individuals.

A study of scribes and discipline can offer new perspectives on Buddhism in the Nara period (710-794) that avoids the extremes of the two aforementioned approaches. For one, the regulation of Buddhism must be understood within a broader East Asian religious context of ritual protocol. In this way, as Michael Como (2009, xvii) has recently noted, the court was not only the producer of an ideology but also a consumer of it. Namely, in regulating the purity of scribes, the state followed ideas of ritual shared widely throughout East Asia by diverse tradi- tions. For this reason, it is necessary to first explore the continental context. Second, the regulation of ritual officiants was not unique to the state; eighth-century scribes operating independently of the court engaged in practices that mirrored those performed in the capital. As I will show, the regulation of ritual purity and the view that sutra copying functioned as a form of religious practice was widespread in early Japan. Finally, and perhaps most importantly, it is necessary to clarify how power, which I see as a contingent combination of forces, was enacted on the body through specific institutions such as an official state scriptorium. Rather than examining regulation from a top-down perspective, I will explore how the disciplinary regimes that scribes were subjected to enabled them to cultivate certain techniques and habits that would prove useful for their religious pursuits.4

Purity and Scripture Copying on the Continent

Ritualized methods of copying scripture first appear in extant sources from seventh-century China. The earliest example that I have found is a colophon dated to 620 ce from Dunhuang appended to the third scroll of the Vimalakirti Sutra:

A disciple burned incense and summoned the sutra copyist Zhu Lingbian ... from afar. [The scribe] diligently upheld the abstinential precepts [Ch. zhaijie ...],5 bathed with fragrant water, put pure robes over his body, and stayed in a quiet chamber practicing the path at all times. (n no. 66, 2: 6)

This colophon suggests that a patron hired a lay scribe and had him take abstinential precepts, perform ablutions, and wear special garments known as "pure robes" before copying scripture. The practices described in this colophon appear frequently in widely circulating seventh- and eighth-century Chinese narrative tales.6 For example, in one story found in numerous East Asian collections, the monk Tanyun ... desires to copy the Lotus Sutra but cannot find anyone of like mind. Without notice, a scribe mysteriously appears who agrees to help Tanyun purely copy the scripture (Ch. jiejing xiejing ...). This entails bathing, donning pure robes, and receiving the eight precepts. After performing these practices, the scribe enters a pure chamber (Ch. jingshi ...). Once inside, he consumes sandalwood, burns incense, and hangs banners to further sanctify the space and his body. He finishes copying the scripture and vanishes as suddenly as he had appeared, leaving Tanyun with a beautiful sutra that proves to be empowered against rotting.7 Here, the method for "purely" copying sutras refers to the same practices we saw in the 620 ce Dunhuang colophon: performing ablutions, wearing special garments, and upholding precepts.

Another story from Records of Flower Garland Sutra Lore (Ch. Huayan jing zhuanji) shows that concerns over purity extended to the preparation of materials used in sutra transcription.8 In this tale, a monk named Deyuan ... decides to copy the Flower Garland Sutra. He first grows paper mulberry (Ch. chu ...) in a pure garden (Ch. jingyuan ...) with fragrant grasses. He bathes every time he enters the garden and waters the mulberry with fragrant water for three years. The paper is made while maintaining purity (Ch. hujing ...) by craftsmen who uphold abstinential rules and change their clothes each time they enter the chamber. After preparing the paper, Deyuan constructs a hall for copying scripture and even washes the timbers in fragrant water before using them for construction. He hires a scribe, whom he has bathe three times in a fragrant bath, receive the abstinential rules, wear pure clothing (Ch. jingfu ...), and don a crown of flowers before beginning to copy the text. As a result, the sutra shines forth light that illuminates the room and numerous good omens appear (t 2073.51.1070c-1071a).

Deyuan is not alone in his close attention to detail-similar tales abound in Buddhist anthologies.9 These tales repeatedly stress the need for scribes to uphold purity. The words of a pious nun named Miaozhi, who appears in the Biographies in Broad Praise of the Lotus (Hongzan fahua zhuan), summarize this sentiment: "[It is best to] hire a scribe who is able to maintain purity. It does not matter if he is skilled or clumsy or exalted or base" (t 2067.51.45b). Failure to follow Miaozhi's advice could compromise ritual efficacy. In one story, a man named Linghu Yuangui ... purely copies a group of sutras in his own hand while upholding purity. When he returns home, he stores the texts carefully, but a fire breaks out and burns the building to the ground. He orders his servants to search through the ashes and they are fortunately able to recover the sutras. Although the boxes containing the texts have been reduced to ash, the scriptures themselves emerge for the most part unscathed. Upon closer examination, Yuangui notices that the title of the Diamond Sutra is charred black. Yuangui had hired a scribe with particularly fine calligraphy to transcribe the title of the Diamond Sutra. This scribe, however, "ate various things [zashi ...]... and did not maintain purity [buhu jiejing ...]." As a result of the scribe's indiscretions, the section transcribed by the title copyist failed to become ritually efficacious and burned as if it were an ordinary text.10

Documentary evidence suggests that some Chinese and medieval Japanese scribes followed these ritualized procedures. We have already examined one Dunhuang manuscript that cites similar practices; other colophons from 662 ce describe how copyists "purely" (Ch. jiejing ...) transcribed an entire canon sponsored by a monk and a fellowship group from the Yu ... district (ex. p 2056; n no. 69, 2: 20; bd 14496). In Japan, a ritualized form of transcription known as according with the correct method (nyoho ...), spread widely in the Heian (794-1185) and Kamakura (1185-1333) periods (Kabutogi 1983, 3-143; Moerman 2007, 252-53 and 2010, 76-77; Nakano 2009, 118-33).11 By early medieval times, various sects produced detailed ritual manuals that describe the process of making paper and copying scripture in a manner quite close to that undertaken by Deyuan, but with increased codification and additional elements such as penance rites and liturgies.12 Before exploring the Nara period precedents for these later practices, let us step back from sutra copying for a moment to reflect more broadly on the notion of purity in early Japan.

Defining Purity

The above stories frequently referred to practices related to diet, dress, and body as means of "upholding purity." It is still necessary, however, to consider how purity was understood in early Japan and on the continent and to explore why it played a central role in ritualized sutra transcription. Terms translated as purity include sei (also kiyoshi ...), jo (also kiyoshi ...), and ketsu (also isagiyoshi and kiyoshi ...)-all of these logographs share a sense of cleanliness. More specifically, these terms imply the absence of dirt and defilement (Jp. kitanashi and kegare ..., ...), which are generally viewed in negative terms. In Buddhist sources, words that are translated as purity have connotations of the absence of negative features and the presence of virtuous ones. The Sanskrit suddha (Ch. qingjing; Jp. shojo ...), one of the more common words used for purity in Buddhist texts, is said to be acquired by developing good roots and removing mental and bodily hindrances (for example, Mohe bore boluomi jing [Maha-prajña-paramita-sutra] t 223.8.340a). Brahma-carya or pure conduct (Ch. fanxing; Jp. bongyo ... or Ch. jingxing; Jp. jogyo ...) refers to behavior in line with the Buddhist precepts and is often used as a euphemism for sexual abstinence.13 Purity, therefore, has a negative quality-namely, the absence of bodily defilement and the avoidance of physical acts and mental states that hinder spiritual progress-and a positive one of engaging in practices conducive to enlightenment.

Purity, in the sense of removing defilement, cleansing the body, and upholding virtuous behavior, was particularly important in the ritual sphere of early Japan. Some of the earliest Chinese reports of life on the Japanese archipelago suggest that individuals practiced sexual abstinence, performed ablutions, and avoided meat at specific times including periods of mourning or while awaiting the return of seafarers (San guo zhi "Wei shu" 30.855). In legal codes from the Nara period, those who had been in contact with death or illness or who had consumed meat were prohibited from participating in state rites (Yoro ritsuryo, nst 3: 213). This is closely related to the idea that appearing before the kami ... in a defiled state could evoke a wrathful response. As a result, great care was taken to ensure that all those engaged in kami ritual avoided defilement (Okada 1982; 1989).14 Similarly, numerous ninth-century Buddhist tales describe the karmic punishments one can receive for failing to uphold purity (Kurosu 2004). Early Japanese ideas about the connection between physical purity and ritual efficacy derived from diverse sources including continental religious practice, canonical Buddhist texts, and indigenous beliefs (Ooms 2009, 253-66).15

Whereas impurity could provoke danger, the cultivation of pure states promised numinous rewards. For one, tales describe pure individuals as possessing extraordinary powers. In one Japanese story from the Records of the Numinous and Strange from Japan, a woman purifies her body (kesshin ...) through bathing and diet and is then able to ascend into the heavens (Nihon ryoiki i.13, snkbt 30: 26-27; translated in Nakamura 1973, 124-25). Similarly, in a canonical Buddhist tale that was known in Nara Japan, a practitioner named Vimala (Ch. Wugou ...)-a name that in Chinese translation literally means "without defilement"-proves his purity by levitating in the air (Da baoji jing, t 310.11.596c-597a). Many sutras connect upholding purity to the Bodhisattva path. For example, in the Maha-prajña-paramita-sutra, the Buddha explains, "Through pure body and mind (Ch. shenxin qingjing ...), one can transcend the stage of the voice-hearers and the pratyeka-buddhas and enter the rank of the bodhisattvas" (t 223.8.340a). In other early Mahayana texts, lay bodhisattvas are advised to maintain purity (jingjie ...) by avoiding alcohol and sexual intercourse (for example, Banzhou sanmei jing, t 418.13.910b). The ultimate Mahayana goal of becoming a bodhisattva was achieved through the cultivation of pure physical and mental states. To borrow the words of Talal Asad (1993, 77), the inability to advance on the Bodhisattva path was at least partially "a function of untaught bodies." As we will see, the fact that purity at once functioned to ensure ritual efficacy and provided a means of cultivation is central to understanding how scribal practices at once produced ritually efficacious sutras and helped pious copyists advance on the monastic path.

On Pure Grounds: The Office of Sutra Transcription

An extensive documentary record shows that Nara period scribes followed many of the ritualized procedures depicted in narrative tales. These documents, which have been preserved in the imperial treasure house known as the Shosoin ..., provide a detailed depiction of how sutra copying was conducted at the Office of Sutra Transcription (shakyojo ...).16 The origins of this office can be traced to a household scriptorium run by Komyoshi ... (701-760), the future Queen Consort (kogo ...) to Emperor Shomu ... (701-756). When Komyoshi became Queen Consort in 729, the management of the household scriptorium came under the authority of her Palace Agency (kogo gushiki ...).17 From here, the scriptorium continued to undergo numerous name changes but eventually became affiliated with the great Nara temple Todaiji .... It continued to be active through 776 when it was permanently shut down.

Although the Office of Sutra Transcription was an institution staffed by lay employees, it was housed on temple grounds. In the early phases, when the scriptorium was connected to the Queen Consort's Palace Agency, it was located at small temples and chapels (in ...) in the grounds of Komyoshi's palace. From 741, the Office of Sutra Transcription was moved to Fukujuji ..., a temple that would later become Todaiji. For most of the scriptorium's history, it remained there with a brief one-year stint at Ishiyama-dera ... in 762.

In Nara Japan, temples were considered pure spaces, as is reflected in the 5/15/716 edict that declares, "In operating and maintaining a Buddhist temple, purity (shojo ...) is primary" (Shoku Nihongi 5/15/716, snkbt 13: 10-12).18 Significantly, many Chinese stories mention patrons building a pure chamber for copying scripture. One popular Chinese story that was also known in Japan and Silla describes a nun from Hedong ... who hires a scribe to copy the Lotus Sutra.19 She installs a bamboo pipe through the wall of the sutra-copying chamber and makes the scribe breathe through the pipe whenever he exhales in order to maintain the purity of the room.20 In a Japanese story that will be discussed below, a scribe is karmically punished by death for defiling temple grounds while copying scripture (Nihon ryoiki iii.18, snkbt 30: 154-55; translated in Nakamura 1973, 245-46). The location of the Office of Sutra Transcription at chapels and temples, therefore, must be understood within this broad discourse of setting aside a pure space for copying scripture. This would have at once enhanced the efficacy of the sutra-copying project and provided devout scribes with an ideal location for Buddhist practice.

