Miracles of Book and Body Buddhist Textual Culture and Medieval Japan

ONE

The Ontology of Sutras

The final chapter of the Lotus Sutra opens on a very curious scene. The bodhisattva Fugen (Sk: Samantabhadra), long abiding in the eastern quarter of the cosmos, has heard that a buddha is preaching the Lotus Sutra on Earth. He arrives, with a multitude of beings trailing him, at the foot of the historical Buddha. After circumambulating the Buddha seven times, he announces, "I have come to listen receptively. I beg of the World Honored One to preach [the Lotus Sutra] to us!" He immediately follows this request with a question: "After the extinction of the Thus Come One, how may a good man or good woman attain this" Lotus Sutra? The Buddha provides a brief answer to Fugen's question, and the bulk of the remainder of the chapter is then given over to the various vows Fugen makes to guard and protect anyone who accepts, keeps, reads, recites, copies, explains, or practices this scripture. In the course of these lengthy praises, Fugen goes so far as to maintain that if a person "forgets a single phrase or a single gāthā [verse] of the Lotus Sutra, I will teach him, reading and reciting together with him," so that he is able to remember it.

What is curious about this scene is that Fugen asks for something he clearly does not need. He arrives in great excitement and requests to hear a sutra that he presumably has not heard before. And yet a mere paragraph later-and still having heard nothing of the sutra-he vows to help other beings study the sutra, which he clearly already has completely memorized, down to the last phrase and verse. In the tiny gap of white space between Fugen's request and his question there is narrative slippage, a textual fissure that allows us to peer more deeply into the motivating concerns and pivotal themes of the sutra. On a narrative level, Fugen's false request produces a litany of desperation rather than the expected preaching of the Lotus Sutra. Each of Fugen's vows begins with the prefatory phrase, "In the last five hundred years, in the midst of a muddied, evil age ... if there is anyone ..."

Thus, Fugen's dramatic last-minute appearance underscores two fundamental things about the Lotus Sutra. First, it will be exceedingly difficult for any of us who live in this degenerate later age to encounter the sutra, and, second, we can never be certain that the text we have encountered is precisely the Lotus Sutra because what we have of the sutra consists largely of narrative about the sutra: we hear of its powers and its rarity, we hear bodhisattvas and buddhas from other lands praise it, we hear what happens to those who slander it, and so forth. Many of these narrative gestures concentrate on increasing readerly desire while constantly obscuring the object of that desire. Is the sutra called the Lotus Sutra-the book I can hold in my hands and read aloud-the same thing as the teaching called the Lotus Sutra-the thing Fugen was so eager to hear? I would argue that by denying the simple satisfaction of readerly desire, texts like these involve their readers in an insatiably desirous relationship to writing and, in doing so, thematize the reader-text relationship, marking this as one of the crucial concerns that the text is trying to explore and for which it is trying to script the rules of engagement.

Other scholars, on whose work I build here, have explored the various complicated and compelling ways that Mahāyāna sutras cut across oral and written modes of discourse, navigate between aural and visual sense worlds, and perhaps even disrupt various norms of mainstream Buddhist traditions. This last point, concerning a distinction between Mahāyāna and "mainstream" Buddhism, is tied into an abiding concern in modern scholarship with the origins of the Mahāyāna in India-not the intellectual focus of this book, but an issue that requires some clarification before we can move forward.

Imagining the Origins of the Mahāyāna

The last half century has seen multiple attempts to theorize or generalize the origins of the Mahāyāna and to define the core concerns that caused it eventually to split from the Buddhist mainstream. Survey materials, textbooks, and encyclopedia entries often characterize the Mahāyāna as a reform movement that arose out of "dissatisfaction" with perceived "shortcomings in the Theravada tradition," such as a primary concern with "individual salvation" rather than the liberation of all beings. Hirakawa Akira, starting in the 1950s, offers a second approach, stressing the importance of laity and the centrality of stupa worship in the formation of the early Mahāyāna. Though his theses have been heavily criticized and largely disproved, his work may be credited with generating sustained academic conversation about the shape of early Mahāyāna. Responding to Hirakawa, several scholars, particularly beginning in the 1980s and 1990s, advanced a third argument: that the Mahāyāna was "the work of a predominantly monastic order of meditators engaged in strenuous ascetic practices, people asserting, in short, that the Buddha is to be found in and through the realization of the dharma, not the worship of relics." Importantly, this third position re-centers the focus, from stupa to dharma, and reasserts the centrality of the Buddhist teachings as contained in the written sutras. In this sense, it bears a surface similarity to Gregory Schopen's early work on the "cult of the book," which provides a fourth angle on Mahāyāna origins: that the Mahāyāna distinguished itself by establishing cultic centers organized not around stupas, but rather around written sutra texts that were recited, worshipped, honored, and circumambulated.

It is absolutely essential, however, to decouple the question of origins from the culture of books. In his later work, Schopen moves beyond the analysis of the linguistic record (what sutras say) to an analysis of the material record (what given, datable sutra texts look like). He points out that sources do provide evidence of a working "book-cult" in India-manuscripts whose "covering boards or first leaves were often heavily stained and encrusted from continuous daubing with unguents and aromatic powders," for instance-but that this "evidence is almost a thousand years later than it should be" (at least according to his earlier thesis). That is, although the linguistic text of sutras, composed between the first and fifth centuries C.E., seems to call for book worship, material evidence for such worship dates to no earlier than the eleventh to fifteenth centuries C.E. in India. Arguing in a similar vein, Jan Nattier has noted that passages concerning sutra worship are often interpolations, sections and phrases that were stitched into the sutras in their later versions, and she concludes that "the emergence of the Mahāyāna ... does not begin with the cult of the book, but rather culminates in it at some point." The book-cult, then, while perhaps not part of the story of Mahāyāna origins, is nevertheless an aspect of Mahāyāna that became increasingly important over time, and which achieved particular significance in Chinese-language translations of the sutras and in East Asian cultures (such as Japan) that relied on these translations. An examination of the Mahāyāna "cult of the book" thus will tell us not about the Great Vehicle's origins in India but about the movement in some of its localized, East Asian forms.

