(Continued from front flap)
Mikael Adolphson is associate professor of history at Harvard University. Edward Kamens is Sumitomo Professor of Japanese studies and professor of East Asian languages and literatures at Yale University. Stacie Matsumoto is a doctoral candidate in history at Harvard University.
T
kamens matsumoto
“This exceptionally rich set of essays substantially advances our understanding of the Heian era, presenting the period as more fascinating, multi-faceted, and integrated than it has ever been before. This volume marks a turning point in the study of early Japanese culture and will be indispensable for future explorations of the era”
Andrew Edmund Goble, University of Oregon
• “As a Japanese historian, I enthusiastically recommend Heian Japan, Centers and Peripheries, the first multi-author English-language academic work to offer a synthetic treatment of the Heian period. Japan’s emperor system is the last remaining sovereignty of its kind in human history, and this volume is indispensable when considering what sovereignty itself means in the present. To that end, the classical patterns established in the Heian period are superbly analyzed in this volume through the dual approach of ‘centers and peripheries.’ ”
Hotate Michihisa,
Historiographical Institute, University of Tokyo
Uni v er sit y of H awa i‘i Pr ess Honolulu, Hawai‘i 96822-1888
•
www.uhpress.hawaii.edu Jacket illustration: In this scene from the Gaki zōshi (Hungry Ghosts scroll, twelfth century), Heian courtiers enjoy a musical entertainment while hungry ghosts search for food from their human hosts. Condemned to this sorry state of invisibility, perpetual hunger, and eating human waste for their sins in a previous life, the ghosts remind the viewer of the dangers of human pleasures. Courtesy of Tokyo National Museum. Image by TNM Image Archives (source: http://tnmarchives.jp) Cover design: April Leidig-Higgins
Heian Japan, Centers and Peripheries
the importance of the early tenth century as a watershed that highlights the institutional and political transformations at court whereby provincial governors were allowed more freedom and, by extension, greater financial benefits. The second point problematizes the notion of a singular dichotomy between center and periphery in Heian Japan. The various essays suggest instead that the nexuses of power were in fact plural, and the periphery was not as peripheral as had been imagined. Thus, rather than conceiving Heian society as a static and one-dimensional formation centering on Kyoto alone, it might better be understood as a society of multiple centers and peripheries. The third point challenges the long-held view that the central government’s lessening of administrative control of the provinces meant an increasing loss of power. Rather, the abandonment of a strict administrative approach in favor of a more effective one allowed elites in the capital to strengthen their hold on the provinces, reflecting an improved integration of centers and peripheries. Fourth, the methods and means of exercising power shifted from one relying solely on official titles and procedures to one that was increasingly based on extra-governmental means, a process of “privatization” that reflected the development of multiple centers of social, political, and economic practice outside the official structures of the state. Heian Japan, Centers and Peripheries presents not only a set of new interpretations of this epochal moment in the Japanese past, but also offers a host of new questions to be addressed in future international and interdisciplinary research modeled on this exemplary volume.
Adolphson
Ja pa n e se history
heia n japa n
center s a nd Per ipher ies • Edited by Mikael Adolphson, Edward K amens, and Stacie Matsumoto
he first three centuries of the Heian period (794 – 1086) saw some of its most fertile innovations and epochal achievements in Japanese literature and the arts. It was also a time of important transitions in the spheres of religion and politics, as aristocratic authority was consolidated in Kyoto, powerful court factions and religious institutions emerged, and adjustments were made in the Chinese-style system of rulership. At the same time, the era’s leaders faced serious challenges from the provinces that called into question the primacy and efficiency of the governmental system and tested the social/cultural status quo. Heian Japan, Centers and Peripheries, the first book of its kind to examine the early Heian from a wide variety of multidisciplinary perspectives, offers a fresh look at these seemingly contradictory trends. Essays by fourteen leading American, European, and Japanese scholars of art history, history, literature, and religions take up core texts and iconic images, cultural achievements and social crises, and the ever-fascinating patterns and puzzles of the time. The authors tackle some of Heian Japan’s most enduring paradigms as well as hitherto unexplored problems in search of new ways of understanding the currents of change as well as the processes of institutionalization that shaped the Heian scene, defined the contours of its legacies, and make it one of the most intensely studied periods of the Japanese past. Throughout, the widely deployed model of “centers and peripheries” is tested as a guiding concept: It serves here as a point of departure for a reexamination of the dynamic tensions among and between literary languages, administrative structures, urban centers and rural regions, orthodoxies and heterodoxies, the status quo and the pressures for adaptation and change, and many other powerful entities and sociocultural forces. An introductory chapter lays out the volume’s four main points. The first emphasizes (Continued on back flap)
Heian Japan, Centers and Peripheries
Heian Japan,
Centers and Peripheries Mikael Adolphson, Edward Kamens, and Stacie Matsumoto, Editors
University of Hawai‘i Press d Honolulu
© 2007 University of Hawai‘i Press All rights reserved Printed in the United States of America 12 11 10 09 08 07 6 5 4 3 2 1 Library of Congress Cataloging-in-Publication Data Heian Japan, centers and peripheries / Mikael Adolphson, Edward Kamens, and Stacie Matsumoto, editors. p. cm. Includes bibliographical references and index. ISBN-13: 978-0-8248-3013-7 (hardcover : alk. paper) ISBN-10: 0-8248-3013-X (hardcover : alk. paper) 1. Japan — History — Heian period, 794–1185. I. Adolphson, Mikael S., 1961– II. Kamens, Edward, 1952– III. Matsumoto, Stacie, 1969– DS856. H424 2007 952'.01 — dc22 2006024714 University of Hawai‘i Press books are printed on acidfree paper and meet the guidelines for permanence and durability of the Council on Library Resources. Designed by April Leidig-Higgins Printed by Edwards Brothers Inc.
contents
List of Maps, Figures, and Tables vii Acknowledgments ix Terminology and Translations xi 1 Between and Beyond Centers and Peripheries 1
Mikael Adolphson and Edward Kamens
Part I. Locating Political Centers and Peripheries 2 From Female Sovereign to Mother of the Nation: Women and Government in the Heian Period 15
Fukutō Sanae with Takeshi Watanabe
Joan R. Piggott
3 Court and Provinces under Regent Fujiwara no Tadahira 35 4 Kugyō and Zuryō: Center and Periphery in the Era of Fujiwara no Michinaga 66 G. Cameron Hurst III
Part II. Shifting Categories in Literature and the Arts 5 The Way of the Literati: Chinese Learning and Literary Practice in Mid-Heian Japan 105 Ivo Smits
6 Terrains of Text in Mid-Heian Court Culture 129
Edward Kamens
7 The Buddhist Transformation of Japan in the Ninth Century: The Case of Eleven-Headed Kannon 153
Samuel C. Morse
Part III. Establishing New Religious Spheres 8 Scholasticism, Exegesis, and Ritual Practice: On Renovation in the History of Buddhist Writing in the Early Heian Period 179
Ryūichi Abé
9 Institutional Diversity and Religious Integration: The Establishment of Temple Networks in the Heian Age 212
Mikael Adolphson
vi | Contents
10 The Archeology of Anxiety: An Underground History of Heian Religion 245 D. Max Moerman
Part IV. Negotiating Domestic Peripheries 11 Famine, Climate, and Farming in Japan, 670 – 1100 275
William Wayne Farris
12 Life of Commoners in the Provinces: The Owari no gebumi of 988 305
Charlotte von Verschuer
13 Lordship Interdicted: Taira no Tadatsune and the Limited Horizons of Warrior Ambition 329 Karl Friday
Part V. Placing Heian Japan in the Asian World 14 Cross-border Traffic on the Kyushu Coast, 794 – 1086 357
Bruce L. Batten
15 Jōjin’s Travels from Center to Center (with Some Periphery in between) 384 Robert Borgen
References 415 Contributors 439 Glossary-Index 441
maps, figures, and tables
Map 1.1. Provinces and highways of Heian Japan xii Map 1.2. Central Japan xiii Map 1.3. Heian-kyō in the mid-Heian age xiv Map 2.1. The Imperial Palace compound (dairi) 20 Map 2.2. The Greater Imperial Palace precincts (daidairi) 24 Map 10.1. Late twelfth-century sūtra burial sites 254 Map 10.2. Sūtra burial sites in Kyushu 255 Map 12.1. Owari Province 306 Map 13.1. Bōsō peninsula 330 Map 14.1. Hakata and vicinity 359 Map 15.1. Jōjin’s travels 386 Figure 2.1. Imperial genealogy of the seventh and eighth centuries 17 Figure 2.2. Imperial genealogy of the early and mid-Heian period 21 Figure 4.1. Sekkanke genealogy during Michinaga’s times 71 Figure 6.1. Detail from Murasaki Shikibu nikki ekotoba 149 Figure 7.1. Eleven-Headed Kannon, Hokkeji, Nara 154 Figure 7.2. Eleven-Headed Kannon, Ryōsenji, Nara 154 Figure 7.3. Eleven-Headed Kannon, Chōenji, Osaka Prefecture 155 Figure 7.4. Eleven-Headed Kannon, Jikō Enpukuji, Wakayama Prefecture 155 Figure 7.5. Eleven-Headed Kannon, Tadadera, Fukui Prefecture 156 Figure 7.6. Eleven-Headed Kannon, Futagami Kannondō, Fukui City 156 Figure 7.7. Eleven-Headed Kannon, Kōgenji, Shiga Prefecture 172 Figure 9.1. Imperial genealogy of the early Heian age 216 Figure 10.1. Diagram of sūtra burial 249 Figure 10.2. Sūtra container 249 Figure 10.3. Chinese porcelain sūtra container 257 Figure 10.4. Bronze plate incised with text of the Lotus Sūtra 263 Figure 11.1. Drought and government orders to plant dry fields 290
viii | List of Maps, Figures, and Tables
Figure 11.2. Fluctuations in the price of brown rice in the famine year of 762 290 Figure 13.1. Heike genealogy 336 Figure 15.1. Portrait of Jōjin 385 Figure 15.2. Genealogy of Jōjin’s family 387 Figure 15.3. Genealogy of key individuals associated with Jōjin 399 Table 3.1. Fujiwara no Tadahira’s career 37 Table 8.1. Kūkai’s interpretive strategy 199 Table 8.2. Correspondence between the five Heart Sūtra sections 203 Table 9.1. Jōgakuji 223 Table 9.2. Betsuin 229 Table 9.3. Matsuji 234 Table 11.1. The incidence of famine in Japan by century, 600–1900 276 Table 11.2. Climatological trends for selected centuries 280 Table 11.3. Famine years, extent and causes, 670–1100 281 Table 14.1. Changes in military administration in Kyushu 362 Table 14.2. The raid of 869 and its aftermath 367 Table 14.3. The events of the 890s 369 Table 14.4. The Toi Invasion and its aftermath 375 Table 14.5. Foreign contacts during the tenure of Fujiwara no Korenori 379
Acknowledgments
Born as an afterthought following the conclusion of the Association for Asian Studies conference in San Diego in March of 2000, this project began in earnest in September of that year, when a planning group, consisting of Mikael Adolphson, G.Cameron Hurst III, Edward Kamens, Stacie Matsumoto, Joan Piggott, and Mimi Yiengpruksawan, convened to propose a conference on Heian Japan. Our goal was above all to stimulate interdisciplinary exchange and to further promote discussions across national, continental, and generational boundaries by addressing a common theme. To test the feasibility of such a project, a twoday workshop, which served as a forum for scholars to discuss their proposed contributions to the conference, was held in September of 2001. An international conference was subsequently held at Harvard University in June of 2002, when scholars from three continents gathered for three days to focus exclusively on the Heian age, a first in the United States. The success of the conference made a strong statement for the vitality of the field, and the discussions that followed bore witness to the many new issues that are still in need of further scrutiny. During the course of this project, we have relied on the good spirit, support, and cooperation of a large number of people, whom we would like to acknowledge. We would like to express our profound gratitude to the directors and staff of the Edwin O. Reischauer Institute of Japanese Studies at Harvard University for their financial and organizational support. The advice and encouragement of Andrew Gordon, director of the institute, in the early stages and during the conference were invaluable, and we are equally grateful for the continued support by his successor, Susan Pharr. We should also like to acknowledge the crucial financial support we received from the Council on East Asia Studies at Yale University (Mimi Yiengpruksawan, chair) as cosponsor. The conference kept the Reischauer Institute staff busy beyond reason in May and June of 2002, and we therefore extend our special thanks to M.J.Scott, Galen Amstutz, Ruiko Connor, Mary Amstutz, and Margot Chamberlain. Yōichi Nakano and Emi Shimokawa provided additional assistance during the busy days of the conference. The formal and informal discussions during the course of the conference provided direction and motivation for the further development of our essays and arguments. We are grateful to our discussants — Martin Collcutt, Janet Goodwin, Esperanza Ramirez-Christensen, Elizabeth ten Grotenhuis, Jacqueline Stone, Michael Puett, Detlev Taranczewski, and Hotate Michihisa — for many insightful comments and suggestions. We would also like to express our special thanks to Mimi Yiengpruksawan and Yoshimura Toshiko for their formal presentations during the conference, as well as to Ethan Segal, Akiko Takata Walley, and Takeshi Watanabe for their assistance at those surprisingly rare moments when translation and interpretation were needed. Anne Rose Kitagawa,