Dietary Discipline

In addition to creating a pure space, patrons and officials at the Office of Sutra Transcription aimed to purify scribes' bodies through enforcing strict diets, frequent baths, proper dress, avoidance of defilements from death and illness, and possibly even sexual abstinence. Most of the practices performed by scribes in Japan have precedents in the Chinese narrative tales and colophons introduced above, but they also must be understood as general methods of ensuring ritual efficacy that were by no means uniquely Buddhist. It cannot be overemphasized that the scribes themselves were lay people, usually of relatively humble rank and status. As we will discuss in more detail below, it appears that at least some scribes had religious ambitions although they were all unordained while serving as sutra copyists.

Ration records from the Office of Sutra Transcription reveal that scribes followed a vegetarian diet and avoided the five pungent roots.21 Food was provided by the scriptorium, so the dietary intake of scribes was strictly controlled. Shosoin documents often refer to such rations as "food offerings" (ryori kuyo ...). The term kuyo implies an offering given to the Buddha or to monks. In another case, a man named Yamato no Oyumi noted in a letter requesting employment that he had been engaging in abstinential dietary practices (saijiki ...) for some number of fortnights (dnk 5: 332; zs 47: 5). Here, it seems that Oyumi viewed vegetarian practices as a prerequisite for scribal labor. The use of such religious language and the vegetarian diet at the scriptorium suggests that scribes occupied an ambiguous space between unordained officials and monks.22

In Japan, prohibitions against meat eating were deeply connected to ritual efficacy. Many of the ideas about the relationship between ritual and meat avoidance were shaped by continental practices shared by Buddhism, Daoism, and the state-cult, as well as what may be described as indigenous ideas of ritual.23 Drawing on these traditions, eighth-century Japanese law required all those participating in state rites to first abstain from meat (Yoro ritsuryo, nst 3: 213; Okada 1989, 24-25; Ooms 2009, 257). Moreover, officials frequently issued temporary prohibitions against meat eating to improve agricultural bounties, cure members of the imperial family suffering from illness, and respond to calamities such as droughts and plagues (Taira 1997; Harada 1993, 70-83 and 2000; Hirabayashi 2007, 168-223; Grumbach 2005, 72-81).24 Furthermore, the section of the Nara period legal code that deals with monastic affairs specifically prohibits the consumption of alcohol, meat, and the five pungent roots by the clergy, except when used as medicine (Yoro ritsuryo, nst 3: 218).25 As Hongo has argued (2005, 34), these laws were instituted to preserve the purity of monks and nuns; the ruling authorities thought that a defiled monastic community could cause natural disasters and disorder. With regard to sutra copying specifically, several tales describe how it is necessary to avoid meat as a means to empower a text.26 When we consider the ration records from the Shosoin collection in this context, it is clear that the vegetarian diet of scribes helped ensure ritual efficacy.

Buddhist discourse on meat is complex but the form of Buddhism that entered Japan lauded vegetarianism as an ideal and strongly associated it with purity. The connection between meat eating and impurity can be traced back at least as far as the vinaya. For example, the Four Part Vinaya (Sifen lü ...)-the most important monastic code in East Asia-forbids the clergy from eating meat that they had seen, heard, or suspected of being butchered specifically for their consumption. The Buddha refers to meat prepared in these manners as impure (Ch. bujing; Jp. fujo ...) (t 1428.22.872b). Although these regulations actually functioned to permit meat eating, the passage is important for our purposes in that it makes an early association between meat and impurity.27 More stringent prohibitions of meat eating would appear in a select group of Mahayana texts that were copied frequently in Nara Japan. The classic and most detailed position appears in the Nirvana Sutra (Da banniepan jing, t 374.12.386a-c). In a particularly vivid passage, the Sutra on Entering [the Country of] Lanka (Ru lengqie jing) likens meat to the impurity of a corpse and encourages those cultivating pure conduct to avoid meat, alcohol, and the five pungent roots (t 671.16.562c-563a-564b). In China, apocryphal texts such as the Brahma's Net Sutra (Fan wang jing) and hagiographical literature would further advance vegetarian ideals (Kieschnick 1997, 22-28 and 2005). These Chinese tales and Mahayana sutras-apocryphal or otherwise-were influential in early Japan and surely popularized the idea that the clergy should abstain from consuming meat. Avoidance of meat and the five pungent roots, therefore, did not only preserve ritual efficacy; it also represented a form of Buddhist practice that may have appealed to pious laborers at the scriptorium.

Washing Away Impurity

In both China and Japan, bathing is an ancient tradition tied to purification.28 Sutra-copying tales frequently mention the ritual bathing of scribes. One detailed story on this topic appears in the early ninth century collection of Japanese tales, the Records of the Numinous and Strange from Japan:

The sramanera Muro ... was from the Enomoto clan. He was self-ordained and without [a clerical] name. Since he was from the Muro district in the Kii province, he shall be called the Novice of Muro. He lived in Arata village in Ate district and shaved his head. He wore a kasaya while following a lay life and supporting a family through maintaining an occupation. He made a vow to purely copy one volume of the Lotus Sutra in accord with the [correct] method [nyoho shojo ni utsushi tatematsuru ...]. Completely by himself, he copied it and each time he urinated or defecated, he washed himself and purified his body [senyoku shite mi wo kiyomu ...]. From when he went into the room, six months passed and at last he finished cleanly copying it.

(Nihon ryoiki iii.10, 143-144; translated in Nakamura 1973, 235-36)

The story continues with a narrative twist similar to the seventh-century Chinese tale of Linghu Yuangui discussed above; a fire erupts that burns down Muro's home. Miraculously, the sutra emerges unscathed from the fire. The ritualized bathing of a copyist, who in this story has a status somewhere between monk and layperson, results in the text becoming fireproof.

These bathing practices, which are portrayed as an ideal in narrative tales, were put into practice at the Office of Sutra Transcription. For one, documents from the early years of the scriptorium show that scribes were issued loincloths for bathing (mizufundoshi ...) (Sekine 1974, 158-59).29 Second, accounting records also show that a bathhouse (yuya ...) existed at the scriptorium (dnk 15: 66; zzs 10: 7 verso). Third, Shosoin documents reveal that firewood was supplied for daily baths (dnk 2: 187-90; zsks 31: 6-8 and dnk 18: 19; zzs 39: 1 verso). This suggests that the scriptorium contained bathing facilities for the scribes and that scribes used them every day. The frequency of the baths is significant, because commoners and elites alike at this time were unlikely to have bathed daily.

Moreover, it appears that a sutra-copying project could be delayed as a result of insufficient bathing facilities. For example, in the brief one-year period that the Office of Sutra Transcription was located at Ishiyama-dera, a project was initiated to copy the Maha-prajña-paramita-sutra. An order issued by the Ishiyama- dera Construction Agency on 1/23/762 requested seven or eight scribes to be summoned for the project, which was set to begin on 2/8/762 (dnk 15: 141-42; zsbs 5: 6 verso and zs 44: 12-13 verso). Workers began to construct a bath slightly before the transcription was scheduled to commence.30 Records show that boards were received for the bath on 2/2/762 and nails were allotted on 2/7/762.31 The allotment of nails on the day before the project was set to begin shows that the carpenters were working on a tight schedule. In fact, the actual copying began on 2/11/762, three days after it was originally scheduled and four days after formal construction of the bath commenced. The most plausible reason for this delay was that the project could not be started until the completion of the bath. Similarly, in 7/764, after a year hiatus from sutra-copying activities, repairs and improvements began in preparation for another project to copy the Maha-prajña-paramita-sutra. Documents show that preparations for the project started with the construction of an ablutions pavilion (chozudokoro ...) and a bathhouse (dnk 16: 517-20; zzs 4: 18).32 Sutra copying, as these examples suggest, could not begin without a bath.

Another Shosoin document further reveals the purifying effects that washing could have for scribes. In this document, a sutra copyist requests four days offfrom work to bathe.33 In asking for time off, the scribe notes that he needs to "bathe and purify the body" (mi mokuyoku shi kiyomen ...) (dnk 22: 215; zzs 40: 1 verso). The character kiyomeru ... means "to purify" and the phrase mokuyoku ... has religious connotations of ritual bathing. Moreover, the requested four-day period implies that this was not a case of regular daily ablutions but perhaps reflects a more serious case of contact defilement, a topic we will return to below. These examples suggest that bathing represented another means through which scribes prepared themselves for ritual activities.

Pure Robes

After bathing, scribes would put on pure robes (joe ...). These robes were issued to new scribes when they started at the scriptorium and frequently appear in budgets for sutra-copying projects.34 In Shosoin documents, joe refers broadly to any type of purified garment used while engaged in religious work (Sekine 1974, 93-95).35 This process of bathing followed by donning pure robes is a common feature of the Chinese sutra-copying tales discussed above. When scribes at the Office of Sutra Transcription returned home from work for periodic vacations after completing their assignments, they needed to turn in their robes before leaving. 36 Scribes donned pure garments to mark their entrance into the ritual sphere and took them offwhen they returned to the secular world. This transition was not merely symbolic. Much of the regulation of early Japanese Buddhism centered on keeping monks and nuns separate from the populace to prevent the clergy from losing ritual efficacy through contact with the defilements of the non-monastic population (Hongo 2005, 34-47). In wearing a pure robe, a scribe's body became that of a ritualist and his actions took on ritually significant effects.

Robes were further purified through periodic laundering. Shosoin documents contain numerous requests for time offto wash a scribe's defiled robes (e ...).37 Documentary records show that female laborers (yatoime ...) were responsible for the laundry (Maruyama 2010, 241-44). As Katsuura Noriko and others have noted, laundering in premodern Japan was not simply a form of housework or cleaning but was deeply connected to purifying garments used in ritual (Katsuura 1995, 197-200; Meeks 2010, 140-41; Nishiguchi 1987, 137-39). Through periodic laundering, robes were transformed from a defiled state to a pure one.

Contact Pollution

Evidence also suggests that scribes would have been expected to avoid contact with defilements stemming from illness and death.38 Death defilement has a long history in Japan. It appears in well-known foundational myths such as the story of Izanagi and Izanami, in which Izanagi becomes defiled after venturing to the underworld in pursuit of his deceased wife.39 In broad terms, those who had come into contact with death defilement were forbidden from engaging in ritual activities in early Japan.40 Concerns over death defilement extended into the Buddhist sphere as well.41 For example, one story from the ninth-century collection of Buddhist tales entitled the Records of Numinous Responses from Japan (Nihon kanrei roku) features a poor man who had been given food and clothing by monks at Gangoji .... Still unsatisfied, the man returns to the temple to steal oil to sell for drinking money. Not only this, he enters the precincts after having come into contact with corpses while working at a burial ground (shiso no tokoro ...). After drinking half of the wine he had purchased with the money he obtained from selling the stolen oil, he once again goes to the temple to seek lodgings. His tale ends with misfortune; the destitute man dies before the night is out. The narrator explains that the thief was punished not only for stealing from the temple but also for defiling it: "Know this! Dirtying pure grounds with an impure body and committing unprincipled acts only invites premature death" (Nihon kanrei roku 12, 74-75).42 The impure body in this story refers to the man's defilement accrued from his contact with corpses.