In addition to distinguishing between the early years of Indian origins and later developments in the sinophone Buddhist sphere, we also need to make a distinction between "what sutras want" and "what sutras were given." Answering the question of what sutras want (for the purposes of this study) is a matter of examining what they ask for, in the Chinese translations that were consulted most often in medieval Japan-what they ask for, but may or may not have gotten. Inquiring into what sutras were given is a matter of examining the material record, which will provide evidence concerning when, where, and how sutras' desires were actualized. In response to the first question-"What do sutras want?"-we know that the Chinese translations of Mahāyāna sutras that were circulating in medieval Japan asked for a wide variety of things, including worship, circumambulation, and offerings; they asked to be read, recited, memorized, and held. In response to the second question-"What were sutras given?"-we know that by at least the tenth century in Japan they were being worshipped, read, recited, copied, expounded upon in sermons, and memorized. A similar array of praxes surrounding Mahāyāna sutras may have developed as early as the third or fourth century in China and as late as the eleventh to fifteenth centuries in India. As will become clear, throughout this study references to sutras treat only the first of these questions, the one concerning the abstract, idealized nature of desire as discernable in the linguistic record, while I use Japanese sources (texts written in Japan, using either the Japanese or the Chinese language) to examine questions about the actualization of sutras' desires and traces these actualizations leave in the material record.

To be clear, then, I come to Mahāyāna sutras specifically from the vantage point of medieval Japanese explanatory tales (setsuwa). My general purpose in this and succeeding chapters is to outline the textual attitudes of a certain body of literature that began in India, was translated into Chinese, and was then reinterpreted in Japan. Rather than speaking of the Indian origins of the Mahāyāna, I am interested in the idea of Mahāyāna textual culture as constructed by the authors of Japanese setsuwa. As a medieval genre, setsuwa are an often-miraculous form of literature that assumes sutras are living beings who may incorporate themselves into human form. Thus, my main interest in sutras has to do with exploring three questions. First, how did sutras come to be alive, that is, what narrative tools propel them into life? Second, now that they are alive, what do they want: as living beings, what are their requirements, their equivalent of food and shelter? And finally, what is the nature of the reader-text relationship they propose; in other words, what happens to human beings who decide to give sutras what they ask for? In short, I am interested in examining the rules and the stakes of sutra engagement and how we, as humans and as readers, play into them.

First, how is it that sutras come alive? Mahāyāna sutras use a variety of literary techniques that modern readers might most readily associate with metafiction. Taken collectively, these metafictional strategies impact the ontological status of Mahāyāna sutras in some important ways. The issue of textual ontology can be generally glossed with the oft-quoted question: "If the Mona Lisa is in the Louvre, where [is] Hamlet?" In other words, where, how, in what plane of action does a literary text exist? I suggest that by obscuring the authorial hand and complicating any notion of a simple origin in oral discourse, Mahāyāna sutras cut themselves off from their point of creation and establish a unique ontology for themselves. Instead of mooring themselves to an external source, as most literature does, Mahāyāna sutras seek to take the authority that might otherwise be invested in an author or in a speaker and invest it in themselves. We see this most plainly in the argument that sutras contain the entire body of the Buddha, and that they are, in fact, the entity that gives birth to buddhas. Mahāyāna sutras thus frame themselves as textual constructs that are alive and eerily self-aware.

Moving forward, now that sutras have a certain sentience, what do they want? I argue that sutras are, in essence, a nervous genre. Aware of their vulnerabilities to the ravages of time and convinced they are always in danger of dying, they evince a strong reproductive drive. Mahāyāna sutras are careful to include within themselves instructions about how they should be propagated, promises of reward for those who agree to do the propagating, and threats to inflict pain on those who do not comply. In short, having successfully placed themselves in a position of agency, sutras speak from this position, demanding very particular things from their readers, only one of which is actually to be read.

Finally, what is the nature of the reader-text relationship that Mahāyāna sutras propose? Here I explore the persistent rhetoric of the "vessel" or "container" through which the mind (Jp: kokoro) of the devotee becomes metaphorically linked to two other Buddhist sites: the stupa and the scroll. This word "mind" needs some explanation. Buddhism recognizes six "sense organs" (Jp: rokkon, literally the "six roots"): the eye is the sense organ that perceives form; the ear, sound; the nose, scent; the tongue, flavor; the body, the "touch" of external stimulus; and, finally, consciousness perceives dharma, itself a multivalent term that in its more specific instances indicates the Buddhist teachings, and at its most expansive encompasses the true nature of reality, the "laws" of the universe. The salient point here is that "consciousness" is embodied, firmly situated in the flesh. This means that the "mind" (Jp: kokoro) is at once the rational and logical center of thought (what in English we might associate with the "brain"), and it is part and parcel of the fabric of our bodies and sensual perceptions (what in English we might associate with the "heart"). This "mind" has a physical locale: it resides somewhere in the region of the chest, a detail that will prove important in the discussions that follow. Whenever I use the word "mind" in this book, this is the entity to which I refer. Considering the reader-text relation, then, ultimately I argue that what Mahāyāna sutras seek is nothing short of a mutually beneficial symbiotic relationship with the human body. In this relationship, the human body serves as a host organism for the sutra-as-symbiont. It is precisely this symbiotic relationship that establishes the ground rules of Mahāyāna textual culture and that becomes the subject of so much miracle literature in medieval Japan.