| Acknowledgments
assistant curator of Japanese art at the Harvard University Art Museums; Anne Nishimura Morse, curator of Japanese art at the Museum of Fine Arts in Boston; and David Ferris, curator of rare books and manuscripts at the Harvard Law School Library facilitated visits to these collections and arranged viewings of rare items that greatly enriched the experience of the conference participants. Our thanks also to W.David Garrahan, Jr., and Bruce Batten for their help with several of the maps and charts included in this volume. The conference organizers were particularly honored by the attendance and participation of several distinguished scholars from Japan. In his remarks at the conclusion of the 2002 conference, Professor Hotate of the Historiographical Institute at the University of Tokyo noted with amazement that so many international scholars from disparate disciplines had come together around one theme to discuss the Heian age. The norms in Japan still call for separate meetings of historians, art historians, religionists, literary scholars, and so forth, but we note that this, too, is changing. But one absence from the conference was most painfully felt. The late Professor Chino Kaori of Gakushūin University, whose works on Heian and Kamakura art and gender have inspired many scholars in both Japan and the West, was part of the small group of scholars who participated in the workshop at Harvard in September of 2001, but she passed away suddenly in January of 2002. It is our hope that this volume may in some measure pay homage to scholars such as Chino, whose insights into and approaches to Heian Japan have encouraged us to look beyond traditional scholarly paradigms. We dedicate this volume to her and to many others who have sought to shed new light on this rich and complex epoch of Japan’s past.
Terminology and Translations
In Japanese studies, scholars within the same field frequently disagree on appropriate usages and translations, perhaps much more so if they belong to different disciplines. In our view, this is not necessarily a weakness since variations may in fact indicate the actual historical usage of a term and point out the differences more clearly between modern and premodern linguistic usages. Thus, while we recognize the importance of consistency, and have indeed encouraged it, the observant reader will also find variations within this volume, but they are relatively few and should not impede interested readers from engaging with the general arguments. Although Japanese terms have unavoidably been used throughout this volume, translations have been provided both in the individual chapters and in the glossary-index. We have followed common academic practices in citing Japanese names with the surname first, followed by the first name. When appropriate, we have retained the genitive no in names of large and highranking families (e.g., Fujiwara no Michinaga) since that was the practice during the Heian era. For years and dates, we use the generally accepted hybrid form — that is, giving the Western year followed by the month and day of the lunar calendar. To help the reader locate entries in diaries that are cited in this study, Japanese era names are consistently listed in the notes.
Map 1.1. Provinces and highways of Heian Japan. Source: Reprinted from Bruce Batten, Gateway to Japan: Hakata in War and Peace, 500 – 1300 (Honolulu: University of Hawai‘i Press, 2006) with permission of the author.
Map 1.2. Central Japan. Adapted from volume 4 of Nihon no rekishi: Ritsuryō kokka (Tokyo: Shōgakkan, 1974).
Map 1.3. Heian-kyō in the mid-Heian age. Adapted from volume 8 of Nihon no rekishi: Ōchō kizoku (Tokyo: Shōgakkan, 1974).
1
d Mikael Adolphson and Edward K amens
Between and Beyond Centers and Peripheries
T
his volume presents new approaches to and interpretations of the first three centuries of the Heian period (794 – 1086), several interrelated aspects of which are reexamined or are analyzed for the first time here, in a group of studies that reach across disciplinary boundaries while sharing a common theme or motif: the real or imagined configuration of “centers and peripheries” and their many manifestations. As we use the term here, “centers and peripheries” can refer to geographical or spatial relationships, but it may also suggest various dynamics in, among, and between institutions and collectives, clans and families, social classes and gender groups, practices and conventions, and other entities, physical and abstract, that come under observation by scholars of this age. This particular part of the Heian period saw some of its most fertile innovations and epochal achievements in literature and the arts, much of the process of the consolidation of aristocratic authority in Kyoto, the emergence of powerful court factions and religious institutions, and important adjustments in the Chinese-style system of rulership as well. At the same time, the era’s leaders faced serious challenges from the provinces that called into question the primacy and efficiency of the governmental system and tested the social and cultural status quo (if, indeed, there was such). This image of crisis needs to be integrated with that of cultural and ceremonial splendor, with the capital as its primary source and site, which many with an interest in Japan will most readily associate with this period. For this reason, a number of scholars have looked outward from the capital (miyako) and toward the provinces (kuni) in recent years in order to understand more about both. In this process, a middle ground of interaction, negotiation, and accommodation has emerged as a site even more key to understanding than any particular center or periphery, be it capital and province or otherwise.1 The capital and its denizens could not have flourished as they did without resources acquired in and from the provinces; those in the provinces followed the cultural leads of those in the capital, but also put pressure on and made their own distinctive contributions to cultural production at home, in adjacent regions, and in the central metropolis itself. When challenges to the prevailing order arose in the provinces, the tremors were felt at every level of the society of the capital and in many of its modes of cultural expression; when structures shifted in the capital, opportunities arose for altering the order in outlying areas as well. The results of these new pressures, intermittent clashes, and ongoing tensions were, of course, changes, some subtle, some more obvious, and not all
| mikael adolphson and edward kamens
at one time or in one place. For these reasons, the authors of the essays in this volume believe that a conceptualization of Heian Japan must not only take into account conditions at the center and the various peripheries, but must also, and above all, address the interrelationships between and among these spheres. The timing of the publication of this volume is especially fitting, given the changes in approaches to texts and sources that have brought the methodologies of many historians and scholars of art, literature, and religion closer together. Hence it is not only a common interest in Heian Japan but also a new awareness of commonalities in our modes of research that have energized this project and have shaped its outcome. While this work self-consciously follows a tradition of edited volumes of essays on premodern Japan, beginning with Studies in the Institutional History of Early Modern Japan (1968), it also breaks with its predecessors in its common focus on one theme explored assiduously and collectively across disciplinary boundaries.2 It is not only in the study of premodern Japan, or in research on this particular period or parts thereof, that the theme or image of centers and peripheries has presented itself as a compelling model, a powerful instigator for new efforts to represent our understandings of the shape and experience of the past — only to resolve or dissolve, chimera-like, as investigations proceed still deeper into the texts and other artifacts of that past that inevitably make the emerging new picture still more blurred and markedly more complex. In the essays in this volume, “center and periphery” serves as a point of departure and in some cases also as a point of return; it emerges as a construct to be contested, redefined, modified, augmented, or elaborated. It is illustrated or mirrored in an array of examples and phenomena that are juxtaposed as singular or plural centers and peripheries, physical and abstract, topographical, institutional, rhetorical, textual, and human. Some of these juxtapositions are explored within individual essays; others will take shape as the investigations of one author or another come into dialogue with still others. This happens, for example, in the encounter between the figures of the wellknown courtier Fujiwara no Michinaga (966 – 1027), treated in detail here by G.Cameron Hurst III, and Taira no Tadatsune (967 – 1031), whose life and career is the topic of Karl Friday’s essay. Michinaga was perhaps the Heian courtier par excellence and the most powerful man at the time when what may be Japan’s greatest literary work, The Tale of Genji (Genji monogatari), was written, while Taira no Tadatsune’s seizure of the provincial headquarters in three provinces in the east from 1028 to 1031 (shortly after Michinaga’s passing) has traditionally been interpreted as a sign of the imperial court’s decline. On the surface, these figures can be said to represent the two contrasting (received) images of the Heian world — tranquil, centralized, and civil at its core; unstable, dispersed, and violent at the edges — and thus to personify the fundamental dichotomy of capital versus countryside. However, as Hurst’s and Friday’s studies show, and as many of the essays in this volume suggest, figures, events, relationships, and policies in Heian society were much more complex than that, and the most interesting part of such stories is often that which is played out in the spaces in between the sites in which such major characters perform and in the often tangled
Between and Beyond Centers and Peripheries |
account of what transpires in or as a result of their interaction. The complexity that is suggested here represents a diffuseness, a multiplicity of forces and loci of action, and a plurality of voices and actors in shifting roles rather than an isolated society centered on one elite, one cultural modus operandi, one set of beliefs, or any seemingly simple pairings of actors or sites. What emerges in place of the former and simpler dichotomous image of a top and bottom, center and periphery Heian Japan is a number of newly observed patterns, configurations, and themes that deserve to be noted for what they indicate in the relations between a variety of centers and peripheries — and in other spaces in between them. If the Heian period was indeed an era of fundamental changes, then there are also some fundamental changes in our images and understandings of it that should be given serious consideration. These are suggested by the essays collected here, and might be organized as follows.