These concerns with death pollution extended to employees at the scriptorium. For example, the following petition submitted by a scribe in 770 or 771 shows that contact with death could be used as a reason to request time offfrom work:43

Petition (ge ...) by Oyake Doji ... regarding the reason for not going [to work]

Younger sister of the same last name with the first name of Yamonko (?) ..., age 51

The above-mentioned person [the sister] had been seriously ill since the end of last month and passed away on the thirteenth day of this month. Starting at this time, I am urgently(?) fearful of defilement [kegare wo awatadashiku osore haberu ...] so for a twelve day period, I hope to aid/cross/extinguish... (dnk 17: 561; zzs 39: 1 verso)44

The document breaks offhere, as it was cut to be reused on the other side. The character that I translate as "aid/extinguish/cross" and "urgently" is difficult to interpret. Another possibility is that the final character ... (sukuu, wataru, seisu, aid/cross/extinguish) reflects a transcription error for ... (Ch. zhai, Jp. sai), which refers to abstinential practices often performed by employees at the Office of Sutra Transcription during times of mourning. Notably, zhai were thought to have purifying effects. Regardless of these interpretive problems, it is obvious from this document that the scribe requested time offfrom work as a result of his contact with death defilement. Significantly, the scribe expressed a fear of defilement; as Okada Shigekiyo has noted, the fear of danger arising from impurity was a central motivation for ritual avoidances in early Japan (1982, 11-61).

Death was not the only source of contact defilement. Another document shows similar concerns over pollution from contact with illness:

On why I have not gone to work for so long.

With regard to the above matter, starting on the twenty-first day of the seventh month, there was an illness in my household. Everyone, both young and old, was bedridden and there was no one leftto go here and there. Only with trepidation did I not go [to work]. However, from this month, on the eighth or ninth day, [their illness] appeared to abate a little. Now, as is necessary, once I have finished cleaning and purifying my house [kansen shite, harae ...], I will return to work. So, I have recorded this letter and sent it with my son Otsutsugu. Respectfully submitted. On the fifteenth day of the eighth month of the fifth year of Hoki. (dnk 22: 589; zzs 40: 3 verso)

Here, the scribe explained his absence from work based on the need to purify his home after encountering illness. The exact means by which he would do this is unclear but it appears to have involved cleaning or washing (kansen ...) and ritual purification (harae or kaijo ...) in addition to an avoidance period.45

The idea that illness was defiling for those involved in religious practices was widespread in Nara and early Heian Japan. Legal codes of the period require that those who call upon the ill first purify themselves before engaging in state rites (Yoro ritsuryo, nst 3: 213-14; Ooms 2009, 257; Okada 1982 104-14 and 1989, 24-27). An edict issued by Emperor Tenmu ... (?-686) shows concern that the old and sick residing at temples may "defile a pure space" (Nihon shoki, nkbt 68: 438-39). Records of Numinous Responses from Japan provides further evidence; one tale mentions the need to purify oneself from illness defilement before entering a temple. In this story, a girl hurries to Gangoji when she hears that the Four Heavenly Kings enshrined at the temple might be able to cure her illness. She wanders behind the lattice where the Four Heavenly Kings stand but is soon flung out by a Yaksa. The story concludes, "Foolish woman... she ought to have proceeded only after purifying [lacuna]. Then what she wished for would not be in vain" (Nihon kanrei roku 3, 62-64).46 Contact with illness could compromise the efficacy of a ritual.

When we consider the above examples together, it seems reasonable to argue that scribes who had come into contact with defilements related to death and illness were permitted to miss work. Notably, we have over a hundred cases of scribes being excused from work due to illness and a dozen cases where they are allowed to miss work due to a death in the family (Sakaehara 1985, 87, and 1987). Surely part of the reason for excusing workers in these situations is consistent with practices in today's work environment: namely, to allow time for healing and mourning. At the same time, many requests for time offfrom mourning specifically mention the performance of abstinential rites (...; Ch. zhai, Jp. sai); we should remember that mourning in Nara Japan was not simply for the emotional well-being of the mourner but was also considered a purifying act (Okada 1982, 290-317).

Sexual Discipline

Sexual abstinence is another practice commonly associated with purity in numerous religious cultures throughout the world. In the East Asian context, abstinence was viewed as an ideal for the Buddhist and Daoist clergy.47 Narrative tales produced in Japan also point to the risks involved in sexual intercourse for scribes. A tale in the Records of the Numinous and Strange from Japan specifically discusses the relationship between sex and sutra copying.

Tajihi... the scribe was a man from the Tajihi district in Kawachi Province. Since his family name is Tajihi, we will use this as his name. Within that district, there was a place of practice called the Nonaka Hall .... There was someone who made a vow, and in the summer in the sixth month of the second year of Hoki kanotoi (771), he invited the scribe [Tajihi] to the hall to have him copy the Lotus Sutra. A group of women gathered there and added pure water [josui ...] to the ink for copying the sutra. At the time of hitsujizaru (2-4 pm), clouds suddenly gathered and rain fell. All entered the hall to escape the rain. Since the back of the hall was small and narrow, the scribe was in the same place as the women. Here, the scribe's lecherous heart kindled and he squatted behind the woman's back, lifted up her skirt, and began to fornicate. As soon as his Mara penetrated her vagina, they died together in embrace.48 The woman died foaming from the mouth.

The narrator explicitly states that the source of Tajihi's demise was related to impurity: "clearly it is known that this is a crime against the guardian of the [Buddhist] law. Even if the fires of lust flame up in one's body and heart, one should not commit a dirty act [egyo ...] based on a lustful heart" (Nihon ryoiki iii: 18, snkbt 30: 154-55). A later version of this tale is even more explicit in declaring that such dirty thoughts have no place in copying scripture:

To think about this, even if a scribe gives rise to a mind full of lust and feels that it is burning his heart, he should stop thinking these thoughts while he is copying scripture.... They truly received punishment for their crimes for defiling a temple and not believing in the scripture.

(Konjaku monogatari shu 14: 26 vol. 3, 331)

In other medieval tales, a scribe's sexual misconduct could result in the copyist and his patrons being summoned to hell (Konjaku monogatari shu 14: 29, vol. 3, 333-38, and Uji shui monogatari 8: 4, 244-45). Lust and sexual deviance could compromise the ritual act of copying scripture and have grave effects on both the patron and the scribe.

It is unclear if scribes employed in the Office of Sutra Transcription were required to be sexually abstinent. In fact, records suggest that some scribes kept families.49 Despite this, it seems plausible that copyists would have been expected to refrain from sexual activities while at the Office of Sutra Transcription. Scribes resided at the scriptorium on temple grounds and could only return home for yearly holidays, for a short vacation after completing a set of scriptures, or for emergencies. In being isolated from their families, scribes were transformed into temporary renunciants, who like their monastic brethren had "leftthe household" (shukke ...).

Beyond the Office of Sutra Transcription

Concern over scribal purity extended beyond the Office of Sutra Transcription into local Buddhist practice. For one, we have seen how many stories in the Records of the Numinous and Strange from Japan associate purity with scribes. The stories take place in multiple locales, including Kawachi and Kii provinces. The mere presence of these tales in the Records of the Numinous and Strange from Japan suggests that the discourse of purity traveled widely in the Nara and early Heian periods. This is particularly true if we accept the claims of some recent scholarship that monks traveling in the provinces likely delivered sermons based around collections of tales such as the Records of the Numinous and Strange from Japan and Records of Numinous Responses from Japan (Suzuki 1994; Kurosu 2004). If this were the case, we can expect that some provincial Buddhists would have been familiar with stories that advocate upholding purity when copying scripture.

Beyond the realm of narrative tales, however, a colophon from a 757 copy of the Maha-prajña-paramita-sutra transcribed by a local Buddhist fellowship group (chishikiyui ...) from Ebara village ... in Kawachi ... province provides more concrete evidence that those in the provinces sought out pure scribes whenever possible: "Since there was no scribe present, a defiled person [esha ...] stained [someru ...] this sutra."50 The colophon uses a clever play on words with the verb someru ..., which I have translated as "to stain." The character itself has a semantic range that includes not only "to stain" but also to defile/to dirty (kegasu) and to write/to dip one's brush (kaku). The usage of this character as defilement is common in Buddhist sources; it appears frequently in compounds such as a defiled mind (Ch. ranxin; Jp. zenshin ...) or polluted (Ch. wuran ...; Jp. osen or wazen).51 As the word at once means to defile and to write, the text of the sutra was defiled as it was written. The implication that such defilement was only performed in the absence of a professional scribe suggests that a sutra copyist was expected to be pure.

Although the connection between scribes and purity may have extended beyond the capital, this colophon suggests that some provincial patrons had trouble finding ritually pure sutra copyists. The story of the Tajihi scribe discussed above further reveals the difficulties that local patrons may have encountered in seeking out a scribe who was calligraphically proficient and ritually pure. These two examples suggest that scribal practices at the Office of Sutra Transcription and those of ad hoc fellowship groups shared common concerns over upholding ritual purity. The main difference between these two groups lies not in the ideology of regulation but rather can be found in the ability (or inability) to secure the necessary manpower to carry out a sutra-copying project.52 In this way, regulation as an ideal was by no means unique to the state. What separated the state from other forms of Buddhism was its capability to staffa large workforce of calligraphically proficient and ritually efficacious scribes.

Cultivating Purity

The discussion so far has adopted a top-down perspective focusing on the ways administrators regulated scribes' behavior to preserve ritual efficacy. In the remainder of this article, I hope to turn this view on its head and explore the effects these disciplinary regimes had on scribes. As Catherine Bell has argued, ritual creates "ritualized agents, persons who have an instinctive knowledge of these schemes embedded in their bodies, in their sense of reality, and in their understanding of how to act in ways that both maintain and qualify the complex microrelations of power" (Bell 1992, 221). In discussing the "agency" of scribes, I follow the recent work of Saba Mahmood in defining agency "not as a synonym for resistance to relations of domination, but as a capacity for action that historically specific relations of subordination enable and create" (2001b, 203). More specifically, the Office of Sutra Transcription's regulations inscribed particular habits related to purity and ritual action on scribal bodies that enabled them to progress on religious paths.

Examples of scriptorium employees who aspired to monastic ordination are revealing in this regard.53 For example, a proofreader named Murayama no Muraji Obitomaro ... applied for monastic ordination in 748. A report of his work activities notes that he performed services of worship to the Buddha (kubu raibutsu ...) (dnk 3: 286; zsbs 40 verso). Sakaehara Towao, who has studied Obitomaro in detail, argues that such practices were not undertaken by all employees; Obitomaro likely took on such responsibilities from his personal convictions (2008, 217-18). He was not alone-several other scribes requesting ordination including Manda no Muraji Emaro ..., Yamato no Fuhito Hitotari ..., and Kume no Atai Kumataka ..., all performed these services.

So why did Obitomaro become interested in Buddhist practice to begin with? Sakaehara points out that he came from the Sayama ... village in the Tajihi district of Kawachi province, a village where the eminent monk Gyoki was known to have been active. Sakaehara speculates that these activities may have planted the seeds of faith in a young Obitomaro or at least those around him. Another employee who requested ordination named Kasuga no Mushimaro ... was from the Nishinari district ... in Settsu ... province, another area where Gyoki was active.54 The background of other scribes reveals religious connections as well; for example, Hata no Yanushi ... and Iwarebe no Kumataka ... spent time working at Nakatomi-dera as attendants (toneri ...) (dnk 9: 325-326; zzs 23: 5 verso). In this way, we can see that many scribes employed at the Office of Sutra Transcription had close connections to Buddhism.