The Sutras

In this chapter I will be toggling between several early Mahāyāna sutras with the aim of creating a synchronic sketch of the genre as a whole and paying particular attention to what sutras say about their own textuality. All of the sutras I survey here were known in medieval Japan through their classical Chinese translations. My point in treating them as a group is to suggest the degree to which their themes and strategies overlap. I treat Mahāyāna sutras as a literary genre rather than as the intellectual property of any given spiritual school or group of schools. This approach reflects my grounding in setsuwa, a genre that transgresses sectarian bounds.

My basic criterion for deciding what sutras to include in this chapter is whether or not a given sutra appears in setsuwa. The Lotus Sutra clearly receives the lion's share of attention, with the Pure Land sutras following in a collective second place and the others considerably further behind. Properly speaking, the Flower Ornament Sutra, which I discuss at length below, did not attract much attention in Japanese setsuwa literature, possibly because its length made many forms of popular devotion (memorization, for instance) quite difficult. Nevertheless, when it does appear in the literature, it makes a powerful impact: Myōe (the monk who severed his ear) was a devotee of the Flower Ornament Sutra. Further, the sutra is particularly germane to this discussion because it is largely given over to describing the special characteristics of bodhisattvas, distinctions that helpfully illuminate the function of memory in the preservation of sacred text. The only other anomaly is the Heart Sutra. Because of its extreme brevity, this sutra was widely memorized and chanted even by illiterate believers. Although the Heart Sutra appears frequently in setsuwa literature, its 276 written characters say little about textual culture, so I omit it from my discussion in this chapter.

One distinguishing characteristic of Mahāyāna sutras is that they were composed in the early centuries of the Common Era, nearly half a millennium after the historical Buddha's death. Nevertheless, they all stage themselves as oral discourses, generally with the historical Buddha serving as the main character. The Lotus Sutra is typical in this regard and begins by placing the Buddha at Vulture Peak on the Indian subcontinent addressing a vast audience. The sutra centers on a series of lengthy dialogues between the Buddha and various interlocutors, the course of which produces a number of famous parables such as the parable of the burning house, which I mention in passing below. The Lotus was translated into Chinese a number of times, most famously by a team of translators led by Kumārajīva (344-413), and it is his translation (T 262) that I refer to in this study.

The Sutra of Immeasurable Life and the Sutra of Meditation on the Buddha Amitāyus (Jp: Amida) also situate the Buddha in the center of the Indian subcontinent at Vulture Peak in the realm of King Bimbisāra. In the Sutra of Immeasurable Life, our historical Buddha, speaking to a large audience primarily through his chief interlocutor and disciple Ānanda, describes how the buddha Amida achieved enlightenment, reiterating that buddha's vows and painting a detailed picture of his Pure Land. This sutra has many translations into classical Chinese, and I refer here to the one by Samghavarman (active third century, T 360). The Sutra of Meditation on the Buddha Amitāyus addresses similar subject matter from a different vantage point. While the Buddha has been at Vulture Peak, the son of King Bimbisāra has killed his father and imprisoned his mother, Queen Vaidehī. When Queen Vaidehī calls on the Buddha to aid her, he appears in her prison cell along with Ānanda and he teaches her to visualize the Pure Land of Amida. I reference the translation by Kumārajīva (T 366).

The Diamond Sutra and Nirvana Sutra move the locale northeast, closer to the base of the Himalayas. The Diamond Sutra places the Buddha at the Jetavana Grove in Śrāvastī. The subject consists largely of a dialogue between Buddha and his disciple Subhūti concerning the true nature of reality. Again, I reference the Kumārajīva translation (T 235). The Nirvana Sutra stages itself at the nearby town of Kuśinagar under a pair of teak trees in whose shade the Buddha has laid himself down to die. The sutra consists mostly of conversations the Buddha has with his disciples Kāśyapa and Manjuśri, and a lengthy dialogue with the bodhisattva Lion's Roar, in which he clarifies earlier teachings and explains the nature of nirvana. My references are to the translation by Dharmakśema (385-433, T 374).

The Vimalakīrti Sutra and the Flower Ornament Sutra depart from the previous patterns in that the Buddha is not the undisputed main speaker in either. The Vimalakīrti Sutra opens with the Buddha at the Amra Gardens in Vaishali near the northeast edge of the Indian subcontinent. Most of the narrative, however, concerns the layman Vimalakīrti, who has taken ill and who uses the opportunity of his illness to preach on a number of topics, most prominently the evanescence of the body. I reference Kumārajīva's translation (T 475). Similarly, although the Flower Ornament Sutra begins by placing the historical Buddha on the Indian subcontinent in the eastern kingdom of Magadha, the bulk of the discourse is carried out by "transhistorical, symbolic beings" who often speak through the power of the Buddha. The actual orator at any time may be the Buddha, two buddhas (the speaker and the Buddha who underwrites him), or even a myriad of buddhas all speaking in unison and often all sharing the same name. Early chapters wend from one topic to the next, though most are concerned with the special characteristics of bodhisattvas. The final chapter, which dwarfs most of the others, follows the progress of the devotee Sudhana as he enters into and pursues the bodhisattva path. The scripture was translated, in whole or in part, multiple times into classical Chinese. The one I reference is the lengthier Śiksānanda (652-710) translation (T 279), though I also consulted the earlier Buddhabhadra (359-429) version (T 278).