The Early Tenth-Century Turning Point Several of the essays in this volume indicate that there was something of a quiet watershed in the first half of the tenth century, in which those holding power at the political centers had no choice but to adopt a less rigid state apparatus. The result was a new system of communicating, ruling, and administrating that relied primarily on the personal and private powers and abilities of the people involved while still retaining the basic structures of the general social hierarchies. For instance, while the imperial court retained the right to appoint governors, the latter were contracted to deliver taxes at a specific rate in return for less supervision and the opportunity to reap financial benefits. Such adjustments were crucial in that they helped avoid what might otherwise have been a debilitating crisis for the imperial court, which found it increasingly difficult to control local powers and to collect taxes under the procedures of the ritsuryō (penal and administrative code) state. Contrary to traditional interpretations, which saw this change as a sign of the imperial court’s collapse, we see in it a means for the capital elites to maintain, if not strengthen, their administrative control of the various peripheries, by accommodating the needs of enough peripheral powers to keep them at bay. The strongest indicator of the early tenth-century transformation is undoubtedly visible in the imperial court’s administration of taxes and the provinces. As both Bruce Batten and Wayne Farris point out, the court came to shift from direct control to a hands-off tax-farming approach. In Batten’s case, this is evident in Kyoto’s involvement in decisions regarding foreign missions as well as attacks in northern Kyushu. Although one might expect a tighter central grip in times of foreign raids along the shores of Kyushu, the court instead came to rely entirely on the individual capacities of officials selected to serve in the area during frequent raids in the late tenth and early eleventh centuries. If such officials were incompetent or unable to deal with the situation, then a more competent person was simply selected until the problems were dealt with. For Farris, the decline in famine reports in the provinces in the tenth and eleventh centuries is not only
| mikael adolphson and edward kamens
an issue of a lack of sources, but also reflects the same kind of change in management of the provinces. Although it is difficult to ascertain how these famines affected the courtiers and whether they represented a decline in revenues for the capital elites, there can be no doubt of the lessening of direct capital management of the provinces. Joan Piggott’s analysis of Fujiwara no Tadahira’s political career at the court, based in particular on his own diary, also indicates that the first half of the tenth century was a time of transition. Tadahira was the most important courtier for close to two decades and was thus responsible for handling the provincial challenges of Taira no Masakado and Fujiwara no Sumitomo in the 930s. Piggott argues that Tadahira was crucial in institutionalizing the regent’s role in the central government, while utilizing both official appointments and patrimonial ties in dealing with provincial matters. His influence over provincial appointments reveal not a decline of central control of the provinces, but rather a mutual dependence with Tadahira’s maintaining the upper hand. Less than a century later, the capital elites had worked out the kinks of this new management style, and the attitude in the court’s handling of Taira no Tadatsune’s rebellion in the early eleventh century aptly demonstrates the success of the new strategies. As Karl Friday argues, despite the seriousness of Tadatsune’s raids and resistance against the first two captains appointed to arrest him, the court never seems to have panicked or felt overly threatened. Rather, it simply replaced the inept representatives with a courtier who was better equipped to handle the insurgence through his own private resources and connections with Tadatsune. Hurst provides further evidence of this change as he notes that a new system, where public land was divided into smaller units (myō) to facilitate the collection of taxes and dues, was put in place at the beginning of the tenth century. As part of these new procedures, Hurst explains, the imperial court “essentially ceased its involvement in the details of provincial administration in return for provincial governors serving essentially as tax collectors.”3 For the influential Michinaga, the assignment of governorships to his own kin and retainers became an effective and beneficial way of maintaining a close tie with the provinces. In a similar way, Adolphson points out that the political elites gave up attempts to control temples through a state network of approved designations in the tenth century. But while one might expect an uncontrolled growth of religious institutions beyond the reach of the capital elites, the central monasteries of Enryakuji, Onjōji, Kōfukuji, Tōdaiji, and Tōji became the new nexuses to which smaller temples came to look for recognition and support. In short, once the ideal of a state-controlled religious hierarchy was firmly abandoned in the tenth century, the well-connected temples emerged as centers in their own right, and peripheral temples throughout Japan came to be more tightly linked to them as branches. This process was above all fueled by the increasing popularity of the Tendai and Shingon beliefs that had begun a century earlier. Similarly, Ryūichi Abé finds the origins of the ideological emergence of more independent Buddhist strongholds in the early ninth century, when Buddhist texts were disassociated from Confucian modes of interpretation governed by the notion of service to the state. This
Between and Beyond Centers and Peripheries |
disassociation spurred further developments in the sphere of religious discourse, as scholarship and writing were separated from the craft of politics that took place in the tenth century. In short, the neatly centered bureaucratic imperial state was replaced by a multicentered and broadened composite in which both secular and religious elites were no longer controlled by the proscriptions of the ritsuryō state, but were still part of the state itself.
Multiple Centers and Peripheries It is clear from the essays in this volume that, contrary to the traditional view of a dominating and static imperial court that eventually fell victim to its own rigid ranking, a singular dichotomy between center and periphery is inadequate. We find both within Kyoto and in the countryside several centers and various degrees of peripheries with fluid boundaries and a substantial middle ground in-between. Even within the imperial court itself, there was more than one nexus of power. Fukutō Sanae demonstrates in her essay on women as rulers and imperial consorts that although women were relegated to the periphery of the bureaucratic and ceremonial workings of the imperial state from the late eighth century, they continued to wield power behind the scenes in spaces symbolically and physically adjacent to the official halls. In that sense, the Japanese description of the age dominated by regents (ca. 858 – 1086) as a royal court state (ōchō kokka) may accurately reflect the centrality of the capital city of Kyoto (Heian-kyō), but it may simultaneously give the false impression that other significant centers did not exist. For example, during that age the most influential monasteries emerged as centers of power in their own right. As Adolphson demonstrates, the creation of branch networks by powerful temples strengthened the presence of the elites in the provinces while also broadening the center itself, inviting more temples to be part of the pyramid of power. Abé’s analysis of texts focusing on the Heart Sūtra in the eighth and ninth centuries also indicates a broadening of the center as the Buddhist texts themselves were disassociated and became independent from the state, even as they continued to be crucial in rituals sanctifying it. In his study of a surge in production of images of Eleven-Headed Kannon, Samuel Morse focuses on Buddhist teachings that traditionally have been seen as peripheral compared to the Tendai and Shingon schools. He criticizes such interpretations for relying exclusively on textual analyses, where the esoteric schools have gotten most attention, in part because of their subsequent dominance in the second half of the Heian age. Instead, Morse shows that the Kannon cult was yet another important aspect of early Heian Japan, further reinforcing our impression of a multifaceted and multicentered society. If the political center thus appears substantially broader than previously thought, then the periphery must also be characterized as appearing less peripheral. For example, both Batten and D.Max Moerman show that the Dazaifu — the governmental headquarters in northern Kyushu — was a center in its own right. Not only did it serve as the nexus for foreign communication in the Heian age, but it was also the center of foreign trade during a period that has tradition-
| mikael adolphson and edward kamens
ally been seen as isolated. Thus, it earned the epithet “the capital of the western periphery” (saikyoku no daijō), a clear reflection of the Dazaifu’s importance to Kyoto’s own status.4 Thus, even as the court relaxed its direct control of the Da zaifu, its main official, the governor of Kyushu, was a man of and from the court. Indeed, it might be argued that the Dazaifu was so important that had the court lost control of it, it would have had a much more difficult time legitimizing its rule. This notion is further supported by Moerman’s piece on sūtra burials, which establishes the particular importance of northern Kyushu to the capital nobles, as it was the most frequent region for such practices. Temples and shrines in northern Kyushu were moreover tied to the court or temples close to the court, and the Hachiman deity, with origins in northern Kyushu, had its own shrine just south of Kyoto itself. Robert Borgen examines the monk Jōjin’s (1011 – 1081) travels between two political centers in East Asia, Kyoto and Kaifeng, both of which were not accustomed to dealing with one another on an official level. It is clear both from Jōjin’s own account and the Heian court’s reluctance to deal directly with the Chinese court in the eleventh century that the Kyoto elites were unsure how to relate to centers outside Kyoto’s direct sphere of influence. As Borgen demonstrates, the Chinese court was eager to establish more frequent exchanges with its Japanese counterpart, but the Chinese empire presented problems for the Japanese worldview, so Japanese leaders, who allowed commercial and other ties to flourish, avoided formal diplomatic exchanges. At the same time, however, there were no intrinsic tensions in religious terms, since such centers were found in multiple locations both inside and outside the urban centers in both countries. Jōjin himself can thus be characterized as the middle ground that connected political and religious centers in eleventh-century Japan and China. Such interaction would not have taken place without the support of members of the imperial court in Heian-kyō. The notion of an isolated Heian world would thus seem exaggerated and misleading. Indeed, Borgen even argues that a premodern and distinct idea of “Japaneseness” is visible in the Heian age, not because of isolation, but in the context of consistent contact with other centers outside the imagined realm of the Heian state. Both Edward Kamens and Ivo Smits demonstrate that the old notions of Japanese writings at the center and Chinese writings in the peripheries, with women identified with the former and men with the latter in the mid-Heian age, do not hold up to closer scrutiny. Rather, there was considerable crossover between the two spheres, open to both negotiation and co-opting. Kamens finds a terrain of texts and usages of techniques between these two extremes, with a substantial and salient middle ground in-between. Smits shows first of all that learning and literary skills in and of themselves were central to the Heian elites, but he demonstrates that, contrary to earlier assumptions, Chinese writing endured as the core of this literacy. Like Kamens, he rejects the notion of a simple dichotomy between women’s and men’s handwritings, claiming instead that there was much more middle ground than exclusive modes for either sex. The overall impression from several of the essays in this volume is that men
Between and Beyond Centers and Peripheries |
and women of the Heian age recognized spheres within competing centers both inside and outside Japan. Moreover, by the mid-Heian age, this multitude of centers had been allowed to grow and become immersed in the state itself. In other words, a broadening and diversification of the center occurred during the early and mid-Heian age, a development that simultaneously resulted in an advancing integration of centers and peripheries.