The case of Karakuni no Hitonari ... provides another example of a scriptorium employee who aspired to become a monk. Hitonari began working at the scriptorium at the age of sixteen, right around the time when the Office of Sutra Transcription was being established as part of the Queen Consort's Palace Agency. Once employed, Hitonari performed a range of tasks. He first appears as a proofreader in a document dated to 2/737 (dnk 7: 101; zzs 43: 1). During his first few years at the Office of Sutra Transcription, he served as an attendant (toneri ...), ran errands such as returning sutras to their respective owners, and checked documents for accuracy. While these activities surely occupied much of Hitonari's time, he also proofread and, most importantly for our purposes, copied sutras.55 Hitonari's administrative duties increased over time, and he eventually took on the position of supervisor (anzu ...) from around 743, when the Office of Sutra Transcription became independent from the Queen Consort's Palace Agency-he was the first individual to hold this position (Yamashita 1999, 224-28). In this role, Hitonari compiled registers and prepared other administrative documents. After eight years of labor, Hitonari applied for ordination on 8/1/745 (dnk 24: 297- 98; zzs 37: 9 verso).56

We have evidence that Hitonari was interested in Buddhism and likely prepared for ordination while employed at the Office of Sutra Transcription. For one, the Shosoin contains a copy of the first two couplets of the Thousand Character Essay believed to be written in Hitonari's hand.57 In the margins beside the main text, Hitonari writes "One hears the three treatises (sanron ...) and other [teachings] and surely believes" (zsbs 48: 5). From this, we can surmise that Hitonari may have had some familiarity with the doctrines of emptiness outlined in the three principle Madhyamaka treatises. Hitonari's application lists several titles he could chant while reading (dokkyo...) including the Sutra of the Path to the Principle (Liqu jing ...), the Sutra of the Medicine King (Yaoshi jing ...), and the Sutra on the One Who Observes the Sounds of the World (Guan shi yin jing ...). These texts reveal a range of devotional commitments and learning.58 Hitonari would have had access to a vast library of manuscripts during his time at the Office of Sutra Transcription that would have enabled him to study scripture. Hitonari's application also cites his ability to chant at Buddhist ritual services (shorai ...); it is plausible that he gained this skill through performing services of worship to the Buddha as part of his scriptorium duties. These pieces of evidence suggest that Hitonari was busy studying Buddhist scriptures during whatever free time he may have had. Moreover, he surely found himself in a community with many like-minded practitioners. For these pious employees, the pure grounds of the scriptorium would have provided an optimal environment for Buddhist practice.59

Altogether, we have records of over thirty scribes requesting ordination. Although these candidates were not all granted their requests, we must still ask why some of them would have been permitted to become monks in the first place. Was there something about their jobs that prepared them for monastic life? One clue can be found in an 11/734 edict that sets the following standards for those wanting to become a monk or nun: the ability to recite the Lotus Sutra and Golden Light Sutra, knowledge of the proper means for performing Buddhist ritual services (raibutsu ...), and a minimum of three years of pure conduct (jogyo ...)-a term that is often interpreted as sexual abstinence but likely referred to a variety of bodily practices related to upholding certain precepts (Shoku Nihongi 11/20/734, snkbt 13: 283; Ruiju sandai kyaku 2, kt 25: 75-76).60 We have already seen how the scriptorium would have offered scribes a chance to familiarize themselves with texts and practice ritual services. As should be clear by now, the cultivation of pure conduct was intimately related to a scribe's job. In fact, whereas most monastic applications from this time contain a section that records the applicant's total number of years spent in "pure conduct," scribal applications all replace the phrase "pure conduct" with "labor" (ro ...). In this way, a sutra copyist's employment functioned as a substitute for pure conduct.61 At the Office of Sutra Transcription, scribes were able to discipline their bodies to properly learn how to perform ritual services and uphold purity in a way that would enable them to pursue monastic careers. Although such practices may have been instituted as a means to empower texts to benefit the patrons, they also served to prepare scribes for a religious vocation.62

Scribes were active in religious practice outside of their official duties at the scriptorium, which suggests that many scribes took an interest in ritual and devotional activities independent of their work. In some cases, scribes copied sutras on their own (dnk 20: 54; zzs 39: 4 verso). They also requested time offfrom work to engage in Buddhist rites such as presenting oil to bodhisattvas, making offerings to the three treasures, performing penance rituals, and observing the six monthly days of purification.63 One document records a group of scribes joining together as a fellowship group to raise funds for a penance ritual (keka ...) dedicated to Amitabha (dnk 17: 111-15; zzs 46: 5). Moreover, sutra copyists participated in private (watakushi...) rites for local spirits including clan deities (ujigami...) and regional gods such as Kamo Okami (transcribed in this document as ...) (dnk 6: 171 and 407; zs 20: 11, 20, and 22; also, dnk 17: 572-73; zzs 39: 2 verso). Altogether nearly ten percent of all requests for time offare related to conducting Buddhist or kami rites. This represents the third most common reason for requesting time offand points to scribes' active involvement in religious life in diverse traditions.64 The active participation in rites by scribes is not surprising, since purity was valued for those engaging in both Buddhist and kami rites. For example, an edict from 7/17/725 declares,

In venerating kami and honoring the Buddha, purity is paramount. Now, it is said that in the shrines of the various provinces, there is much stink and defilement... how could this ever be taken as the proper way to worship the venerable kami? The chiefs of the provincial offices should themselves take up the silk offerings, diligently clean [the shrine], and hold the annual rites. Also for the grounds of the Buddhist temples, they should be thoroughly cleaned and the monks should be made to recite the Golden Light Sutra.65

The disciplinary practices of the Office of Sutra Transcription, therefore, prepared scribes to engage in both Buddhist and kami rites. In this way, disciplinary regimes purified the scribe's bodies in ways that not only served the patrons but also enabled the scribes to effectively engage in diverse ritual occupations both in Buddhist and non-Buddhist spheres.

Although we have already seen that some scribes outside of the Office of Sutra Transcription-at least as portrayed in narrative tales-may have been professional scriveners who had little interest in upholding Buddhist ideals, evidence from colophons of sutras copied by independent fellowship groups show that many provincial scribes engaged in Buddhist practice and aspired toward monastic ordination, much like employees at the Office of Sutra Transcription. Some of these groups were formed around shared cultic concerns, often based on devotion to a particular deity or text. In one case, a group joined together to copy the Yogacara-bhumi, a text intimately connected to Maitreyan worship in early Japan (Sonoda 1972). The patrons and the scribes alike called themselves "disciples of Maitreya" (jishi deshi ...). Some scribes in this group were upasaka (ubasoku ...), devout lay practitioners who in many cases upheld the five precepts and aspired to monkhood. Other members had monastic sounding names that they listed above their "original name" (honmyo ...), which suggests that they may have been novice monks who had not received full ordination (Nara cho shakyo, plates 15-16; Nara ibun, 614-15).

A single scroll from a separate project to copy the Yogacara-bhumi, currently housed at Toshodaiji ..., gives the name of a patron and scribe who were likely associated with a larger fellowship group. Both individuals use the character "Yu" ... in their names, which is the transliteration for the first sound of Yogacara. Moreover, they identify themselves as "Bodhisattvas." The rather uneven hand suggests that the scribe was not a professional scrivener but rather a devotee to the sutra who cooperated with the group out of a sense of piety and perhaps as a form of Buddhist practice (Nara cho shakyo, plate 39; Nara ibun, 619).

In a copy of the Maha-prajña-paramita-sutra, an unordained scribe with an excellent hand named Kawai no Kimi Mizomaro ... refers to himself as a "friend of the Dharma" (hoyu ...; Nara cho shakyo, plate 17; Nara ibun, 613). In another fellowship led by the monk Dogyo ..., one scribe calls himself an upasaka and a second identifies himself as a sramanera-a novice monk (Nara cho shakyo, plates 56-57). Unfortunately, in most cases, these colophons are the only extant historical record mentioning these individuals, so it is impossible to get an intimate look into their religiosity. We have seen that some scribes such as the Tajihi scrivener lacked serious religious commitments, but the presence of upasaka, novice monks, and self-proclaimed Bodhisattvas in many of these groups suggests that some sutra copyists viewed transcribing scripture as a form of religious practice that could prepare them for ordination and allow them to advance on the Buddhist path.

Conclusion

This article has shown that scribes engaged in a set of bodily practices tied to diet, bathing, dress, pollution avoidance, and sexual abstinence. These practices, often described in terms of purity, crossed wide geographic and social boundaries. From the perspective of the patron, such practices helped empower the text to ensure and enhance the creation of merit. Some pious scribes, on the other hand, cultivated pure conduct to prepare for religious careers. In concluding, I will provide a short analysis of how some of the evidence presented here may help us reassess the "state Buddhism" model.

As noted in the introduction, Nara Buddhism is often described in terms of strong regulatory measures imposed on the clergy by the state. Most scholarship that challenges this model has focused on the ineffectiveness of the authorities in regulating Buddhism and highlighted the activities of those operating outside of the state's umbrella. While recognizing the value of these recent revisions, the state's relationship to Buddhism cannot be glossed over and simply replaced with a vaguely defined "popular Buddhism." It is necessary to reassess the regulation of Buddhism by understanding both its broader context and its specific effects.

For one, the regulation of scribal practices should not be reduced to state control of Buddhism. The tendency to see state regulation in terms of secular manipulation of religious establishments reveals more about modern assumptions regarding the separation of church and state than it does about premodern Japanese religiosity. Ritual and statecraftwere inseparable in eighth-century Japan. The disciplinary regimes enforced at the scriptorium arose as a result of concerns over ensuring ritual efficacy. The scribes at the Office of Sutra Transcription were not ordained, so it is clear that the regulation of bodies was not limited to the monks and nuns but extended broadly to all those engaged in ritual. In this way, the regulation of Buddhism was perhaps less about demonstrating the effectiveness of the state's legal and administrative regime to the clergy than it was about ensuring an efficacious ritual system.

Second, efforts to ensure the ritual purity of scribes were by no means unique to the state, to Buddhism, or to the Nara period. Rather, patrons from diverse social and geographic settings, including the continent, aimed to ensure the ritual purity of lay scribes. In fact, the regulation of purity is common to many ritual systems in diverse cultural and temporal settings. Where the state was perhaps unique in comparison with provincial sutra-copying projects was in its ability to secure a constant source of ritually efficacious labor. Since the idea of regulating purity was widely shared by so-called "popular" and state movements alike, it is misleading to consider efforts at regulation as a defining feature of state Buddhism.

A new understanding of Buddhist practice in the Nara period also calls for a closer examination of the particular effects power had in shaping the agency of scribes. Proponents of the state Buddhism model often assume that power is consciously instituted by elites above and passively consumed by the masses below. In place of this two-tiered model, I have considered the specific ways that scribes experienced power through institutions such as the Office of Sutra Transcription. I have tried to show that while on the one hand scribes were disciplined to ensure ritual efficacy, these same regimes produced and enabled certain habits that proved useful for scribes in their own religious practices and careers. Through the bodily practices outlined above, scribes were able to cultivate particular dispositions that allowed them to pursue a monastic path and more effectively engage in rites for local kami. Rather than seeing discipline in a singularly passive sense, this article has explored how scribes were able to cultivate themselves within the structures of the Office of Sutra Transcription and beyond. The disciplinary regimes were mutually beneficial to both pious scribes and their patrons. This symbiotic relationship would continue after scribes received ordination, as they would be expected to perform ritual labor on behalf of the state as monks. To view scribes as agents, therefore, is not to see them as self-empowered individuals opposing repressive regimes, but rather to understand how they cultivated themselves in a manner that corresponded to historically conditioned social and cultural norms largely in agreement with the state's agenda. Such an understanding of agency provides a first step in moving beyond the facile binaries of "state" and "popular" Buddhism.

Footnote

* This article benefited greatly from the feedback I received at the Columbia Center for Japanese Religion's 2010 Purity Workshop held at Columbia University, 18-19 February, as well as from comments by Jacqueline Stone, Jolyon Thomas, and Susan Naquin.

1. My use of discipline as an analytic tool draws heavily from Talal Asad 1993, 159-67. More generally, recent work on the body that explores how specific skills and dispositions are cultivated through pedagogical practices has been helpful in framing this article. For this, see Asad 1993 (esp. 83-167) and 1997 and Mahmood 2001a and 2005. It perhaps goes without saying that Foucault's (1995) seminal study of penal systems influences my thinking on the notion of discipline. In particular, Foucault's emphasis on the contingency of power, and his argument that power structures in disciplinary regimes function to produce particular forms of knowledge, can help us move beyond binary models that reduce the relationship between state and scribe to simple exploitation. Although I have found English terms such as "discipline" and "regime" to be helpful in framing this article, I could have just as easily used Buddhist language connected to the Six Perfections (Sk. Sat-paramita; Jp. ropparamitsu ...). The second through fourth perfections of upholding precepts (Sk. sila; Jp. jikai ...), forbearance (Sk. ksanti; Jp. ninniku ...), and zealous effort (Sk. virya; Jp. shojin ...) similarly require practitioners to accept external norms and punishments while demanding vigor toward cultivating goodness and practicing the path.