As several scholars have pointed out, while Mahāyāna sutras typically situate themselves as the direct discourse of the historical Buddha, they are in fact later compositions. This raises important questions about the literary nature of these texts and the particular ways in which they understand their fictions of place, presence, and orality. In the following sections I take up these arguments, examining the narrative strategies sutras employ and characterizing the genre as a metafictional enterprise.

Mahāyāna and Metafiction

Though metafictional elements have occasionally been identified in earlier works, metafiction is most closely associated with postmodernism, and the word itself was not coined until 1960, so it may seem odd that I am characterizing many Mahāyāna sutras, texts that were written in the first centuries of the Common Era, as metafictional. I use the term in the sense articulated by Patricia Waugh, who defines metafiction as "writing which self-consciously and systematically draws attention to itself as artifact in order to pose questions about the relationship between fiction and reality." While she characterizes metafiction as explicitly postmodern, many of the specific techniques she discusses as its core strategies can be found in abundance in Mahāyāna sutras. In addition to its rhetorical style, metafiction also concerns itself with a critique of genre, often calling attention to its own literariness by parodying or departing from oral convention. Thus, metafiction is both a set of narrative strategies and a reaction against a certain temporal frame, a desire to shake things up, to use literature as a way to depart from convention and, by critiquing convention, to create something new. While some may balk at my use of the word "fiction" to describe sacred texts, the point I mean to stress is that Mahāyāna Buddhism was born out of just such a desire to shake things up and that, furthermore, it chose written scripture-literature-as the primary tool through which to orchestrate that upheaval.

Where did Mahāyāna sutras come from? Did they originate in the teachings of the historical Buddha, kept, as some traditional sources maintain, beneath the sea under the watchful eye of the Dragon King until such time as land-loving beings were ready for them? Are they the latter-day revelations of inspired mystics? Or are they the conscious construction of "wily authors"? I think the last two speculations are the more tenable and, in both of these scenarios, Mahāyāna sutras are "fiction" in the sense that they were created and fashioned by human hands. Over the next several pages I will examine the ways in which metafictional strategies obscure the authorial hand while implicitly critiquing the notion that Mahāyāna sutras originate in the oral discourse of the Buddha. By invoking these strategies, sutras invest themselves with authority and begin to address their audiences directly, asking their readers to treat them in certain ways.

Let me be clear. By using the tools of fiction to think about sutras, I do not mean to question either their efficacy or their validity. Rather, I wish to point out two things. First, sutras as we have them today possess a complex textual history that we, as yet, understand imperfectly. Second, sutras employ a rich range of rhetorical and literary techniques. Drawing concerted attention to these may help us to understand better both the medium and the message: how sutras work and what they say. Approaching sutras in this way is a bit like taking apart a pocket watch. The objective is to understand how the watch functions, not to deny the forward march of time whose movement the watch was created to make visible and quantifiable. Assuming that sutras were created in order to make certain things visible, let us then look closely at how they do so.

Mahāyāna sutras make use of a host of metafictional narrative techniques, some of which operate at a surface level while others cut straight to the core of the Mahāyāna as a particular Buddhist path. As one of the more superficial techniques, many sutras make intertextual allusions to other sutras: the Flower Ornament Sutra references a sutra named Voice of the Cycles of Teaching of All Buddhas (T 10.279.381c19); the Nirvana Sutra glosses passages from the Contemplation Sutra (T 12.374.474a27 ff. and T 12.374.565.604 ff.) and the Lotus Sutra (T 12.374.471a28 ff.); while the Lotus Sutra alludes to the Sutra of Immeasurable Meanings (T 9.262.2b8), which appears to have been composed after the Lotus Sutra in the wake of this imaginative citation. These intertextual references establish a given sutra as a member of a recognized genre of texts. Further, these allusions often serve as the stepping-stone for the sutra at hand to claim its own supremacy, defining itself as containing all the previous teachings and going beyond them.

In addition, Mahāyāna sutras often include as part of their narrative passages detailing the reception of that narrative. These descriptions are sprinkled liberally throughout the text, and most sutras conclude with a scene of joyful reception. In fact, these scenes are such a staple of the genre that the Nirvana Sutra defines "sutra" as any text that "begins with 'Thus have I heard,' and ends with 'were overjoyed, did obeisance, and departed'" (T 12.374.451b22-23). In this way sutras establish what reactions they anticipate from their audiences. As readers, we are expected to receive the teachings with great happiness, to worship and make offerings to them, to vow to protect them, and to promise to teach them elsewhere in the world. Lest we forget, our role is almost always scripted for us in the final paragraphs of the sutra, so that as we turn from its closing lines to face the world outside, we remember the charge we have been given.

The Trouble with Entrustment

In a somewhat more complex gesture, Mahāyāna sutras typically include scenes in which the narrative is named and entrusted. In these interludes, the main character (the historical Buddha) charges a secondary character (usually Ānanda, but sometimes Kāśyapa or another disciple) with the duty of preserving and propagating the story in which they both appear. Imagine, for a moment, reading the final chapter of a novel in which a secondary character turns to the protagonist and asks him, "What's the name of this book we're in, and how would you like me to circulate it after you die?" When the naming and entrustment scene is successful, and when it occurs near the end of the text, it activates the traditional opening of sutras ("Thus have I heard") and reminds readers that they have been consuming a frame narrative. Readers can now identify that "I," who disappeared after line one, as a specific character in the story.