Integration of Centers and Peripheries According to the traditional view, the relaxation of direct administration of the provinces in the mid-to-late Heian age necessarily meant a decline of central control and presence in the peripheries. However, as several essays in this volume indicate, the abandonment of the bureaucratic management of the realm actually helped the central elites strengthen their control and make it more effective. The establishment of private estates (shōen), a process that got under way during the age under scrutiny here but that accelerated markedly during the twelfth and thirteenth centuries, for instance, has often been seen as a sign of decline. In truth, however, the patrons and proprietors at the top of the shōen pyramid belonged to the class of capital elites, and their involvement in the collection of taxes and management of the estates through local stewards shrunk the distance between rulers and ruled. Granted, the relationship between the top and management levels was not always smooth, but neither was that between provincial administrators and the court, as evidenced in Charlotte von Verschuer’s chapter. In the first comprehensive analysis of four of the articles in the famous 988 petition from local officials in Owari Province, she highlights the tensions between ambitious governors and those who were entrusted with matters of provincial administration. By contrast, the private administration of public and private land became more effective from the proprietors’ and administrators’ point of view because of the contractual nature of their relationship. And because local notables gained authority by representing central elites, the presence of the center in the peripheries was more likely to be reinforced than weakened. If the balance between the two parties shifted, then the contract could always be negotiated. In short, the privatization of land helped to bring local elites into the centered hierarchy, thus strengthening the integration of centers and peripheries. It must also be noted, however, that integration should not be taken to mean a one-way exploitative relationship. Von Verschuer’s study reveals the important codependent relationship between local administrators and the central authorities, further proving the integrated nature of the relationship between them. We detect an intricate power triangle of both codependence and competing interests with the imperial court, the offending governor, and the local notables. The offending governor was eventually deposed by the imperial court and could thus not continue his innovative tax-collecting strategies without the endorsement of the court. The capital elites were naturally materially dependent on the provincial population, but local officials needed, and could utilize, the court to protect their own position.
| mikael adolphson and edward kamens
In Hurst’s treatment of Michinaga, the shift away from rigid bureaucratic procedures again reminds us of the center’s concern with and strengthened control of the provinces in the eleventh century. And the contributions by the middle-ranking provincial governors (zuryō) and resident officials were significant and important to the court elites, as indicated by the lavish gifts Michinaga received. In Hurst’s words, “The court depended upon the zuryō for their livelihood — taxes, construction projects, and contributions both public and private — while the governors themselves relied upon the nobles, especially the Regents House, for appointments to the posts that guaranteed access to those resources. Both parties seemed to appreciate this dependency.”5 Friday’s treatment of the Tadatsune insurgence also confirms the multifaceted and codependent nature of centers and peripheries, involving farmers and provincial elites in the local arena as well as ranking courtiers and temples in the capital region. The enduring ties between the center and periphery, he explains, were part of a complex and intense “interplay between the rural and urban elites, reflecting a balance between centrifugal and centripetal forces.”6 Moerman notes that a variety of sūtras from various schools were buried at a number of sites, all indicating a multitude of religious centers and a close integration between regional and national centers. The best example is undoubtedly northern Kyushu and in particular the close cooperation and mutual support between the Hachiman deity and Tōdaiji as well as the establishment of Iwashimizu Hachimangū south of Kyoto. The change of administrative strategies of the Dazaifu in the early tenth century provides further evidence of the increased integration and co-opting of regional forces. In Batten’s words, local powers were “bound to the center by their own parochial interests.”7 These are indications not merely of the capital elites’ co-opting powerful regional beliefs but in fact of strong religious and political bonds. The Northern Kyushu periphery was, after all, not so distant from the Kyoto center, nor even all that peripheral. Adolphson uses the term “centering” in describing the creation of temple networks, which came to stand as a successful sign of the central elites’ strategy of tying provincial powers closer to the center, favoring inclusion over exclusion. The difference from the Nara age (710 – 784) was that the representations of the center were private temples, not state-controlled institutions. Morse’s essay also points to an increased integration between the cults in the capital region and the provinces. The artistic production in regional temples has left us with compelling evidence that monks of the Hossō school centered in Nara actively contributed to the spread of the Eleven-Headed Kannon cult among local residents, perhaps even more so than Tendai and Shingon in certain regions. Regional spiritual preferences as well as the artistic and religious ideas of monks educated at one of the religious centers came together in the production of statues at provincial temples. Finally, Smits offers an analysis of how the peripheries were brought closer to the center through texts dealing with social fringes. The absurd was made accessible through comedic texts, and people on the social peripheries, whether poor or by occupation, were brought closer to the center through Chinese writings (kanbun), the main medium of the learned at the center. In this way, the mode
Between and Beyond Centers and Peripheries |
of kanbun on the one hand reflected its centrality, while its usages — topics, categories, and social decorum — all show its depth as a terrain of negotiation and integration with the peripheries.
Privatization The emergence of multiple centers outside the immediate structures of the bureaucratic state reflects a phenomenon some scholars have referred to as a process of privatization, an older interpretation that seems supported by the essays in this volume even if today’s scholars may prefer a different terminology. This was a slow yet clearly visible trend of increased use of extragovernmental means to rule and administer, whose roots could be traced to well before the Heian age. For example, before the Nara age, hereditary ties and private assets played an important role in politics. The “century of reform” from the mid-seventh to the mid-eighth centuries represented an overhaul to replace a conglomerate of powerful families with a state based on bureaucratic rules and codes. John Whitney Hall described the adjustments that began in the late ninth century as “a return to familial authority.”8 It is safe to conclude that these trends toward extragovernmental means of rulership and lordship are visible almost everywhere and that they had a tremendous influence on developments both in the late Heian age and well after. In Fukutō’s piece, we learn that principal imperial consorts exerted enormous influence over political matters at the imperial court, not as empresses, as during the seventh and eighth centuries, but as heads of the imperial family and mothers or grandmothers of reigning emperors. Fukutō sees this development as an important stage in the development of the imperial house itself as a private entity and faction within the court, leading to the rule by retired emperors (insei) that came to characterize much of the twelfth century. The same trend is equally visible within the imperial court in the cases of Fujiwara no Tadahira and Fujiwara no Michinaga as demonstrated by Piggott and Hurst respectively. Hurst’s essay deals directly with the effects of the use of private ties and offices in matters of the state. In Piggott’s contribution, Tadahira stands out as one of the main figures in institutionalizing the usage of such extragovernmental means and personnel. For example, when the Shingon center Tōji complained against alleged intrusions into its property in the Ōyama estate in 920, it filed its petition directly to Tadahira instead of using the official and established channel through the Council of State. The religious sector went through a similar transformation during this age. Beginning with the imperial court’s acknowledgment and sponsoring of privately funded and founded cloisters in the ninth century, influential patrons in the capital and provinces continued to support their own set of religious ceremonies and networks. By the tenth century, the most popular temples came to cater directly to the needs of the capital aristocrats, thus gaining increasing freedom from the religious parameters proscribed in the bureaucratic codes. Abé’s description of Kūkai’s disengaging of Buddhist texts from the state-mandated Confucian paradigm speaks to the beginning of this process, and Moerman’s essay shows that
10 | mikael adolphson and edward kamens
sūtra burials were frequently as concerned with the private welfare of the family of the donor as with the public good. The Fujiwara also had a very active hand in the final release of religious institutions into fully private units. According to Adolphson, the court abandoned its efforts to regulate temples throughout Japan, and allowed instead prestigious or important temples to serve the state and receive official titles, something that the Northern Fujiwara promoted and themselves benefited from. The establishment of private temple networks under the most powerful institutions was the natural outcome of this privatization process. In the end, perhaps it is in the provinces that we find the strongest evidence of the move away from bureaucratic procedures to more reliance on private assets and procedures. As Friday points out, Taira no Tadatsune’s insurgence was not only caused by private competition but was also solved through an effective use of private connections between the appointed officer and Tadatsune himself. Von Verschuer’s essay on the Owari petition lends support to the idea of Kyoto as the political center of the age, but it also reflects changes in its role since the early Heian age. The provincial administration had already reached a considerable degree of privatization, a direct cause of the abuse by the governor, who considered public land and fees as much his own as those of the state. And the court’s lack of interest in making official reports of famines in the provinces, as described by Farris, can in part be understood in terms of less direct involvement by the organs of the state in provincial administration and the adoption of a tax-farming system. If these conclusions and the essays in this volume can further research in the field of Heian studies and encourage new and innovative approaches along multidimensional and multidisciplinary lines, then this work has more than served its purpose. In particular, we find the notion of inclusion rather than exclusion especially significant. Whereas previous accounts have tended to emphasize the isolationist and remote character of the capital elites as well as of Japan itself, the essays in this volume suggest that although still heavily hierarchical, Heian society moved toward a more inclusive system as part of the adjustments that took place in the tenth and eleventh centuries. It was by including emerging powers, be they provincial warriors, temples, or shrines, and recognizing other nexuses that the imperial court survived as long as it did (by some accounts to the late Kamakura age) as a force and source of political authority. As a result, Heian society became more diverse, more diffuse, and more complex through gradual, deliberate, and controlled adjustments. Our studies have made it clear that this process can never be fully understood if the focus remains exclusively on the centers and peripheries as dichotomies; rather, we should direct our attention to, and include the important terrains that lay between, the space of interaction, transformation, integration, and seemingly unlimited potential for further change.