2. My use of the term "agency" is in line with recent work in anthropology by scholars such as Saba Mahmood (2001b; 2005 ) and Talal Asad (1993; 2003). I will return to this issue later in the article.

3. For a more detailed analysis of how the state Buddhism model emerged in the nineteenth and twentieth centuries, see Lowe 2012, 5-34.

4. I borrow the terms "technique" and "habit" (or habitus) from a tradition of scholarship originating with Marcel Mauss. According to Mauss (2006, 149), techniques are "traditional actions combined in order to produce a mechanical, physical, or chemical effect." Although Mauss sets these ordinary actions apart from religious ones, I see the two as inseparable. With regard to the Aristotelian notion of habitus that Mauss invokes, I rely on Saba Mahmood's reading, where she defines the term as "an acquired excellence at either a moral or practical craft, learned through repeated practice until that practice leaves a permanent mark on the character of the person" (Mahmood 2005, 157-58). I have used the word "habit" in place of habitus to avoid confusion with Bourdieu's better known usage of the term.

5. I will provide Chinese transcriptions for terms appearing in sources written on the continent and Japanese transcriptions for those appearing in Japanese sources throughout this article.

6. Many of these stories are collected in the sutra copying chapters of collections such as the Fahua zhuanji (t 2068.51.80a-87c); Hongzan fahua zhuan (t 2067.51.42b-47c); and the Huayan jing zhuanji (t 2073.51.170c-172a, which are all anthologies of tales dedicated to specific sutras. For an introduction to this genre of writing, see Stevenson 1995 and 2009. There is also a similar work from Silla entitled Pop'wagyong chip'omgi that was compiled by Uijok ... (ca. 681-705). This work contains a chapter on sutra copying, but all of the stories are borrowed from tales in the above Chinese collections. At the very least, this source suggests that such tales were circulating widely on the continent. For an overview of the contents of the Pop'wagyong chip'omgi, see Takahira 2008. I would like to thank Daniel Stevenson for helping me obtain a copy of this text. Many of the stories also appear in earlier encyclopedic compilations such as the Fayuan zhulin and the Datang neidian lu.

7. I have most closely followed the version in the Fayuan zhulin (t 2122.53.421b), but also consulted the Hongzan fahua zhuan (t 2067.51.45a); Fahua zhuanji (t 2068.51.83b); Jishenzhou sanbao gantong lu (t 2106.52.428b); Datang neidian lu (t 2149.55.340b); and the Pop'wagyong chip'omgi.

8. This collection garnered attention in Japanese court circles from the 750s. It was copied based on a source text borrowed from the monk Shindo ... (dates unknown) in 751 to be used at Shima no In ..., a chapel that likely refers to the Outer Shima no In ..., a center of Kegon studies in early Japan located at Hokkeji .... After this, it was borrowed from the Office of Sutra Transcription several times, including by the well-known monk Zenju ... (723-797) and by the palace through order of the Shibichudai ..., an expanded form of the Queen Consort's Palace Agency. For the 751 copy, see dnk 3: 487 (zzs 15: 10), 529-30 (zsbs 50 verso); dnk 9: 44 (zzs 35: 6), 603 (zzs 15: 9); dnk 10: 418-19 (zzs 36: 1); dnk 11: 18 (zzs 26: 7), 159, (zzs 27: 4), 364 (zzs 37: 4), and 490 (zzs 38: 1). For later borrowing, see dnk 3: 630 (zz 50: 1), dnk 12: 261 (zzs 2: 11: 1); dnk 13: 38 (zzs 16: 4) and 194 (zzs 15: 4). As Miyazaki Kenji has pointed out, Kegon scholarship began to expand from 751 largely due to the activities of Jikin ... (sometimes read Jikun, 691-777), a monk who resided at the Outer Shima no In. For the Shima no In and Jikin's role in initiating Kegon studies in early Japan, see Miyazaki 2006, 232-39 and 347-60. In citing Shosoin documents, I will first give the published edition in Dai Nihon komonjo and then cite the photographic reproduction of the manuscript.

9. For some citations in secondary literature, see Drége 1991, 204; Kabutogi 1983, 3-8 and 117-31; and Stevenson 1995, 450-51.

10. I have relied mostly on the version in the Fahua zhuanji (t 2068.55.83b-c), but have also consulted the Jishenzhou sanbao gantong lu (t 2106.52.428a-b); Datang neidian lu (t 2149.55.340a); Fayuan zhulin (t 2122.53.416b-c); Hongzan fahua zhuan (t 2067.51.45a); and the Pop'wagyong chip'omgi. All the versions mention that the title copyist did not maintain purity, but only the Fahua zhuanji and the Fayuan zhulin specifically mention that the scribe "ate various things." The phrase "ate various things" implies meat eating.

11. Since Kabutogi is primarily interested in how the word nyoho is used, he leaves out much of the eighth century narrative, as the term nyoho seldom appears in this period. Although I do not explicitly provide a genealogy of the nyoho sutra copying practice in this article, I believe that the data in the following sections fill an important lacuna in understanding how the nyoho practice arose in Japan.

12. For later Japanese ritual texts in both the Lotus and Pure Land traditions, see Nyoho kyo genshu saho (t 2730.84.890a-898a) and Jodo sanbu kyo nyoho kyo shidai (jz 9: 370-72).

13. This general meaning of "pure conduct" as upholding the precepts is common in some of the earliest Buddhist literature. See for example the Zengyi ahan jing (t 125.2.714c). In other texts, it is closely associated with cutting offdesire, and sexual abstinence in particular. Commentaries suggest that the term continued to have both meanings of abstinence and upholding the precepts in medieval China. See Jizang's ... (549-623) Fahua yishu ..., t 1721.33.557b, for an example. Other meanings of the term are well summarized and cited in Mochizuki 1954- 1963, 5: 4678-9.

14. Of course, the association between impurity and danger is common to numerous societies throughout the world, as is famously argued in Douglas 1966.

15. For more on the way that the discourse of purity in Japan was shaped in dialogue with Buddhism, see Hongo 2005 (especially 5-52); Taira 1997; Takatori 1979; Stone 2007 and 2010.

16. For the best introduction to the Shosoin collection, see Sakaehara 2011. For an English language overview, see Farris 2007. The Shosoin corpus has been frequently used by historians, but has largely been ignored by scholars of religion. This is particularly true in discussions of purity. For example, Okada Shigekiyo's (1982) monumental survey of purity and avoidance in early Japan makes no mention of Shosoin documents.

17. Recent overviews of the institutional history of the Office of Sutra Transcription can be found in Yamashita 1999, 17-136; Sakaehara 2000 and 2003; and in English see Lowe 2012, 76-100.

18. For more on temples as pure spaces, see Hongo 2005, 5-30.

19. See Fahuan zhuanji (t 2068.51.85b); Jishenzhou sanbao gantong lu (t 2106.52.428b); Datang neidian lu (t 2149.55.340b); Fayuan zhulin (t 2122.53.486c); Hongzan fahua zhuan (t 2067.51.43b); Mingbaoji (t 2082.51:789a); and Pop'wagyong chip'omgi. In Japan, the story is cited in the Nihon ryoiki iii.10, snkbt 30: 144, and also appears in Konjaku monogatari shu 7: 18, snkbt 34: 120.

20. The breath is thought to be defiling in Buddhism. Even today, monks usually do not extinguish a candle with their breath, but instead use a snuffer or their hand. I would like to express my gratitude to Funayama Toru for mentioning this connection with breath defilement to me.

21. This was first pointed out in Ishida 1966, 204. For one document that details the diets of scribes, see dnk 13: 284-317 (zzs 8: 19). More exhaustive studies of scribal diets can be found in Ishida 1966, 210-14 and Sekine 1969. Lists for the five pungent roots vary from text to text, but a ninth-century Japanese legal commentary of the Soniryo lists garlic (ohiru or ninniku ...), asafoetida (koko or kureno'omo ...), and three types of green onion (these include wakegi or jiso ..., nira or kakuso ... (also written ...) and mehiru or ranso ...). See Ryo no shuge, kt 23: 83. This list appears to be based on the Fan wang jing (t 1484.24.1005b). Prohibitions about the five pungent roots can also be found in the Daoist tradition (Kohn 2003, 127).

22. Recent archaeological analysis of fecal parasites has confirmed that other state officials did consume meat, so the vegetarian diet appears to be particular to employees of the Office of Sutra Transcription. For the archaeological report, see Imai et al. 2010, 126-31. We also have records of meat being offered to senior officials at the Todaiji Construction Agency. Maruyama speculates that some scribes may have had the chance to share some of these fares, but she admits that this is only based on her own imagination-an imaginary vision that I see to be rather fanciful (Maruyama 2010, 238-40).

23. For a good outline of the continental practices, see Grumbach 2005, 50-70. For early indigenous practices, see San guo zhi ("Wei shu") 30.855. In this text, residents of "Wa" are described as abstaining from meat for ten days during periods of mourning and as selecting one representative who would abstain from meat when others embarked on seafaring journeys to ensure their safe return.

24. We should be careful about interpreting these prohibitions as expressions of an indigenous concern with kegare. As Harada (1993 and 2000) and Hirabayashi (2007) have argued, kegare as a general concept probably did not emerge until the mid to late ninth century-as seen in texts such as the Jogan kyakushiki ... (compiled between 820 and 868 and completed in 871; it is no longer extant but has been partially pieced together through quotations found in other sources such as the Ruiju sandai kyaku) and the Engi shiki ... (compiled in the late ninth century and presented to the emperor in 905). At the same time, concerns over purity in the more narrow contexts of ritual are constantly referred to in Nara and early Heian sources.

25. Notably, scribes appear to have consumed alcohol at the scriptorium, often under the guise of medicine. For this, see Shinmura 1985, 166-77 and Maruyama 1998, 119-44 and 2010, 197-200. Overall, Buddhism generally permits the use of alcohol as medicine. For example, see Sifen lü t no.1428, 22.672b. For the role of alcohol in Chinese Buddhism, see Michihata 1970, 214-348; Benn 2005; Liu 2008, 398-435. Numerous primary sources suggest that alcohol was frequently used as a medicine and perhaps even manufactured at temples in China and Japan. Some monks such as Saicho, however, seem to have been opposed to the practice, whereas Kukai's disciples appear to have endorsed it. For some of these diverse positions in early Japanese Buddhism, see Eizan daishiden in ddz 5 furoku: 39; Goyuigo in kdz 2: 799; and Nihon ryoiki ii.32, snkbt 30: 109-11; translated in Nakamura 1973, 203-5.

26. For example, see the tale of Yuangui discussed above. In a later Japanese tale, both the patron and the scribe alike are sent to hell after the copyist eats meat while transcribing scripture. See Konjaku monogatari shu 14: 29, vol. 3, 333-38, and Uji shui monogatari 8: 4, 244-45.

27. I should note that the Four Part Vinaya suggests monks and nuns struggled over whether or not it was acceptable to receive meat as offerings. In the above passage before the rule is declared, a nun is confronted with a slaughtered cow that a wealthy patron had prepared for the Buddha and his entourage. The nun reacts to the scene by "throwing her hands in the air, wailing, and lamenting the injustice." Her surprise and the need for the vinaya redactors to discuss the issue at all shows that meat eating was a contested topic amongst the Indian monastic community and developed in a largely pragmatic dialogue over the proper relationship between lay patrons and the clergy. We would be wrong to view the early monastic community as simply accepting meat eating.