This frame narrative also establishes something of a phantom link between the sutra's authority and its purported origin in the word of the Buddha. The other metafictional strategies, however, largely work to undermine this tentative link and, in fact, naming and entrustment scenes are at times quite problematic. The Nirvana Sutra, for instance, is virtually a case study in everything that can go wrong with entrustment. Put another way, the sutra showcases a variety of ways in which a written sutra, by disrupting the smooth functioning of entrustment, can stop being merely the object that is passed around (orally entrusted by one person to another) and can instead begin to seize authority for itself.

When the Nirvana Sutra opens, the Buddha is on his deathbed. Myriad beings appear, all grieving and wailing that once the Buddha has passed into extinction, the teachings, too, will begin to die. Responding to these worries, Buddha encourages the monks in the assembly not to despair, because "I now entrust [Jp: fushoku] all the unsurpassed true dharma to Mahākāśyapa. This Kāśyapa will henceforth be the one on whom you may rely. Just as the Thus Come One is the refuge of all beings, so it is with Mahākāśyapa. He is now your refuge." The monks, however, have something else in mind and they make a counterproposal, arguing, "If this dharma treasure is entrusted to Ānanda or any other monk, it will not abide for long. Why is this? Because all of the voice-hearers including Mahākāśyapa will die.... Thus, you should entrust this unsurpassed buddha dharma to all the bodhisattvas."

At this point in the Nirvana Sutra narrative, the Buddha has not even suggested Ānanda as an option, though he was widely regarded in pre-Mahāyānic tradition as the monk who had heard and remembered more of Buddha's teachings than any other. The monks' point, though, is that any one of them may die at any moment, so it would be far safer to entrust the teachings to all of the advanced practitioners rather than any single person. The Buddha recovers quickly from this criticism and praises the monks' forethought, noting that he had already considered this contingency. He accepts their proposal and formally announces that he hereby entrusts the dharma to all bodhisattvas. The narrative continues, however, as if this protest had never been lodged, and a few pages later Kāśyapa formally vows to "protect and hold the true dharma," expounding it widely to all other beings and beating down those who do not have faith in it in the way that hail or frost beats down the tender grasses (T 12.374.382c20-23).

What we have here, in somewhat muted form, is something like a deathbed squabble over the inheritance. To whom will the treasure of the dharma be handed down? At this point, the treasure under discussion is the entire dharma, and not just any single named sutra, so whoever receives the formal assignment stands to become the next leader of the assembly. One faction stands behind Ānanda, another behind Kāśyapa, and yet another espouses pluralism. Though the Buddha, as developed in this sutra, agrees to pluralism in principle, in practice he anoints Kāśyapa. At least for now.

The troubles are not over. Though Kāśyapa may have been chosen as the receptacle for all previous teachings, the Buddha is clearly beginning another sermon. Although naming and entrustment scenes usually happen near the end of sutras, Kāśyapa moves preemptively in this case, asking the Buddha, "O World Honored One, what is this sutra to be called? How should bodhisattva-mahāsattvas make offerings to it and hold it?" (T 12.374.385a2-3). Kāśyapa here positions himself as the keeper of all the Buddha's teachings, up to and including this final dispensation. The Buddha, appearing to accept Kāśyapa's bid, names the sutra that he is in the process of delivering and then returns to his sermon, making a number of provocative points that seem to contradict some of his earlier teachings. Kāśyapa cannot help but point out several of these glitches, and eventually his persistence seems to annoy the Buddha, who replies somewhat testily, "If you have doubts in what I say, you should not be the one to accept [the teaching]!" (T 12.374.397b11), and he counsels Kāśyapa to throw off the deceptions of Māra, the Evil One. Kāśyapa does so and finally agrees to accept the teachings as given. The issue of entrustment appears to be resolved. The sermon draws to a close, and the layman Cunda is chosen to present the last offerings.

At this point, however, Manjuśri stands up, says that he has some questions, and engages the Buddha in a lengthy conversation, an addendum to the sermon, which Kāśyapa occasionally interrupts. The Buddha concludes the sermon a second time, saying, "Manjuśri and the rest of you, you should expound the great dharma widely, to all people. Now, I entrust this dharma to you [Manjuśri] and, when Kāśyapa and Ānanda arrive, you should entrust this true dharma to them, too." Haven so spoken, he lies down on his right side, succumbing again to the pain of his bodily illness.

This second entrustment scene raises a rather thorny question, under which is hidden an equally thorny textual problem. Where has Kāśyapa gone? At several points earlier in the sutra there have been references to the notion that, although virtually all beings in the cosmos appear to be gathered around the Buddha's deathbed, Kāśyapa and Ānanda are missing. In truth, Ānanda has not yet shown up as an active character in the narrative, but, as we have already seen, Kāśyapa has been serving for dozens of pages as Buddha's main interlocutor. He has already been entrusted with (a truncated version of) the Nirvana Sutra, he has already clashed with Manjuśri over who will receive the final teaching, and, to make matters particularly perplexing, the Buddha was just talking to him. In fact, after a few more chapters (dominated by a dialogue between Buddha and Manjuśri), Kāśyapa will interject himself into the conversation once again, seeming to have been present and listening all the while.