Between and Beyond Centers and Peripheries | 11
Notes 1. The concept of the “middle ground” as not only spatial but also phenomenal and functional is adapted here from Richard White, The Middle Ground: Indians, Empires, and Republics in the Great Lakes Regions, 1650 – 1815 (Cambridge; New York: Cambridge University Press, 1991). See also Bruce Batten, To the Ends of Japan, 9. White states, “The middle ground is the place in between: in between cultures, people, and in between empires and the nonstate world of villages....On the middle ground diverse people adjust their differences through what amounts to a process of creative, and often expedient, misunderstandings. People try to persuade others who are different from themselves by appealing to what they perceive to be the values and practices of those others. They often misinterpret and distort both the values and the practices of those they deal with, but from these misunderstandings arise new meanings and through them new practices — the shared meanings and practices of the middle ground” (p. x). It is in these shared meanings and practices that we may be able to understand the commonalities of those living in the various centers and peripheries. 2. Studies in the Institutional History of Early Modern Japan, ed. Marius Jansen and John Whitney Hall (Princeton, N.J.: Princeton University Press, 1968). Other works in this genre include Medieval Japan: Essays in Institutional History (New Haven, Conn.: Yale University Press, 1974), ed. Jeffrey P. Mass and John Whitney Hall; Japan in the Muromachi Age, ed. John W. Hall and Toyoda Takeshi (Berkeley: University of California Press, 1977); Japan before Tokugawa, ed. John Whitney Hall (Princeton, N.J.: Princeton University Press, 1981); Court and Bakufu in Japan: Essays in Kamakura History, ed. Jeffrey P. Mass (Stanford, Calif.: Stanford University Press); The Bakufu in Japanese History, ed. Jeffrey P. Mass and William B. Hauser (Stanford, Calif.: Stanford University Press); and The Origins of Japan’s Medieval World, ed. Jeffrey P. Mass (Stanford, Calif.: Stanford University Press, 1977). 3. See Hurst, p. 67 in this volume. 4. See Moerman, p. 258 in this volume; Nihon Montoku tennō jitsuroku, Ninju 2 (852) 2/8. 5. See Hurst, p. 91 in this volume. 6. See Friday, p. 349 in this volume. 7. See Batten, p. 381 in this volume. 8. See John Whitney Hall, Government and Local Power in Japan 500 to 1700, 99; and Jeffrey P. Mass, Antiquity and Anachronism in Japanese History, 13 – 14.
part i
Locating Political Centers and Peripheries
2
d Fukutō Sanae, with Takeshi Watanabe
From Female Sovereign to Mother of the Nation Women and Government in the Heian Period
T
he Heian age stands out for the contributions made by noblewomen in the production of literary works and as sponsors of the arts and religious ceremonies. Their participation in political decisions at the center, however, has traditionally been seen as limited, especially when compared to the two hundred years between Empress Suiko’s accession (592) and Empress Shōtoku’s death (770), when men and women reigned in equal numbers and shared similar spans of rule. This context seems to be unique to ancient Japan as one is unlikely to find this many female rulers during such an extended period in other premodern societies. Still, scholars have until now understood female sovereigns primarily as interim rulers, who came to reign when confusion surrounded the succession of a male to the throne or when the heir was still a child. According to this view, female sovereigns wielded little actual power as rulers because they owed their position to male sponsors. However, recent research now suggests that empresses matched emperors in demonstrating real authority, though their accession may have been considered provisional.1 Still, a momentous change occurred at the end of the eighth century, when female sovereigns came to yield less power in the government and all but disappeared from the throne. But they continued to exert considerable influence through other means, as mothers of emperors, and were often granted titles such as kōtaibunin (imperial mother) or kōtaigō (grand imperial dowager). Commenting on these new positions, Satō Nagato recently observed, We should not exaggerate the influence of the temporary courts of the grand imperial dowagers (kōtaigō) during this period. Although one might expect that her opinion be solicited for political matters because of her daily contact with the emperor, we must not automatically conclude that she was actively participating in matters of the state. And, when the emperor was an adult, other than in minor matters, she was not directly or publicly participating in governmental affairs, even if she maintained a strong relationship with the emperor.2 Satō accordingly refutes the notion that imperial consorts were involved politically. It is true that the imperial government structured itself around male sovereigns, especially after the reign of Emperor Kanmu (r. 781 – 806). From that time on, women were excluded not just from rule, but also from the bureaucracy,
16 | fukutō sanae with takeshi watanabe
and were driven from the center of politics to the peripheries. However, Satō’s interpretation raises several important questions. Did empresses, who until the eighth century stood on nearly equal terms with their male counterparts, really relinquish their political rank? Did this marginalization of official duties within the imperial court mean a complete loss of influence? In short, did female participation in government truly disappear? In this chapter, I will explore the place of women within the Heian court government from the ninth through the eleventh centuries, a period usually seen as one dominated by rule of regents and known as the royal court state (ōchō kokka). I will examine the sociopolitical relationship between men and women at the core of the Heian state. To understand these developments, an overview of the status of female rulers in the preceding era will be helpful.
Women and Government in the Seventh and Eighth Centuries: The Era of Female Sovereigns In recent years, research on female sovereigns has flourished along with scholarly interest in imperial rule in the seventh and eighth centuries. This has allowed us to reevaluate the role of women as rulers during this early period. The first verifiable female sovereign was Empress Suiko (r. 592 – 628), who ascended the throne in 592.3 Her road to rulership was complicated, but she began her career at court as the principal consort of Emperor Bidatsu (r. 572 – 585), who died in 585. After his death, his half brother Yōmei ruled, but he died after only two years on the throne. Another brother, Sushun, then ascended the throne in 587, but power struggles between the Mononobe and Soga clans led to his assassination five years later. It was in the midst of such tumultuous dispute over the imperial succession that influential court aristocrats selected the previous consort as their ruler, thus making Suiko empress. The influential historian Inoue Mitsusada asserted that it became customary for consorts to assume leadership as female sovereigns during such succession disputes. Inoue’s ideas came to be known and accepted as the “provisional female sovereign theory” (jotei chūkei setsu).4 Although Empress Suiko did in fact become sovereign during such circumstances, this theory has as its backdrop a gender-biased presumption that women could not exercise political authority. Not only were female sovereigns provisional in this line of thinking; they were also merely puppets for their male sponsors. The provisional female sovereign theory has recently been subject to reexamination. Newer research shows that while the circumstances may have dictated their accession, and their role may have been considered provisional, female sovereigns actively possessed and exercised political power equal to that of their male counterparts.5 For example, the consort of Emperor Tenmu (673 – 686) is believed to have fought alongside her husband during the Jinshin War of 671 – 672, when they battled Tenmu’s nephew for the throne. After the death of their son, Prince Kusakabe (662 – 689), she became Empress Jitō (645 – 702; r.
From Female Sovereign to Mother of the Nation | 17
Figure 2.1. Imperial genealogy of the seventh and eighth centuries
690 – 697), until her grandson, Prince Karu, who would rule as Emperor Monmu (r. 697 – 707), reached adulthood (figure 2.1). During her reign, Empress Jitō promulgated the Asuka Kiyomihara Law Codes and transferred the capital to Fujiwara, in effect establishing the foundation of the bureaucratic ritsuryō (penal and administrative codes) state. These accomplishments are commonly accepted, and would seem to contradict Inoue’s interpretation.6 Moreover, recent scholarship indicates that Empress Jitō not only ascended the throne without the customary selection process headed by the nobles, but also removed competitors, such as Tenmu’s sons by other consorts, and forcefully took control by herself.7 In the following century, political condi-
18 | fukutō sanae with takeshi watanabe
tions changed as nobles close to the throne gained more influence, but empresses continued to exercise authority in the same fashion. For example, both Empress Genmei (r. 707 – 715), who ascended the throne without any experience in governing, and Empress Kōken (r. 749 – 758), who was a forceful sovereign groomed to rule, ascended the throne without having ruled alongside a male emperor as principal consort. Of the female sovereigns who ascended the throne in the seventh century, many were imperial dowagers (ōkisaki). Women were esteemed within the imperial family, and as the mother of the emperor, the ōkisaki was involved in determining the succession. In the imperial bureaucratic state of the seventh century, the principal imperial consort (kōgō) assumed the role of imperial dowager. As suggested by the phrase “rule from behind” (shirie no matsurigoto),8 the principal consort stood equal in importance to the sovereign in state affairs and inherited all the authority of the throne after the death of the emperor. This provides an interesting contrast with China, where only the kōtaigō retained an official rank, and presided over a special court to support the child emperor. Even if she were not the birth mother, she had to hold that title to effectively rule in his place while helping the young emperor maintain his position. In Japan, the Taihō Law Codes of 701 stipulated that even if the birth mother of the emperor did not become the principal imperial consort, she could become either senior imperial consort (kōtaihi) or imperial mother. These titles made it possible for nonprincipal consorts to wield the same authority and power as the empress dowagers to support the emperor. In other words, it allowed the imperial consort to exercise power in her own name by prolonging the emperor’s reign after his death.9 The eighth century saw new developments that further augmented the role of imperial consorts. The principal consort of Emperor Shōmu (701 – 756; r. 724 – 749), Fujiwara no Kōmyōshi (701 – 760), established the Office of the Imperial Consort (Kōgō gūshiki) outside the inner Imperial Palace, which was independent of the emperor’s bureaucracy and maintained its own assets.10 Kōmyōshi became Grand Imperial Dowager Kōmyō when her and Shōmu’s daughter ascended the throne as Empress Kōken in 749. After Shōmu’s death in 756, Kōmyō came to support her daughter’s rule as mother of the empress, and it was her nephew, Fujiwara no Nakamaro (706 – 764), who took command of court affairs by becoming chief of the Office of the Imperial Consort. Women who participated in politics did not consist only of imperial consorts. In the seventh century, female officials (miyabito) served at the center of the palace, and they were organized into twelve bureaus of the rear palace, in accordance with the ritsuryō codes. Their main responsibility was to serve as channels of communication between the court council and the emperor or empress. The miyabito continued to act as the emperor’s liaison to the bureaucracy into the eighth century.11 In this way, male and female officials both worked as one to run the government; and on ritual occasions, men and women took on comparable work, though they were segregated.12 Empress Kōken retired in 756 in favor of Junnin, but she reascended the throne in 764 as Empress Shōtoku, reflecting the uncertain and competitive conditions
From Female Sovereign to Mother of the Nation | 19
at the imperial court at the time as she asserted her own control at the expense of her Fujiwara opponents. In either case, her second reign marks an important boundary, since the position of women in the court would change drastically after her death in the tenth month of 770, followed by the ousting of her main ally, the monk Dōkyō. First, we find no more female sovereigns following Shōtoku until they appear as puppet empresses in the Edo period, but as we shall see, this is only part of the story.13 Without a doubt, the loss of imperial power signaled a fall in the political standing of women. Second, the office of the principal consort, which had been located outside the inner palace and functioned independently, was moved inside the main palace area, beginning with the principal consort of Emperor Kōnin (r. 770 – 781), Princess Inoue.14 The principal consort, who used to stand alongside the emperor as a ruler and who used to have a right to succeed after her husband’s death, now stood below her spouse. From this time, the prerogatives of the throne rested solely with the male emperor. Third, the governmental responsibilities of female officials gradually lessened. Conversely, and in part to replace the roles of these female officials, an increasing number of consorts and relatives of the imperial family began serving at the court. In short, while the status and role of female officials began to erode in the public domain, the participation of noblewomen in the private sphere continued to be prominent, if not increased.