28. For a sweeping overview of ancient and medieval Chinese bathing practices, see Schafer 1956, esp. 66-69; and Yates 1997, 513-16. Both scholars note the importance of bathing for those engaging in ritual. For bathing in the medieval Daoist tradition, see Kohn 2003 114-19. Early Chinese accounts of the people of "Wa" describe ritual bathing that occurs after funerals for purification. See the San guo zhi ("Wei shu") 30.855. Also, in the well-known story from the Kojiki, Izanagi bathes to purify himself after emerging from the land of the dead. See Kojiki, nkbt 1: 62-69 and Nihon shoki, nkbt 67: 90-95. In English, see Philippi 1969, 61-70 and Aston 1972, 24-28. For a historical survey of baths in early Japan, see Butler 2005, especially 2-4. Butler notes how temples often had baths used for purifying priests before conducting religious services.

29. Sekine notes that these loincloths only appear in documents dating up until 739 and suggests that bathing may have been characteristic of the early scriptorium. As we will see, however, other documents related to firewood and the construction of tubs for scribes show that the practice likely continued long after the loincloths disappeared.

30. It should also be noted that the construction of bathhouses is thought to be meritorious in Buddhism. For example, see Wenshi xiyu zhongseng jing (t 701.16.0802c-803c). This text was copied in Japan at the order of Roben ... (689-773) in 748 and lectures were given on it, so court patrons also likely had interest in the merit-making aspects of bathhouses; see dnk 9: 597 (zzs 19: 10 verso).

31. See dnk 5: 41 (zzs 45: 3 verso). For the nails, see dnk 15: 295 (zzs 45: 2), dnk 5: 61 (zzs 45: 2 verso), and dnk 15: 317 (zzs 44: 6). Two baths were made at this time: one for the scribes and one for the prelate (sozu ...). The fact that scribal baths were built at the same time as the prelate's shows the important role bathing was viewed to have. Roben served as the prelate at this time and was the founder of Ishiyama-dera; see Sogo bunin, dbz 65: 4.

32. For a good overview of this period in the scriptorium's history, see Yamamoto 2002, especially 548-49. I would like to thank Kuwabara Yuko for suggesting that I look into the Ishiyamadera bath. Personal correspondence, 14 January 2010.

33. Studying requests for time off(seikage ...) has been greatly facilitated through the work of the coe for Research on Formation and Characteristics of Ancient Japan ... at Nara Woman's University. They have published an annotated collection of all the Shosoin seikage. See Nara Joshi Daigaku 21 Seiki coe Puroguramu 2005-2007. Many of the seikage are also compiled in Nara ibun 2: 573-609.

34. dnk 8: 579 (zzs 44: 10). Early discussions of pure robes can be found in Ishida 1966 and Kishiro 1982.

35. Workers performing various tasks such as proofreading and assembling paper all wore pure robes at the Office of Sutra Transcription. Moreover, workers engaged in temple construction, including onmyoji performing rites at a new temple, were also issued pure robes. See dnk 25: 321 (zzs 45: 5).

36. For the rule, see dnk 17: 607 (zzs 20: 2 verso). For examples, see dnk 8: 580 (zzs 44: 10) and dnk 17: 559-60 (zzs 39: 1 verso). I have learned a great deal about the relationship between laundering pure robes and defilement discourse from Takemoto Akira's (2004) unpublished presentation on the topic and am grateful to him for sharing his handout from this talk with me.

37. See for example, dnk 4: 347 (zs 20: 2), dnk 6: 288 (zs 20: 14), dnk 17: 573 (zzs 39: 2 verso), dnk 18: 542 (zzs 3: 8 verso), and dnk 20: 56 (zzs 39: 4 verso).

38. On the category of contact defilement, see Kim 2004, 23-26

39. See the references to the Izanagi story in footnote 29 above. As Okada Shigekiyo has noted, early chronicles such as the Kojiki and Nihon shoki often refer to death as defilement (kegare ..., owai ..., aiaku ...) and describe the world of the dead as a defiled realm (kitanaki kuni or owai no kuni ...); see Okada 1982, 274-317 and 1989, 18. Even earlier references to death pollution can be found in Chinese reports of the kingdom of "Wa"; see San guo zhi ("Wei shu") 30.855. For a good overview of later Heian developments with regard to death defilement, see Kim 2004, 82-103 and Yamamoto 1992, 14-27. This is not to say that concerns over death defilement are at all uniquely Japanese. For example, the notion that corpse demons would infect a family by traveling from the deceased to the living existed in medieval China (Strickmann 2002, 37-39). For contemporary accounts of death pollution in China, see Watson 1982.

40. The Yoro administrative code banned those who were in mourning or who had inquired after illness from engaging in state rites. See Okada 1982, 97-103 and 1989, 24-25. For later developments, see Yamamoto 1992, 14-15 and Kim 2004, 77-78.

41. For later developments in Buddhism, see Stone 2007 and 2010.

42. This collection is no longer fully extant, but what remains can be found in Tsuji 1981. Tsuji provides both the original text and annotations. I refer to the collection by first giving the tale number and then the page number in Tsuji.

43. Despite the fragmentary character of the document, we can assume that it dates to 770 or early 771, as the backside of the paper for such requests for time offwere usually reused soon after being submitted. In this case the verso dates to 3/12/771.

44. I have followed the kundoku and annotations of Nara Joshi Daigaku 21 Seiki coe Puroguramu 2005-2007, 114. Readers familiar with classical Chinese will notice the odd syntax, but Japanese grammatical patterns frequently appear in Shosoin documents written in "Chinese."

45. The characters ... are frequently glossed as harae in the early chronicles of Japan. It appears that the harae was often performed at the end of the avoidance period before returning to one's regular duties. See Yamamoto 1992, 164-85 and Kim 2004, 180-87. Although Yamamoto and Kim primarily deal with later sources, their interpretation seems plausible for this example, as the petitioner submits his request six or seven days after the illness abates and states that he will perform a harae before returning to work.

46. Katsuura Noriko has recently examined this tale and reached a different conclusion. She suggests that the woman's impurity may have arisen from menstruation rather than illness. She argues that it would be a contradiction for the woman to first have to purify herself from illness in order to pray to be cured of her ailments. The story itself provides no evidence whatsoever that the woman was menstruating, and we have no sources that portray menstruating women being forbidden from temples in this period (although there is some evidence that blood in general was viewed as a pollutant). I prefer to either accept the contradiction as a contradiction or interpret "purification" as referring to a temporary purification conducted before praying. This reading is strengthened by the fact that other sources (described above) provide evidence that those who had come into contact with illness were prohibited from engaging in ritual in early Japan. For Katsuura's view, see Katsuura 2009a, 9. This article has also appeared in English in Katsuura 2009b.

47. For abstinence in the Daoist monastic context, see Kohn 2003, 119-23. For an overview of Buddhism's complex attitudes toward sexuality, see Faure 1998. For abstinence as an ideal in Chinese Buddhism in particular, see Kieschnick 1997, 17-22. There are also early depictions of abstinence amongst the people of "Wa" with regard to the person ritually responsible for the safe return of seafarers; see San guo zhi ("Wei shu") 30.855.

48. Mara, the infamous demon king, is frequently used as a euphemism for a penis in Buddhist texts.

49. Most often family members such as children and spouses are mentioned in requests for time offto mourn the dead or to care for the sick. For children, see dnk 2: 191 (zs 31: 12 verso), dnk 4: 494 (zzs 3: 4 verso), dnk 15: 99 (zzs 3: 4 verso), and dnk 17: 603-4 (zzs 39: 2 verso). For wives, see dnk 17: 598 (zzs 39: 2 verso).

50. The colophon appears in Tanaka Kaido's catalog of sutras (Tanaka 1973, 62). Unfortunately, the manuscript itself is no longer extant, as it was lost to flooding in 1953 (Gorai 1956). Without the extant manuscript, we have no way of confirming that the colophon is written in the same hand as the copyist.

51. For one classic depiction of the defiled mind, see Dasheng qixin lun (t 1666.32.577c). In Japan, this compound of defiled mind appears in the Nihon kanrei roku and is used in contrast to the term "pure mind" (shojo no kokoro ...); see Nihon kanrei roku 4, 64-65.

52. This is not to suggest that the scribes at the Office of Sutra Transcription necessarily maintained purity. It seems highly plausible that some scribes did break rules. In fact, we have one letter submitted to the Office of Sutra Transcription by a scribe, who apologizes for committing an error (ayamachi ...) that was both grave (jinshin ...) and went against decorum (reigi ...); he then expresses regret (sange ...) and promises to show more diligence (shojin ...) in the future; see dnk 6: 162 (zs 19: 6). The letter is signed by eleven employees of the Office of Sutra Transcription. Of course, we have no way to confirm whether his transgressions had anything at all to do with purity, but we should not assume that every scribe obeyed the regulations of the scriptorium.

53. See dnk 3: 78 (zs 28: 8). For more on this document, see Inoue 1966, 417-20; Sakuma 1971; Kito 1977, 115; and Horiike 1980, 1: 573-75. My own research suggests that around half of the candidates were granted their request; this number is slightly higher than these previous studies suggest

54. I have used the relatively strict criteria established by Inoue Mitsusada and Sakaehara Towao for determining areas where Gyoki was active. See Sakaehara 2006, 94-98. At least two other employees come from areas that appear in the Gyoki nenpu, but these sites cannot be definitively connected to Gyoki through comparison with other more reliable sources.

55. For a close look at his copying activities, see Lowe 2011, 28-34. For a detailed look at his life including his poetic and calligraphic output, see Lowe 2012, 132-58 and 190-95.

56. We can assume he was granted ordination because he stopped working at the scriptorium soon after his recommendation. There are a few documents listed in standard reference sources such as Takeuchi et al. (1958-1977) that record his name for dates after he was recommended for ordination. A few of these, such as dnk 24: 132 (jk 4: 2-4) and dnk 24: 368 (jk 33: 18: 2), are simply misdated in the Dai Nihon komonjo series. This is a common occurrence, but the correct dating for these documents can be found in the more recent Shosoin monjo mokuroku (Tokyo Shiryo Hensanjo 1987-). The other post-8/1/745 documents that appear in Takeuchi et al. (1958-1977), dnk 9: 382 (ss 41: 2 verso) and dnk 24: 341 (zzs 26: 6), simply record activities that Hitonari had performed in 744 and early 745. The fact that his name is blackened out on the request may also be related to the approval of his application. We have one other example of a request where a significant portion of the document is blotted out, but the character pass (go ...) appears in the margin; see dnk 24: 299-300 (zzs 34: 1 verso).

57. Naito attributes this fragment to Hitonari based on the character of his hand. Although judging hands can be a notoriously problematic endeavor, I find Naito's assessment convincing. See Naito in Shosoin jimusho 1964, 45.

58. Here, the Liqu jing most likely refers to chapter ten of the Da bore boluomiduo jing (t 220.7.986a-991b); see Kaiyuan shijiao lu (t 2154.55.651a, 662c). For the various translations of this text, their transmission to Japan, and their importance in the development of Japanese esotericism, see Abé 1999, 247-60. The Yaoshi jing is most likely the single scroll version, whose full title is Yaoshi liuliguang rulai benyuan gongde jing (t 450.14.404-409). Guan shi yin jing refers to the Guanyin chapter of the Lotus Sutra, which often circulated independently; see Miaofa lianhua jing (t 262.9.56c-58b). All of these sutras are relatively common in applications by lay practitioners for monastic ordination (ubasoku koshin no ge ...); see Horiike 1980 2: 406-13 and Yoshida 1988, 177-87.

59. Of course, we should not dwell on these "success" stories. Many scribes employed at the Office of Sutra Transcription lived lives marked by tragedies of personal loss and slaved for long hours with little chance of advancement (Sakaehara 1987, 196-202; 1991, 233-34; Farris 2007, 429). For some scribes, the scriptorium provided a place to engage in religious training, but for others, the scriptorium offered little more than a means to get by.

60. The extent to which these requirements were enforced is unclear, but at the very least they represent an ideal monk in the eyes of the state at this time.