What this snag in the narrative tells attentive readers is that the text of the Nirvana Sutra that we now have is clearly a composite document, one that has been stitched together from at least two separate sources. It is ironic that the entrustment scene is the segment of the narrative that most obviously exposes the seams in that narrative's textual history. If the conventional work of the entrustment scene is to provide a clear point of oral origin for the written text of the sutra, that is manifestly not how this particular entrustment scene works. I would not go so far as to call it a parody of convention, but it is definitely a departure, and a critical one at that.

As if this were not enough, there is a third entrustment scene, this one properly situated very near the end of the narrative, but no less troubling for its place of prominence. Following the second entrustment, the Buddha engages in a series of intensive dialogues, first with Manjuśri and Kāśyapa, and then with the bodhisattva Lion's Roar. Buddha and Lion's Roar agree that Ānanda is the best candidate to receive the final teachings and make the final offering, and they discuss his memory skills at length. Buddha entertains several more questions from the assembly and then asks for Ānanda, presumably preparing to entrust him with the entirety of the Nirvana Sutra. Ānanda, however, has never arrived. One of the assembly informs Buddha that Ānanda has been waylaid by Māra, who has turned himself into sixty-four thousand billion likenesses of Buddha, each of which is preaching a different doctrine. Ānanda is "greatly pained" by this and by his inability to break free, and, though he stays his mind on the Thus Come One, strangely no one has come to his aid (T 12.374.600c25-26).

Manjuśri then makes what could be interpreted as a power play, assuring the Buddha that certainly, among all the beings congregated, there is at least one capable of receiving and retaining the teachings. He persists, "Why ask where Ānanda is?" (T 12.374.601a10). Buddha lectures Manjuśri at length about Ānanda's special qualities, and particularly his power of memory, before concluding, "That is why I ask where Ānanda is, [because] I desire for him to accept and hold this Nirvana Sutra." Buddha sends Manjuśri to rescue Ānanda and Ānanda returns, but he is never formally entrusted with the teachings and never takes a vow to preserve or propagate them. In narrative terms, all his appearance has done is to cast doubt on Manjuśri and Kāśyapa's powers of memory (which are inferior to his own) and draw attention to the fact that, at the sutra's close, Buddha still has not settled the question of who should be entrusted with preserving the teaching or if, in fact, different disciples hold responsibility for different portions of it. If Ānanda is, indeed, the Buddha's choice (or one of the Buddha's choices), then Ānanda will have to depend on the cooperation and memory skills of his fellow disciples since he, in fact, never heard any of the sutra with whose preservation he may have been charged.

Scenes like these reveal that Mahāyāna sutras draw on the narrative conventions of oral delivery and transmission while undermining some aspects of that structural logic. The point is that the way Mahāyāna sutras were created (through a writing process that at times resulted in a composite text) and the way they suggest they were created (through the transcription of a perfect memory of an entrusted verbal teaching) do not match up. It seems certain that at least some Mahāyāna sutras were stitched together from a number of disparate texts into a single literary document that comes apart at the seams when read attentively from beginning to end. Entrustment scenes can reveal these seams and some of the stitchwork done to conceal them. What they do not reveal, however, are the identities of the tailors: early Mahāyāna sutras have no (named) authors. Thus, they are severed both from a verifiable history of oral discourse and from an identifiable act of literary fabrication. Instead, they float free from both these traditional locales of textual authority.

The Anxiety of Text

If Mahāyāna sutras occasionally advert to unsettling questions about their textual pasts, in another metafictional strategy they are much more overtly anxious about their ability to survive into the future. A substantial portion of many Mahāyāna sutras consists of rhetoric built to engender fear that the narrative will be corrupted, destroyed, or otherwise lost to circulation. In the Sutra of Immeasurable Life, for instance, the Buddha maintains matter-of-factly that, in the ages to come, "all scriptures and paths will perish, but out of compassion and pity, I will especially preserve this sutra and maintain it in the world for a hundred years more." Similarly, in the Lotus Sutra the Buddha enunciates a lengthy verse concerned with the fragility of the teachings, studded throughout with lines like, "If after the Buddha's extinction / in the midst of an evil age / one can preach this scripture / that is difficult.... / If after the Buddha's passage into extinction / in the midst of an evil age / to read this scripture for but a moment / that is difficult.... / If after my extinction / one can hold this scripture / and preach it to even one person / that is difficult." Part of what makes these actions so difficult is the simple passage of time, during which memories fade, pages rot, and lineages of oral transmission tangle and break.

Mahāyāna sutras also worry about enemies other than time. The Lotus Sutra speaks of human adversaries who "malign the scriptures," acting on "hatred and envy" (T 9.262.15b26-27). The Vimalakīrti Sutra admits to similar anxieties, as when the Buddha instructs Maitreya to "employ supernatural powers to propagate sutras such as this, spreading them" throughout the world "and never allowing them to be wiped out [Jp: danzen]." The Nirvana Sutra provides the most explicit and extended nightmare of textual desecration. The Buddha suggests to Kāśyapa that a mere eighty years after his passage into Nirvana,

there will be many evil monks who will abridge this sutra and cut it into many pieces, so that the color, fragrance, beauty, and flavor of the true dharma are lost. All these evil monks ... will insert worldly phrases, grand and decorative, but devoid of the essential. Or they will chop off the beginning and add it to the end, chop off the end and add it to the beginning, or put the beginning and end in the middle and the middle at the beginning and end. You should know that evil monks such as these are friends of Māra.