Women and Government in the Ninth Century: The Power of the Kokumo From the age of Emperor Kanmu to that of Emperor Saga (r. 809 – 823), the political role of women and their standing changed greatly. Female sovereigns no longer appear from the late eighth century, and the role of principal consorts changed as well. The introduction of the imperial consort’s palace inside the inner Imperial Palace at the very end of the Nara period under Princess Inoue continued and was institutionalized in the new capital of Nagaoka (784 – 794) and in Heian-kyō. A hall named the Jōneiden was designated as an office for the public affairs of the imperial consort in the center of the rear palace within the palace compounds of Kyoto (map 2.1). On occasion, this hall also came to be used by the imperial consort for various ceremonies.15 Thus, although the Office of the Imperial Consort was incorporated into the inner Imperial Palace, indicating a diminished formal function, the public rituals held at the Jōneiden reflect a continued official function still played by the imperial consort. On the other hand, Tachibana no Kachiko (786 – 850), the imperial consort of Emperor Saga, had the emperor’s officials running her office. When one considers how Grand Imperial Dowager Kōmyō had her own freestanding office, the loss of independence for the imperial consorts and mothers seems clear.16 The change of the imperial consort’s administrative office also meant a change in her role. Fujiwara no Otomuro (760 – 790) was appointed kōgō of Emperor Kanmu on the fourteenth day of the fourth month of 783 because of her status as the mother of Kanmu’s eldest son, later Emperor Heizei (figure 2.1). Emperor
Map 2.1. The Imperial Palace compound (dairi)
From Female Sovereign to Mother of the Nation | 21
Figure 2.2. Imperial genealogy of the early and mid-Heian period
Saga, whose imperial consort was Tachibana no Kachiko, instituted a number of Chinese-inspired reforms with an emphasis on virtue, and there was a renewed emphasis on the spouse of the emperor in accordance with Chinese notions of how each gender had its respective roles.17 Until that point, Japan had developed its own tradition by having the imperial consorts, along with the emperor, receive white robes for sacred ceremonies (haku no kinu), in stark contrast to the habits of Tang China. However, in the second month of 820, it was stipulated that the imperial consorts’ robes be “auxiliary” (josaifuku). It is thus beyond any doubt that the position of the imperial consort became secondary to the emperor’s from Emperor Saga.18 From the time of Tachibana no Kachiko onward, the phrase “within the palace” (konchū) was added to edicts appointing imperial consorts: “I hereby declare
22 | fukutō sanae with takeshi watanabe
[Tachibana no Kachiko] imperial consort and the ruler within the palace.”19 In other words, the imperial consort’s authority was defined as being within the compound of the rear palace, reflecting a major change in her function and status.20 Until then, the imperial consort stood next to the emperor in the Audience Hall (Daigokuden) to hear the New Year’s greetings on the morning of the first day of the year. However, from the time of Kachiko, the imperial consort withdrew and no longer participated in the audience on New Year’s Day. Instead, a new ceremony was initiated on the second day, when she received New Year’s greetings from the crown prince, retainers, and female officials, a format that followed the Chinese model for the imperial consort and empress dowagers.21 In China, however, the imperial consort received greeting only from higher-ranking female attendants, whereas in Japan, all female attendants participated, symbolizing her position at the apex of all ranking women at the imperial court.22 A political transformation among female officials also occurred within the imperial court. From the New Year’s Day greetings to many other ceremonies and banquets, the participation of women, who at first stood side by side with men, lessened between the time of Emperor Kanmu and the early Heian period.23 The promotions ceremony (joi no girei) used to occur with women and men gathered together in rows at the court. However, from 813, the ceremonies were divided, with the men holding their ceremony on the seventh and the women on the eighth, although in reality the women’s ceremony became an all but empty shell as the actual lining up of the women to be promoted was eliminated.24 Rather, the female officials (naishi), who formerly held responsibilities in the administrative apparatus of the state, came to be replaced by male secretaries known as kurōdo in the aftermath of the Kusuko Incident of 810, which pitted Retired Emperor Heizei against Emperor Saga.25 From then on, as female officials came to serve the emperor mainly in minor matters, their political role declined.26 The responsibilities at the court, which had formerly been executed by both men and women, were now divided: the political authority rested with men, while the private, including the sexual, rested with women.
New Developments in the Latter Half of the Ninth Century As noted, in the early Heian period, especially at the court of Emperor Saga, the roles of men and women underwent a radical transformation. In the latter half of the ninth century, there would be further change. First, the power of the mother of the reigning emperor, as kōtaigō or kōtaibunin, became more pronounced. The change is first apparent on the fifteenth day of the seventh month of 842, two days after Retired Emperor Saga died. In what came to be known as the Jōwa Disturbance, Crown Prince Tsunesada was deposed under pressure from the Fujiwara. However, the imperial decree following the incident also notes that the decision was made in accordance with “the words of the senior grand imperial dowager [taikōtaigō].”27 To further complicate matters, Kachiko was the grandmother of Tsunesada, and the prince’s mother, Princess Seishi, was naturally upset with her own mother. Seishi is said to have “shaken in anger, cried out in grief, and re-
From Female Sovereign to Mother of the Nation | 23
sented her mother.”28 The involvement of Kachiko in this incident indicates that, after Saga’s death, she assumed actual control of the imperial house.29 The new crown prince following the Jōwa Disturbance was Michiyasu, the son of Fujiwara no Yoshifusa’s sister Junshi and Emperor Ninmyō, reflecting also the increased influence of the Northern Fujiwara (see figure 2.2). Since “the ranking nobles repeatedly submitted [that Michiyasu be appointed],”30 it is clear that the twelve ranking courtiers supported and accepted Kachiko’s decision. Emperor Ninmyō and Princess Seishi were twins, but in choosing Prince Michiyasu, Kachiko favored the paternal line, a stance apparently shared by the courtiers and obviously strongly promoted by the influential Yoshifusa. Prince Michiyasu ascended the throne as Emperor Montoku (r. 850 – 858). Extant documents do not reveal exactly how his birth mother, Fujiwara no Junshi, supported Montoku, but the imperial edict four years later promoting her to kōtaigō offers a glimpse. On the twenty-first day of the third month of 850, Emperor Ninmyō died in his imperial palace. Crown Prince Michiyasu then moved into the Eastern Palace (Tōgū), where he received the imperial regalia (map 2.2). On the seventeenth day of the fourth month, Junshi was promoted from consort (nyōgo) to imperial mother. The imperial proclamation of four years later explains how this came about: The senior grand imperial dowager of Saga [Tachibana no Kachiko] and the grand consort of Junna [Princess Seishi] were both present at the imperial court at the same time. Although it is clear from ancient customs that I [Emperor Montoku] revere my mother [Fujiwara no Junshi], I can only lament that merit in life and death are different. Thus, I could only grant my Fujiwara mother the title kōtaibunin. I hope that this modesty will invite happiness in the next world.31 This statement reveals that Emperor Montoku had actually wanted to give his mother the honorific title of kōtaigō from the beginning, but he could not, since two senior consorts were already present. Having no choice, he thus made her kōtaibunin. On the fourth day of the fifth month of 850, however, Senior Grand Imperial Dowager Tachibana no Kachiko died. Four years later, on the twenty-sixth day of the fourth month of 854, Emperor Montoku was finally able to elevate Junna’s consort Seishi to senior grand imperial dowager and Junshi to grand imperial dowager. To modern readers, such squabbling may appear pedantic, especially since both titles in question referred to the imperial dowager. However, in the world of the Heian court, such titles not only reflected status and hierarchy, but also provided important precedents for future successors to the throne and thus had a tremendous impact on the fate of individual noble families. It is therefore not surprising that Princess Seishi disagreed with attempts to promote Junshi. In fact, when Seishi herself was promoted to taikōtaigō, she declined the honor because she did not want to vacate her current kōtaigō title, which would then have been granted to Junshi.32 There was naturally some tension between Seishi and Junshi, since the latter’s son had become crown prince at the expense of
Map 2.2. The Greater Imperial Palace precincts (daidairi)
From Female Sovereign to Mother of the Nation | 25
Seishi’s own son in the Jōwa Disturbance. In any event, one can tell that, at this time, the kōtaigō was considered higher in rank than the kōtaibunin even though they occupied the same rank according to the Taihō Codes.33 The elevation of the kōtaigō in the latter half of the ninth century above that of the kōtaibunin is probably derived from the Chinese classics, which mention grand empress dowagers running the imperial court. Following the death of Emperor Montoku in the eighth month of 858, Junshi is recorded as doing the following: On the twenty-seventh day, the kōtaibunin34 [Junshi] went to the Eastern Gojō Palace [where the crown prince resided]. Two days later, the crown prince and the kōtaibunin returned to the crown prince’s eastern palace, riding in the same palanquin. The procedure was the same as for imperial progressions, except that no announcement was made for the cart. She did this to support the accession of the young prince.35 Nishino Yukiko believes that this is a strong indication of Junshi’s support in making her young grandson emperor.36 In particular, the words “to support the accession of the young prince” unmistakably point to the sovereign’s grandmother, Junshi, as the head of the imperial house, supporting the first documented accession of a child emperor. She accordingly played a role similar to that played by Tachibana no Kachiko, dominating matters at court as the head of the imperial house. This notion is further reinforced by her placement behind the elevated throne during the enthronement ceremony held at the enthronement hall, an expression of her central role in the accession of the child emperor.37 The following year, Junshi moved to the Eastern Gojō Palace, after which Fujiwara no Meishi (also read Akiko), the mother of Seiwa (r. 858 – 876), came to live with and support the young sovereign. Meishi became imperial mother on the seventh day of the eleventh month of 858, less than three months after her son’s accession to the throne.38 Emperor Seiwa’s coming-of-age ceremony was held on the first day of 864, a few days after which Meishi was promptly promoted to grand imperial dowager. Following Emperor Seiwa’s move from the Eastern Palace to the Jijūden, centrally located within the Imperial Palace compound, on 865/11/4, Meishi moved to the Jōneiden soon afterward (see map 2.1).39 As already noted, the Jōneiden was the official administrative hall of the principal imperial consort from the late eighth century, but it was now taken over by the mother of the emperor, who became known as “mother of the nation” (kokumo), a new title that reflected the position of imperial mothers from the time of Princess Seishi. In fact, the kokumo subsequently became the main supporter of the emperor’s rule.40 Thus, while Emperor Saga promoted the sinified ideal of an empress dowager controlling the court, it was the mother of the nation that came to assume that role. Meishi moved to the Shiki no Mizōshi (located just east of the imperial compound; see map 2.2), in the northeastern part of the palace compound, where the office of the grand imperial dowager was located, on the seventeenth day of the second month of 874. The Shiki no mizōshi was also where the influential