61. The other scribal applications only list the number of years engaged in labor but do not list texts studied or other ritual abilities. This is true for all post-745 recommendation letters. Notably, Hitonari does not list the Lotus Sutra or the Golden Light Sutra-which was required by the 734 edict discussed above. This may be related to the fact that his application comes right at the point when the application was simplified and requirements for sutra recitation may have been dropped. Japanese scholars have studied this simplification of the recommendation process in detail. Increased demand for manpower for the construction of Todaiji, changes in state policy, and institutional shifts in the place where these recommendations were processed all contributed to this change. Good introductions to the genre of monastic applications can be found in Horiike 1980, 1: 566-82; Kito 1977; Nemoto 1974 and 1976; Nakabayashi 1993; Sakuma 1971; Sato 1993; and Yoshida 1988, 155-86.

62. I should note than in pursuing a religious vocation, scribes would be furthering imperial interests, as performing rituals for the protection of the realm and the health of the emperor was a central feature of a monastic career. This is in line with the two senses of regime described in the introduction to this article.

63. For offering oil to bodhisattvas, see dnk 17: 557 (zzs 29: 1 verso) and dnk 14: 178 (zzs 34 verso). For offerings to the three treasures, see dnk 17: 598 (zzs 29: 2 verso). For penance rites, see dnk 17: 5888 (zzs 39: 2 verso). For abstinential rites, see dnk 15: 90 (zzs 3: 4: 3: 12 verso) and 17: 595 (zzs 29: 2 verso). Okusa Hiroshi (2010) has studied abstinential practices in detail and reached similar conclusions that many scribes were actively interested in Buddhist practice; see also Minowa 1999.

64. The most common reasons for requesting time offare those due to illness and after the completion of a set of scriptures (Sakaehara 1985, 87).

65. Shoku Nihongi 7/17/725, snkbt 13: 160-61.

References

references

abbreviations

bd Beijing Dunhuang manuscript. Housed in the National Library of China. Reproductions in Guojia tushuguan cang Dunhuang yishu .... 146 vols. Beijing: Beijing tushuguan chubanshe, 2005-2012.

dbz Dai Nihon Bukkyo zensho .... 100 volumes. Tokyo: Kanko Suzuki Gakujutsu Zaidan, 1970-1973.

ddz Dengyo Daishi zenshu .... 5 vols. Tokyo: Sekai Seiten Kanko Kyokai, 1989.

dnk Dai Nihon komonjo: Hennen monjo .... 25 vols. Tokyo: Tokyo Teikoku Daigaku, 1901-1940.

jk Jinkai .... Published in Shosoin komonjo eiin shusei..., 17 vols. Tokyo: Yagi Shoten, 1988-.

jz Jodo shu zensho .... 23 vols. Tokyo: Sankibo Busshorin, 1970- 1972.

kdz Kobo Daishi zenshu .... 8 vols. Koyasan: Mikkyo Bunka Kenkyujo, 1965-1968.

kt Kokushi taikei.... 60 vols. Tokyo: Yoshikawa Kobunkan, 1964- 1967.

n Nakamura Fusetsu Collection. Reproduced in Taito Kuritsu Shodo Hakubutsukan shozo Nakamura Fusetsu kyuzo uiki bokusho shusei .... 3 vols. Tokyo: Sokatsuhan, 2005.

nkbt Nihon koten bungaku taikei .... 100 vols. Tokyo: Iwanami Shoten, 1958-1966.

nst Nihon shiso taikei .... 67 vols. Tokyo: Iwanami Shoten, 1970- 1982.

p Manuscript from the Pelliot Collection. Housed in the Bibliothèque nationale de France with digital editions on http://idp.bl.uk. Also published in Faguo guojia tushuguan cang Dunhuang Xiyu wenxian ... . 34 vols. Shanghai: Shanghai guji chubanshe, 1994-2004.

snkbt Shin Nihon koten bungaku taikei .... 100 vols. Tokyo: Iwanami Shoten, 1989-2005.

ss Seishu .... Published in Shosoin komonjo eiin shusei .... 17 vols. Tokyo: Yagi Shoten, 1988-.

t Taisho shinshu daizokyo .... 85 vols. Tokyo: Taisho Issaikyo Kankokai, 1924-1932.

zs Zokushu .... Published in Shosoin komonjo eiin shusei ... . 17 vols. Tokyo: Yagi Shoten, 1988-.

zsbs Zokushu besshu .... Published in Shosoin komonjo eiin shusei .... 17 vols. Tokyo: Yagi Shoten, 1988-.

zsks Zokushu koshu .... Published in Shosoin komonjo eiin shusei .... 17 vols. Tokyo: Yagi Shoten, 1988-.

zzs Zokuzokushu .... Published in Shosoin monjo .... 16 reels.

primary sources and collections

Banzhou sanmei jing .... t vol. 13, no. 418.

Da banniepan jing .... t vol. 12, no. 374.

Da baoji jing .... t vol. 11, no.310.

Da bore boluomiduo jing .... t vols. 5.7, no. 220.

Dai Nihon komonjo: Hennen monjo. 25 vols. Ed. T.ky. Daigaku Shiry. Hensanjo ... S. Tokyo: T.ky. Teikoku Daigaku, 1901.1940.

Dasheng qixin lun .... t vol. 32, no. 1666.

Datang neidian lu... t vol. 55, no. 2149.

Eizan daishiden ... B. ddz, vol. 5.

Engi shiki ... kt 26.

Fahua yishu ... vol. 33, no. 1721.

Fahua zhuanji .... t vol. 51, no. 2068.

Fan wang jing .... t vol. 24, no. 1484.

Fayuan zhulin .... t vol. 53, no. 2122.

Goyuig. .... kdz, vol. 2.

Gyoki nenpu .... In Zokuzoku gunsho ruij. ...] 3: 428.37. Tokyo: Heibonsha, 1969.

Hongzan fahua zhuan .... t vol. 51, no. 2067.

Huayan jing zhuanji .... t vol. 51, no. 2073.

Jishenzhou sanbao gantong lu ... t vol. 52, no. 2106.

J.do sanbu ky. nyoh. ky. shidai ... jz vol. 9, 370.72.

Kaiyuan shijiao lu .... t vol. 55, no. 2154.

Kojiki .... nkbt, vol. 1.

Konjaku monogatari sh. ... W. snkbt, vols. 33.37.

Miaofa lianhua jing .... t vol. 9, no. 262.

Mingbaoji .... t vol. 51, no. 2082.

Mohe bore boluomi jing .... t vol. 8, no. 223.

Nara ch. shaky. .... Ed. Nara Kokuritsu Hakubutsukan ....

Nara: Nara Kokuritsu Hakubutsukan, 1983.

Nara ibun .... 3 vols. Ed. Takeuchi Riz. .... Tokyo: Toky.d., 1962.

Nihon kanrei roku .... In Nihon kanrei roku no kenky. ..., Ed. Tsuji Eiko ....Tokyo: Kasama S.sho, 1981.

Nihon ry.iki .... snkbt, vol. 30.

Nihon shoki .... nkbt, vols. 67.68.

Nyoh. ky. genshu sah. .... t vol. 84, no. 2730.

Pop'wagyong chip'omgi .... In Tokyo daigaku toshokan zo: Hokekyo shogenki .... Tokyo: Kich. Kotenseki Kank.kai, 1981.

Ruiju sandai kyaku .... kt 25.

Ru lengqie jing ... t vol. 16, no. 671.

Ryo no shuge .... kt 23.

San guo zhi.... Beijing: Zhonghua Shu Ju, 1959.

Shoku Nihongi .... snkbt, vols. 12.16.

Shosoin komonjo eiin shusei .... 17 vols. Tokyo: Yagi Shoten, 1988-.

Sifen ... t vol. 22, no.1428.

Sogo bunin .... dbz vol. 65, no. 484.

Uji shui monogatari.... nkbt, vol. 27.

Wenshi xiyu zhongseng jing .... t vol. 16, no. 701.

Yaoshi liuliguang rulai benyuan gongde jing .... t vol. 14, no. 450.

Yoro ritsuryo .... nst, vol. 3.

Zengyi ahan jing .... t vol. 2, no. 125.

secondary sources

Abé, Ryuichi

1999 The Weaving of Mantra: Kukai and the Construction of Esoteric Buddhist Discourse. New York: Columbia University Press.

Asad, Talal

1993 Genealogies of Religion: Discipline and Reasons of Power in Christianity and Islam. Baltimore: Johns Hopkins University Press.

1997 Remarks on the anthropology of the body. In Religion and the Body, ed. Sarah Coakley, 42-52. Cambridge: Cambridge University Press.

2003 Formations of the Secular: Christianity, Islam, Modernity. Stanford: Stanford University Press.

Aston, W. G.

1972 Nihongi: Chronicles of Japan From the Earliest Times to A.D. 697. Rutland, VT: C. E. Tuttle Co. (original edition 1896)

Bell, Catherine

1992 Ritual Theory, Ritual Practice. Oxford: Oxford University Press.

Benn, James A.

2005 Buddhism, alcohol, and tea in medieval China. In Of Tripod and Palate: Food, Politics, and Religion in Traditional China, ed. Roel Sterckx, 213-36. New York: Palgrave MacMillan.

Butler, Lee

2005 Washing offthe dust: Baths and bathing in late medieval Japan. Monumenta Nipponica 66: 1-41.

Como, Michael

2009 Weaving and Binding: Immigrant Gods and Female Immortals in Ancient Japan. Honolulu: University of Hawai'i Press.

Douglas, Mary

1966 Purity and Danger. London and Boston: Routledge and Kegan Paul.

Drége, Jean-Pierre

1991 Les Bibliothéques en Chine au temps des Manuscrits. Paris: École Française d'extreme-orient.

Farris, William Wayne

2007 Pieces in a puzzle: Changing approaches to the Shosoin documents. Monumenta Nipponica 62: 397-435.

Faure, Bernard

1998 The Red Thread: Buddhist Approaches to Sexuality. Princeton: Princeton University Press.

Foucault, Michel

1995 Discipline and Punish: The Birth of the Prison. New York: Vintage Books. (originally published 1977)

Gorai Shigeru ...

1956 Kishu Hanazono mura Daihannyakyo no shosha to ruden ... . Otani shigaku 5: 1-23.

Grumbach, Lisa

2005 Sacrifice and Salvation in Medieval Japan: Hunting and Meat in Religious Practice at Suwa Jinja. PhD dissertation, Stanford University.

Harada Nobuo ...

1993 Rekishi no naka no kome to niku: Shokumotsu to tenno, sabetsu ... . Tokyo: Heibonsha.

2000 Kodai Nihon no dobutsu kugi to sessho kindan: Noko girei to nikujiki kinki o megutte ... . Tohoku gaku 3: 150-77.

Hirabayashi Akihito ...

2007 Kamigami to nikushoku no kodaishi .... Tokyo: Yoshikawa Kobunkan.

Hongo Masatsugu ...

2005 Ritsuryo kokka Bukkyo no kenkyu .... Kyoto: Hozokan.

Horiike Shunpo ...

1980 Nanto Bukkyo shi no kenkyu .... 3 vols. Kyoto: Hozokan.

Imai Koki ..., Kunitake Sadakatsu ..., and Kaneda Akihiro ...

2010 Toho kanga chiku no chosa: Dai 440, 446 ji ... 440-446.... Nara bunkazai kenkyujo kiyo 2010: 126-35.

Inoue Kaoru ...

1966 Naracho Bukkyoshi no kenkyu .... Tokyo: Yoshikawa Kobunkan.

Inoue Mitsusada ...

1971 Nihon kodai no kokka to Bukkyo .... Tokyo: Iwanami Shoten.

Ishida Mosaku ...

1966 Shakyo yori mitaru Naracho Bukkyo no kenkyu ... . Tokyo: Toyo Bunko. (original edition 1930)

Kabutogi Shoko ...

1983 Hokke shakyo no kenkyu .... Tokyo: Daito Shuppansha.

Katsuura Noriko ...

1995 Onna no shinjin: Tsuma ga shukkeshita jidai .... Tokyo: Heibonsha.

2009a Josei to kegare kan .... Bukkyo shigaku kenkyu 51: 1-20.

2009b Women and views of pollution. Acta Asiatica 97: 17-37.