This violent editorial cutting and pasting will obstruct beings' access to the scripture, Buddha argues, keeping them from getting at its true meaning. Like Roland trying to control the arrangement of his corpse on the battlefield, the Buddha wants his textual corpus to maintain a certain shape and order, lest the evidence be misread and his deeds misunderstood. The imagery supports the notion of a great physical battle in which sutras are unstrung, sliced and chopped, then sewn back together into unsatisfying semblances of their former selves.

Passages such as these work to establish an overarching atmosphere of desperation, and they stress the importance of the recipient in transmitting and preserving the text in these violent times. In addition to elaborating an anxious tone, these passages also suggest an important characteristic of sutras as material objects, namely, that they are not capable of sustaining themselves. With the Buddha having died, the sutras are in danger of dying, too, unless someone else makes special, even supernatural, efforts to preserve them. Even when these efforts are made, however, beings find them "difficult," and inevitably the sutras will all eventually perish. The world will pass through a spiritual dark age, which will end only when another buddha appears to start the cycle turning again.

The Sutra Library

Happily, there is an inverse to this nightmare. If the Nirvana Sutra imagines the tatters in which a material copy of a sutra may end up, the Vimalakīrti Sutra provides an equally extreme account of the opposite, a fantasy of completely perfect preservation. In the thirteenth chapter of the Vimalakīrti Sutra, the Buddha offers the following story. Long, long ago there was a buddha named Medicine King who lived and taught in a world called Great Adornment. His chief patron was a king who had one thousand sons, one of whom was named Moon Parasol. One day Moon Parasol, observing all the various offerings of his father and brothers (flowers, incense, banners, music, etc.), wonders to himself what the best of all offerings is. Suddenly, a heavenly being appears in the sky and tells him the most superior is the offering of the dharma. Neither Moon Parasol nor his father knows how to make this offering, so they ask Medicine King to explain. He says, "Good man, the offering of the dharma means the profound sutras preached by all the buddhas. The people of the world all find them difficult to believe and difficult to accept, for they are wonderfully subtle and difficult to see, clean and pure and without stain.... They are contained in the storehouse of the bodhisattva and are sealed with the dhāranī seal." Moon Parasol, the Buddha tells us, was one of his former incarnations and thus, in delivering the teachings of the Vimalakīrti Sutra, the Buddha completes the offering of the dharma that he, as Moon Parasol, desired to make. I will deal more closely with this sort of pretextual history soon, but for the moment let us concentrate on the issue of textual perfection.

In this story, the buddha Medicine King speaks of a vast storehouse filled with text, in other words, a library. The holdings of this library consist of every sutra ever given, each of which is preserved whole, with no smudges or stains, and each of which is stamped with a seal that marks it as genuine. This library is not exactly open to the public, however, because it exists inside the bodhisattva who has achieved such an advanced spiritual stage that he has the power of perfect recall (Sk: dhāranī). The bodhisattva's memory is thus a library that only he can read. This is just as well, because normal, worldly people have trouble seeing such a bright, clear text (a frustrating experience perhaps remotely akin to trying to read a glossy page under a high-powered fluorescent light). The offering of the dharma, therefore, consists of making the perfectly preserved sutras visible to everyday beings by expounding upon them in sermons. These sermons form the basis of various individually named sutras.

The Flower Ornament Sutra contains a similar passage that helps delineate between the perfect, and perfectly sealed, sutras of the bodhisattva and the sutras that we, as human beings, might hope to encounter in this world. In a section explaining the tenth and final intuition of a bodhisattva on the path to enlightenment, the sutra states that the Buddha's wisdom "exists within the body of every living being," a fact that often goes unrecognized, literally "unseen." The sutra glosses this situation as follows: "It is as if there were a single sutra scroll ... and on this scroll were recorded all things in the multi-thousand-fold world, [leaving] nothing not [recorded]." This sutra scroll is then envisioned as existing within a mote of dust, and in fact every mote of dust is similarly possessed of a sutra scroll. Sutras, therefore, are everywhere, millions upon millions of them within our very bodies, but they are obscured and hidden, illegible to the vast majority of living beings. We have a sutra library inside of us, but we do not have access to its holdings. Rendering the sutra scrolls legible requires nothing short of an act of readerly fission in which a person possessed of "penetrating wisdom" and a "pristine pure divine eye" (T.9.278.624a8)-in other words, a buddha or very advanced bodhisattva-trains the sacred optic on the mote of dust and splits it asunder, releasing the sutras into the world, where they can now be of benefit to living beings.

There is a distinction to be drawn, then, between the sutra as "text" and the sutra as "work." Editors of critical editions have long noted the various discrepancies that occur between different copies of a single work of literature. A copy may be illustrated or not, include marginalia or not; it can be produced as manuscript or as type, in different fonts, on different substances, with various page and line breaks; some authors, like Walt Whitman, published numerous versions of a single poem; and so forth. Proponents of textual studies assert that each of these differences, though they may at times seem to be superficial, exerts a subtle pull on the work, creating a greater or lesser sense of variability and pointing to the instability of literature as an art form. Thus the famous question comparing the singularity of Mona Lisa with the bewildering multiplicity of Hamlet.