26 | fukutō sanae with takeshi watanabe
Chancellor Fujiwara no Mototsune (836 – 891), Meishi’s brother, had his imperial palace office. An entry in Nihon sandai jitsuroku from 881 states, After being appointed chancellor, [Fujiwara no Mototsune] declined the honor four times.41 During that time, the Council of State was deadlocked in many matters. The noble council therefore held meetings, and sent a secretary to report governmental matters to Mototsune at his office within the palace (he used the Shiki no Mizōshi for his office). After Mototsune, this became customary.42 The quarters of sisters or daughters serving at the palace often became the offices of noblemen, and so Mototsune, the most powerful courtier of his time, made his office in his sister’s Shiki no mizōshi. By sharing space with Meishi, who was the mother of Emperor Seiwa as well as the grandmother of Emperor Yōzei (r. 876 – 884), Mototsune could use the authority of the mother and grandmother of the emperor to solidify his own power, acting as the kokumo’s deputy.43 Therefore, the Northern Fujiwara chieftains’ successful rise to power as regents, and by extension the regency age (858 – 1086) itself, occurred against the backdrop of the authority of the “mother of the nation.” The origins of the regent’s power have in fact not received enough attention by scholars. It is important to note, for example, that the regent could control the imperial court from the Shiki no Mizōshi, which he shared with the kokumo — usually the mother or grandmother of the sovereign — precisely because of the kokumo’s history of supporting the emperor. At the beginning of the ninth century, consorts and women were no longer allowed to ascend the imperial throne themselves: women had been pushed from the center toward the periphery of the government. Instead, the emperor’s mother assumed a new status as mother of the nation and came to take power behind the throne, a position that became even more powerful in the case of child emperors. It was against the backdrop of such parental authority and informal influence on reigning emperors that Fujiwara fathers and brothers were also able to exercise power.
The Mother of the Nation in the Tenth Century: Princess Hanshi and Fujiwara no Onshi From the latter half of the ninth century, the imperial mother thus assumed the role of the caretaker of the child emperor. Her role gradually became institutionalized, a trend that was further reinforced when the imperial mother’s office also came to serve as the living quarters of the kokumo. This trend continued in the tenth century.44 The mother of Emperor Uda (r. 887 – 897), Princess Hanshi (853 – 900), exerted great influence over the selection of a consort for her grandson, Emperor Daigo (r. 897 – 930). It is well-known that she made her daughter, Princess Ishi, the consort of Emperor Daigo.45 A later source explains: “Retired Emperor [Uda] received a command from his mother [Hanshi], and stopped the imperial consort [Onshi] from entering the palace.”46 From this statement, we can see that
From Female Sovereign to Mother of the Nation | 27
Retired Emperor Uda followed his mother’s orders regarding the selection of a consort for his son, Emperor Daigo. If an imperial consort gave birth to a prince, the child would naturally become a candidate for the throne. The selection of a consort was consequently of great importance for the imperial succession and the lineage, and the imperial mother’s ability to influence such decisions reflects her powerful position in the mid-Heian age. Emperor Daigo greeted his grandmother, Princess Hanshi, and Retired Emperor Uda on the third day of the seventh month of 897, when he entered adulthood and ascended the throne. When the new emperor moved to the imperial residence at the Seiryōden, the previous emperor moved to the Kokiden, a palace just north of the imperial residence.47 Just after that, on the ninth day of the eighth month, Princess Hanshi and Retired Emperor Uda both moved to a palace at Higashi Sanjō where they lived together until the second month of the following year, when Uda moved to the Suzaku Palace, located south of the Greater Imperial Palace area.48 From their choice of residences, we can deduce that Hanshi, as the mother of the nation, lived with her son not only while he ruled, but after his reign as well, continuing to exert her influence. The mother of Emperor Daigo, Fujiwara no Inshi, had died by the time of Emperor Daigo’s accession, so the daughter of Mototsune, Fujiwara no Onshi (882 – 907), became his adopted mother. She was promoted to imperial mother “and her office was established the same day,”49 when Onshi moved from the Imperial Palace to the Horikawa Palace in the Higashi Sanjō area. The following year, she “moved from the Gojō Palace to the Suzaku Palace,” an indication that she was not active in political matters.50 She did not, in other words, assume the role of mother of the nation. It was another Fujiwara no Onshi (885 – 954), the birth mother of Emperor Suzaku (r. 930 – 946), who proceeded to strengthen the power of the kokumo. After the untimely deaths of the crown princes Yasuaki and Yoshiyori, widely believed to have been caused by the angry spirit of the unfortunate Sugawara no Michizane (845 – 903), the crown prince who was to become Emperor Suzaku was raised with much care and caution.51 In 930, Suzaku became emperor at the mere age of nine. Seven days after the enthronement ceremony, Retired Emperor Daigo died; the following year, Retired Emperor Uda died as well, leaving Onshi as the head of the imperial house to look after Suzaku. Onshi continued to reside with her son not only after his adult initiation ceremony, but also after his accession, as well as after the designation of imperial consorts.52 Onshi’s influence over imperial succession matters aptly reflects her high status. According to the Ōkagami, a chronicle describing the rise to prominence of the Northern Fujiwara, Onshi appealed to Emperor Suzaku about the selection of his younger brother, Prince Nariaki (926 – 967). Nariaki ascended the throne as Emperor Murakami (r. 946 – 967), and Onshi continued to exert great influence on succession matters after his enthronement as well. For example, when Murakami wished to make his son by Fujiwara no Anshi, Prince Norihira (950 – 1011), heir to the throne, Onshi mediated between Retired Emperor Suzaku and Murakami, ensuring the stability of the imperial lineage. According to Satō Nagato, Onshi “privately discussed the matter with Morosuke [the Fujiwara
28 | fukutō sanae with takeshi watanabe
chieftain], and demanded cooperation from Suzaku” in order to make Norihira crown prince.53 Emperor Reizei’s maternal grandfather, Minister of the Right Fujiwara no Morosuke (908 – 960), describes in his journal Kyūreki how his daughter’s words carried great weight in the selection of Prince Norihira as the heir.54 In fact, Prince Norihira was born at Morosuke’s residence on the twenty-fourth day of the fifth month of 950. Upon receiving the news, Emperor Murakami was delighted to have an heir from a ranking consort, although a lower-ranking consort, Fujiwara no Yūhime, had already given birth to another prince (Hirohira). Murakami immediately ordered prayers for the newborn prince’s protection and, in his response to the news of the birth, expressed a strong desire to make him the crown prince.55 Onshi’s response was said to be the same as the emperor’s. The gifts brought to the third-night banquet celebrating the birth also reveal Murakami’s intention to name Norihira the crown prince. Morosuke noted with a mix of joy and concern that the gifts Onshi had received upon giving birth to Emperor Daigo’s heir, Crown Prince Yasuaki, were the same, confirming in Morosuke’s view the likelihood of Norihira’s becoming emperor. By the same token, however, it also reminded him that Yasuaki had died before it was time for him to ascend the throne.56 According to Morosuke’s diary entry for the tenth day of the sixth month of 950, Murakami was already pushing hard to name Norihira crown prince, while Morosuke suggested that the following year would be more suitable. The details of this negotiation are complicated, but suffice it to say that a direct meeting between the emperor and Morosuke occurred on the fifteenth day of the sixth month. At that meeting, Murakami ordered that the investiture of the crown prince be held the following month, and that Onshi be notified. Ten days later, Murakami noted to Morosuke that he had sent a letter to Onshi and received her reply stating that the ceremony should be carried out with care, since Onshi was concerned about a possible falling out between Murakami and his brother, the former Emperor Suzaku. On the twenty-seventh, when she and Morosuke met face to face, Onshi expressed her doubts, saying, “Recently, I have heard rumors of an investiture of a crown prince of Emperor Murakami. I stated that Retired Emperor Suzaku should be informed if this ceremony is to take place soon, yet the emperor says he will only report it after it has taken place. I don’t know what he is thinking.”57 Satō claims that Onshi “met secretly with Fujiwara no Morosuke and sought Suzaku’s cooperation regarding Prince Norihira’s investiture.”58 However, the reason for her asking Murakami to inform Suzaku about Prince Norihira was not to seek his assistance. Until then, Retired Emperor Suzaku did not have a child, but that year, his imperial consort, Kishi, gave birth to a princess.59 Therefore, there was still a strong possibility for the young Retired Emperor Suzaku to father imperial offspring. Onshi, the birth mother of Suzaku, must have considered that a son who would be eligible to become crown prince might be born as well. For this reason, Onshi considered Suzaku’s approval for Prince Norihira’s investiture a necessary step to avoid a rift within the imperial
From Female Sovereign to Mother of the Nation | 29
family. In the end, Emperor Murakami did not follow Onshi’s advice, and secretly made his son crown prince without obtaining his elder brother’s consent. From this sequence of events, one may argue that the mother of the nation did not have much say in the selection of the crown prince. However, I would like to emphasize that Murakami did speak with the kokumo about the investiture of the crown prince, and sought her approval. On the other hand, he kept the matter a secret from Retired Emperor Suzaku. I concur with the interpretation of Fujiki Kunihiko, who stated that Onshi’s “opinion as the most senior member and actual head of the imperial family was sincerely respected.”60 It should also be noted that Onshi’s influence over the selection of consorts (kisaki) is another salient manifestation of her powers as kokumo.61 Such influence was not as publicly visible as in issues of imperial succession, yet it was tremendously important in shaping the imperial lineage.