Kieschnick, John

1997 The Eminent Monk: Buddhist Ideals in Medieval Chinese Hagiography. Honolulu: University of Hawai'i Press.

2005 A history of Buddhist vegetarianism in China. In Of Tripod and Palate: Food, Politics, and Religion in Traditional China, ed. Roel Sterckx, 186-212. New York: Palgrave MacMillan.

Kim, Jayne Sun

2004 A History of Filth: Defilement Discourse in Medieval Japan. PhD dissertation, Columbia University.

Kishiro Shuichi ..

1982 Nihon bunka no shuhen .... Tokyo: Yuzankaku. (original edition 1932)

Kito Kiyoaki ...

1977 Nihon kodai toshiron josetsu .... Tokyo: Hosei Daigaku Shuppankyoku.

Kohn, Livia

2003 Monastic Life in Medieval Daoism: A Cross-Cultural Perspective. Honolulu: University of Hawai'i Press.

Kurosu Toshio ...

2004 Kodai ni okeru kudoku to shite no seiso: Nihon ryoiki jokan dai jusan en no ichikosatsu ... . In Nara Bukkyo to zaichi shakai ..., Nemoto Seiji ... and Samuel Crowell Morse ..., eds., 155-78. Tokyo: Iwata Shoin.

Liu Shufen ...

2008 Zhonggu de Fojiao yu shehui .... Shanghai: Shanghai Guji Chuban She.

Lowe, Bryan D.

2011 Texts and textures of early Japanese Buddhism: Female patrons, lay scribes, and Buddhist scripture in eighth-century Japan. Princeton University Library Chronicle 73: 9-36.

2012 Rewriting Nara Buddhism: Sutra Transcription in Early Japan. PhD dissertation, Princeton University.

Mahmood, Saba

2001a Rehearsed spontaneity and the conventionality of ritual: Disciplines of Salat. American Ethnologist 28: 827-53.

2001b Feminist theory, embodiment, and the docile agent: Some reflections on the Egyptian Islamic revival. Cultural Anthropology 16: 202-36.

2005 Politics of Piety: The Islamic Revival and the Feminist Subject. Princeton: Princeton University Press.

Maruyama Yumiko ...

1998 Nihon kodai no iryo seido .... Tokyo: Meicho Kankokai.

2010 Shosoin monjo no sekai: Yomigaeru Tenpyo no jidai ... . Tokyo: Chuo Koron Shinsha.

Mauss, Marcel

2006 Techniques, Technology and Civilization. Ed. Nathan Schlanger. New York: Durkheim Press/Berghahn Books.

Meeks, Lori R.

2010 Hokkeji and the Reemergence of Female Monastic Orders in Premodern Japan. Honolulu: University of Hawai'i Press.

Michihata Ryoshu ...

1970 Chugoku Bukkyo shi no kenkyu .... Kyoto: Hozokan.

Minowa Kenryo ...

1999 Nihon kodai ni okeru hassaikai no juyo .... Indogaku bukkyogaku kenkyu 48: 28-34.

Miyazaki Kenji ...

2006 Nihon kodai no shakyo to shakai .... Tokyo: Hanawa Shobo.

Mochizuki Shinko ... and Tsukamoto Zenryu ...

1954-1963 Bukkyo daijiten .... 10 vols. Tokyo: Sekai Seiten Kanko Kyokai.

Moerman, D. Max

2007 The archeology of anxiety: An underground history of Heian religion. In Centers and Peripheries in Heian Japan, Mikael S. Adolphson, Edward Kamens, and Stacie Matsumoto, eds., 245-72. Honolulu: University of Hawai'i Press.

2010 The death of the Dharma: Buddhist sutra burials in early medieval Japan. In The Death of Sacred Texts: Ritual Disposal and Renovation of Texts in World Religions, ed. Kristina Myrvold, 71-90. Aldershot, UK: Ashgate.

Nakabayashi Takayuki ...

1993 Ubasoku(i) koshin seido no tenkai.... Shosoin monjo kenkyu 1: 40-80.

Nakamura, Kyoko Motomochi

1973 Miraculous Stories From the Japanese Buddhist Tradition: The Nihon Ryoiki of the Monk Kyokai. Cambridge, MA: Harvard University Press.

Nakano, Chieko

2009 "Kechien" as Religious Praxis in Medieval Japan: Picture Scrolls as the Means and Sites of Salvation. PhD dissertation, University of Arizona.

Nara Joshi Daigaku 21 Seiki coe Puroguramu ...

2005-2007 Shosoin monjo no kundoku to chushaku: Seika fusan ge hen ... , vols. 1 and 2. Nara: Nara Joshi Daigaku 21 Seiki coe Puroguramu.

Nemoto Seiji ...

1974 Ubasoku koshin no ge ni tsuite ... 17: 41-55.

1976 Kodai ni okeru ubasoku ubai ni tsuite: Toku ni tenpyo junana nenki o chushin to shite .... In Nihon shi ni okeru minshu to shukyo..., ed. Shimode Sekiyo ..., 55-83. Tokyo: Yamakawa Shuppansha.

Nishiguchi Junko ...

1987 Onna no chikara: Kodai no josei to Bukkyo .... Tokyo: Heibonsha.

Okada Shigekiyo ...

1982 Kodai no imi: Nihonjin no kiso shinko .... Tokyo: Kokusho Kankokai.

1989 Imi no sekai: Sono kiko to hen'yo .... Tokyo: Kokusho Kankokai.

Okusa Hiroshi ...

2010 Shakyosei to Bukkyo .... Otani daigaku daigakuin kenkyu kiyo 27: 75-103.

Ooms, Herman

2009 Imperial Politics and Symbolics in Ancient Japan: The Tenmu Dynasty, 650- 800. Honolulu: University of Hawai'i Press.

Philippi, Donald L.

1969 Kojiki. Princeton: Princeton University Press.

Sakaehara Towao ...

1985 Miyako no kurashi .... In Kodai o kangaeru: Nara ..., ed. Naoki Kojiro ..., 77-120. Tokyo: Yoshikawa Kobunkan.

1987 Heijokyo jumin no seikatsushi .... In Tojo no seitai..., ed. Kishi Toshio ..., 187-266. Tokyo: Chuo Koronsha.

1991 Tenpyo no jidai .... Tokyo: Shueisha.

2000 Nara jidai no shakyo to dairi.... Tokyo: Hanawa Shobo.

2003 Nara jidai shakyoshi kenkyu .... Tokyo: Hanawa Shobo.

2006 Ono-dera no doto no chishiki .... In Kodai Izumi-gun no rekishiteki tenkai ..., 93-110. Izumi: Izumi Shishi Hensan Iinkai.

2008 Gyoki to Nakatomi kei shizoku: Ise shinko to Bukkyo .... In Chiiki no naka no kodaishi ..., ed. Noda Reishi ..., 203-23. Tokyo: Iwata Shoin.

2011 Shosoin monjo nyumon .... Tokyo: Kadokawa Gakugei Shuppan.

Sakuma Ryu ...

1971 Ubasoku koshin no ge no hitobito .... Shinano 23: 15-24.

Sato Fumiko ...

1993 Ubasoku koshin no jitsuzo to sono shiteki igi ... . Shiso ... 50: 58-76.

Schafer, Edward H.

1956 The development of bathing customs in ancient and medieval China and the history of the Floriate Clear Palace. Journal of the American Oriental Society 76: 57-82.

Sekine Shinryu ...

1969 Naracho shokuseikatsu no kenkyu .... Tokyo: Yoshikawa Kobunkan.

1974 Nara cho fukushoku no kenkyu .... Tokyo: Yoshikawa Kobunkan.

Shinmura Taku ...

1985 Nihon iryo shakaishi no kenkyu: Kodai chusei no minshu seikatsu to iryo .... Tokyo: Hosei Daigaku Shuppankyoku.

Shosoin jimusho ... ed.

1964 Shosoin no shoseki .... Tokyo: Nihon Keizai Shinbunsha.

Sonoda Koyu ...

1972 Chishiki to kyoke: Kodai ni okeru shuhasei no kigen .... In Kokushi ronshu ..., ed. Akamatsu Toshihide Kyoju Taikan Kinen Jigyokai ..., 243-59. Kyoto: Akamatsu Toshihide Kyoju Taikan Kinen Jigyokai.

Stevenson, Daniel B.

1995 Tales of the Lotus Sutra. In Buddhism in Practice, ed. Donald S. Lopez, 427-51. Princeton: Princeton University Press.

2009 Buddhist practice and the Lotus Sutra in China. In Readings of the Lotus Sutra, Stephen F. Teiser and Jacqueline I. Stone, eds., 132-50. New York: Columbia University Press.

Stone, Jacqueline I.

2007 The dying breath: Deathbed rites and death pollution in early Medieval Japan. In Heroes and Saints: The Moment of Death in Cross-Cultural Perspectives, Phyllis Granoffand Koichi Shinohara, eds., 173-246. Newcastle, UK: Cambridge Scholars.

2010 Do kami ever overlook pollution? Honji suijaku and the problem of death defilement. Cahiers d'Extrême Asie 16: 203-32.

Strickmann, Michel

2002 Chinese Magical Medicine. Stanford: Stanford University Press.

Suzuki Keiji ...

1994 Tohikan kotsu to zaichi chitsujo: Nara-Heian shoki no Bukkyo o sozai to shite .... Nihonshi kenkyu 379: 34-59.

Taira Masayuki ...

1997 Sessho kindan no rekishiteki tenkai .... In Nihon shakai no shiteki kozo-Kodai, chusei ..., ed. Oyama Kyohei Kyoju Taikan Kinenkai ..., 149-71. Kyoto: Shibunkaku Shuppan.

Takahira Myoshin ...

2008 Hokke kyo shugenki ni kansuru ichikosatsu .... Indogaku bukkyogaku kenkyu 56: 668-71.

Takatori Masao ...

1979 Shinto no seiritsu .... Tokyo: Heibonsha.

Takemoto Akira ...

2004 Shakyojo kara mita kegare kan .... Unpublished presentation delivered at the Shosoin monjo kenkyukai ..., Osaka, 30 October.

Takeuchi Rizo ..., Yamada Hideo ..., and Hirano Kunio ..., eds.

1958-1977 Nihon kodai jinmei jiten .... Tokyo: Yoshikawa Kobunkan.

Tanaka Kaido ...

1973 Nihon koshakyo genson mokuroku .... Kyoto: Shibunkaku.

Tokyo Shiryo Hensanjo ..., eds.

1987- Shosoin monjo mokuroku .... Tokyo: Tokyo Daigaku Shuppankai.

Tsuji Eiko ...

1981 Nihon kanrei roku no kenkyu .... Tokyo: Kasama Shoin.

Watson, James L.

1982 Of flesh and bones: The management of death pollution in Cantonese society. In Death and the Regeneration of Life, Maurice Bloch and Jonathan P. Parry, eds., 155-86. Cambridge: Cambridge University Press.

Yamamoto Koji ...

1992 Kegare to oharae .... Tokyo: Heibonsha.

Yamamoto Yukio ...

2002 Shakyojo monjo no kisoteki kenkyu .... Tokyo: Yoshikawa Kobunkan.

Yamashita Yumi ...

1999 Shosoin monjo to shakyojo no kenkyu .... Tokyo: Yoshikawa Kobunkan.

Yates, Robin D. S.

1997 Purity and pollution in early China. In Zhongguo kaoguxue yu lishixue zhi zhenghe yanjiu ..., 479-541. Taipei: Academica Sinica.

Yoshida Kazuhiko ...

1995 Nihon kodai shakai to Bukkyo .... Tokyo: Yoshikawa Kobunkan.

2006a Kodai Bukkyo o yominaosu .... Tokyo: Yoshikawa Kobunkan.

2006b Religion in the classical period. In Nanzan Guide to Japanese Religions, Paul L. Swanson and Clark Chilson, eds., 144-62. Honolulu: University of Hawai'i Press.

Yoshida Yasuo ...

1988 Nihon kodai no bosatsu to minshu .... Tokyo: Yoshikawa Kobunkan.