Peter Shillingsburg provides a useful set of terms for grappling with the realities of textual dispersion. According to his terminology, a piece of literature that might be expressed in any number of forms (in folio or quarto, on vellum or paper, in various editions, etc.) but that is not reducible to any single one of those expressions should be termed the "work." Distinct apprehensions of the work that are expressed in words and punctuation form the "linguistic text." This linguistic text, in turn, may be housed in any number of material forms or physical "containers." The combination of linguistic text and physical container forms the "material text." While the work and the linguistic text exist in the abstract as inaccessible and intangible, the material text (any single iteration of the work expressed in language and contained in some physical form) is available to human senses. We only ever read material texts.

To gloss this terminology with respect to the literature at hand, the sutra as a linguistic text (a string of words and line breaks) exists abstractly in this world and may survive by lodging itself in a container, the most popular being memory or some sort of external surface (leaves or paper). This physically lodged material text is subject to the ravages of time and the violent attacks of editors or rivals and, even given the best efforts of its devotees, will eventually fade and die. The sutra as work, however, continues to exist, perfect and undamaged, and this abiding teaching is at least one of the concepts encompassed in the Sanskrit word dharma. In later chapters of this book I will tend to bracket the idea of the "dharma," the perfect "work," since it is inaccessible to the average believer. Instead, I will be focusing on the "material text," which is complicated enough. After all, this text can comprise any of the physical, material copies of that work that may be written on leaves, carved into bark or stone, brushed onto silk or paper, or inscribed on the surface of the mind. It is these interconnected surfaces that so intrigued the authors of medieval miracle literature and that found their ways into the sermons of medieval priests. Unless otherwise specified, then, I intend the word "text" to refer to the material text as housed in a specific container.

Remembering Prehistory

A number of Mahāyāna sutras contain within them stories of the Buddha's past lives. Occasionally, in yet another metafictional narrative move, these sutras tell us that the Buddha, over the course of one of these previous incarnations, has encountered the very sutra that he is currently preaching. In the Lotus Sutra, for instance, the Buddha explains to the assembly of monks that, in a certain kingdom long ago, there was a buddha named Victorious Through Great Penetrating Knowledge who often gave sermons to the king and his sixteen sons. Hearing the dharma expounded for the fourth time, all of the sixteen princes, along with myriad other beings in the kingdom, renounce their worldly riches and take up vows, embarking on a path of formal practice. Following a hiatus of twenty thousand kalpas, during which the princes have been perfecting their training, the buddha preaches again, this time finally delivering the long-awaited sutra entitled, of course, the Lotus Sutra. After expounding the Lotus for eight thousand kalpas, Victorious Through Great Penetrating Knowledge retires to a quiet chamber where he spends eighty-four thousand kalpas in meditation. In the meantime, those assembled have one of three responses: joyful acceptance (the sixteen princes), a wait-and-see attitude (some of the other voice hearers), and doubt (everyone else). The sixteen princes commit the sutra to memory, then preach it widely. Eventually they all become buddhas, the sixteenth of whom is none other than the historical Buddha Śākyamuni, who is telling his audience this story about his past. In other words, readers of the Lotus Sutra encounter a narrative in which the orator of the narrative tells a story about how he heard the story from someone else. The story that the "someone else" told, the story that the Buddha is now telling, and the narrative that we are now reading all have the same title.

One point these narrative backstories make is that the Buddha is not the author of any particular sutra, nor is he the wellspring from which the work originates. He is, rather, the latest in a long line of enlightened beings who have committed themselves to memorizing and spreading copies of the text. Another implication of these stories has to do with narrative drift. When the Buddha features as a character in his own narrative, the line between story and storyteller blurs. He claims to have memorized the sutra, the same sutra with the same title that he is now preaching, when he was a young ex-prince, but with two major (and unacknowledged) alterations. First, material has been added: the story of the sixteenth prince could not have been present in the sutra that the sixteenth prince originally heard, and therefore our text includes a later addition. Second, material has clearly been deleted. English translations run roughly four hundred pages, complete with notes. One would have to preach quite slowly indeed for this text to require eight thousand kalpas to transmit. Thus, the particular relation between the text of the Lotus Sutra and the abiding work remains ever in question, but in a promising and alluring way: the story can expand to include audience members as characters.

Three variations on the motif suggest how this metafictional strategy can encompass figures other than the Buddha, drawing an increasing range of beings into the textual organism. The first variation involves a question. The Nirvana Sutra includes a lengthy dialogue between the Buddha and the bodhisattva Lion's Roar, during the course of which Lion's Roar asks the Buddha about the accuracy of a typical metaphor that describes nirvana as a flame blown out. Rather than answer the question, the Buddha responds that ages ago a buddha named Well Gained delivered a teaching of the Nirvana Sutra that took three billion years. At that time, "I, along with you, was among those congregated. I asked that buddha this same question. At the time, that Thus Come One was in right samādhi for the sake of all beings and did not reply. Excellent, great one! You do well to remember this incident." Lion's Roar plays, in this textual generation, the same role that the Buddha played in the last generation with the unspoken promise being that, in a future iteration, Lion's Roar will be the buddha to whom the question is posed. While still operating within a very elite circle-a bodhisattva who is the chief interlocutor of a living buddha-this narrative move creates more room for play in the borders of the text.

The second variation works with the issue of expanding text in a different way, by using nominal sameness to refract time and space into an endless hall of mirrors. The Buddha, still speaking to Lion's Roar, recalls hearing a sutra called the Nirvana Sutra delivered by a buddha named Śākyamuni, who was born to a king and queen with the same names as the historical Buddha's mother and father. This doppelgänger buddha has disciples who also share names with the Buddha's disciples Śāriputra and Ānanda. After listening to that buddha preach the Nirvana Sutra, the Buddha