Fujiwara no Senshi (961 – 1001) In the Eiga monogatari, a tale glorifying the rule of the Northern Fujiwara, we find a famous story of how kokumo Higashi Sanjō’in (Fujiwara no Senshi) forced her way into the bedchamber of her son, Emperor Ichijō (r. 986 – 1011), and coerced him into granting her brother Michinaga (966 – 1027) private inspection powers (nairan), a move tantamount to naming him a regent.62 From this story, we see how Senshi was able to intervene in important personnel appointments at the imperial court. Her influence over appointments can be seen in other rec ords as well. On the fifth day of the seventh month of 997, Fujiwara no Michi tsuna (955 – 1020) was promoted to the post of grand councilor (dainagon), even though he had just become middle councilor (chūnagon) in the fourth month of the previous year. Fujiwara no Sanesuke (957 – 1046) had served longer and was senior to Michitsuna as middle councilor, and would thus be ranked higher, since he had been named supernumerary middle councilor (gon no chūnagon) in the eighth month of 997. Sanesuke became upset and resentfully complained about Michitsuna: “He has written his name for yet a little while, and knows neither one nor two [about matters at the imperial court].” He then added, “The emperor’s officials control matters of the state, as the imperial mother (bokō) makes affairs of the court solely her own.”63 From Sanesuke’s critical remarks, it is clear that ranking courtiers were well aware of the kokumo’s influence over appointments. In fact, many more documents exist demonstrating Senshi’s intervention in court personnel appointments. The aforementioned Sanesuke was on the receiving end of her grace in 990, when he was promoted to junior third rank for his work in supervising the constructing of gates in the palace area. In the previous year, Regent Fujiwara no Kaneie (929 – 990) noted to Sanesuke, “I have added you to the project of building gates, since Senshi has recommended you. Therefore, you were appointed.”64 Other more senior courtiers, such as Fujiwara no Yasuchika and Fujiwara no Kinsue, had also hoped for this appointment. Kinsue even
30 | fukutō sanae with takeshi watanabe
had the recommendation of his elder brother, Minister of the Right Fujiwara no Tamemitsu, but the decision was made in accordance with Senshi’s wishes. Having achieved junior third rank, Sanesuke promptly went to Senshi’s palace to express his gratitude. Experiences such as these must have prompted him to write, only a few years later, that “the imperial mother makes affairs of the court solely her own,” regarding Michitsuna’s rapid rise in the court hierarchy, which at that point, ironically enough, worked against Sanesuke himself.65 In either case, such interventions were commonplace during Senshi’s time. As with many of her predecessors, Senshi exercised tremendous authority in the selection of consorts. When Fujiwara no Shōshi (988 – 1074; later known as Jōtōmon’in) was appointed principal imperial consort on the twenty-fifth day of the second month of the year 1000, it was Senshi who made the formal appointment. The decision was made late in the previous year, when the head chamberlain (kurōdo no tō) of Emperor Ichijō, Fujiwara no Yukinari (972 – 1027), served as an emissary, delivering messages between the emperor, Senshi, and Fujiwara no Michinaga, Shōshi’s father. It was also decided that her investiture as consort should be made even though Ichijō already had a principal imperial consort.66 On this occasion, Yukinari forcefully argued that having two principal consorts for one emperor was acceptable, especially since the Fujiwara-born consorts had all taken the tonsure and were thus dedicated to Buddhism. They would therefore not be able to worship the Fujiwara ancestral gods as required. In other words, he claimed that although there was already a current principal consort, it would be appropriate to appoint another one, who could be devoted to the Fujiwara kami. This argument came to serve as the foundation for the principle of two principal consorts for every emperor henceforth, and it is well-known that Michinaga was greatly impressed with Yukinari’s argument. It made it possible to initiate an unprecedented arrangement, which beyond any doubt must have had the consent of the mother of the nation, Higashi Sanjō’in Fujiwara no Senshi.67 Finally, it should also be noted that Fujiwara no Senshi was the first mother of the nation to become a “retired lady” (nyoin), by taking the tonsure, yet remaining active in palace politics. In part, this took place because the three positions of taikōtaigō, kōtaigō, and kōgō were becoming less effective. By leaving the three posts to her junior consorts, she not only secured their future positions,68 but she also began to demonstrate the authority she could exercise as a nyoin, as the matriarch of the imperial family.69
Jōtōmon’in Fujiwara no Shōshi: Bridging the Regency Government and Rule by Retired Emperors Acting as the head of the imperial house, Jōtōmon’in Shōshi was the mother of the nation who exercised the most authority by looking after her grandsons and great-grandsons. Shōshi was born the daughter of Fujiwara no Michinaga and Minamoto no Rinshi in 988. In 999, at age eleven (twelve according to contemporary Japanese custom), she entered the palace of Emperor Ichijō as his consort. The following year, on the twenty-fifth day of the second month, she
From Female Sovereign to Mother of the Nation | 31
was named a principal imperial consort. Another principal consort, Fujiwara no Teishi (977 – 1000), was eleven years her senior and a rival. However, Teishi died while giving birth on the sixteenth day of the twelfth month of 1000, and Shōshi then adopted Teishi’s son Prince Atsuyasu. In 1008, Shōshi gave birth to Prince Atsuhira (later Emperor Go-Ichijō), and the following year, Prince Atsunaga (Emperor Go-Suzaku) was born. From Emperor Ichijō’s death in 1011 until her own death at the age of eighty-seven in 1074, she oversaw, for a span of more than sixty years, the rule of her sons Go-Ichijō (r. 1016 – 1036) and Emperor GoSuzaku (r. 1036 – 1045), her grandsons Emperor Go-Reizei (r. 1045 – 1068) and Emperor Go-Sanjō (r. 1068 – 1072), and the first years of the reign of her greatgrandson Emperor Shirakawa (r. 1072 – 1086). During this period, her younger brother Fujiwara no Yorimichi (992 – 1074) acted as regent, and the path toward rule by retired emperors (insei), which came about after Shirakawa’s retirement in 1086, was initiated. Many studies have been devoted to this late period of the regency, but little attention has been paid to the important role of Fujiwara no Shōshi in this transition. For instance, the shift in balance in favor of the imperial family can be seen in Shōshi’s support of reigning emperors over her own brother Yorimichi. Below, I will point to a few characteristics of Shōshi’s authority as the head of the imperial house. First, Shōshi had great influence over court appointments. The following quotation from the diary of Fujiwara no Sukefusa speaks for itself: “The emperor, the regent [Yorimichi], and the principal imperial consort [Jōtōmon’in] fill positions with people with whom they have strong connections. Such behavior can only be a sign of an age of decline.”70 Moreover, I have shown elsewhere how records in the Kojidan, a thirteenth-century tale, report that Yorimichi tried to have his son Morozane succeed him as regent, but when he went to announce this to Shōshi, the lady, who was already in bed at the time of the visit, abruptly arose and stated that her father’s will prescribed that the office should be passed to Yorimichi’s younger brother Norimichi.71 Fujiwara no Sanesuke’s frequent visits to Shōshi, his adopted son Sanehira’s success in climbing the court hierarchy, and his use of the well-connected court lady and writer Murasaki Shikibu as his intermediary to send messages further reflect the private nature that a control of the imperial court entailed, as well as Shōshi’s authority over personnel appointments.72 Second, Shōshi had the right to determine consorts. As already noted, the selection of a consort was the first step in the continuation of the imperial line, and the ramifications of such decisions could be far-reaching, as all nobles in Kyoto knew. Fujiwara no Yoshinobu, Shōshi’s younger brother by a different mother, sought to have his adopted daughter become consort of Crown Prince Takahito, and he clearly felt compelled to gain the approval of Shōshi: “The other day, I received the blessing of the regent [Yorimichi], and when I went to the retired lady [Shōshi] in the same matter, I received her permission as well.”73 The experience must have been similar for other ranking nobles hoping to have their female relatives climb the ladder at the court in this fashion. However, from Emperor Go-Ichijō to Emperor Go-Sanjō, all consorts were descendants of Michinaga on either the maternal or the paternal side. Using her powerful position, Shōshi
32 | fukutō sanae with takeshi watanabe
forged enduring connections between the imperial house and Michinaga’s descendants. Her success came to set Michinaga’s lineage apart from the rest of the aristocratic class, especially when it came to matters of the imperial succession, which in turn enabled the imperial house to reassert its control over imperial succession.
Conclusion Until the latter half of the eighth century, the status of men and women at the court did not differ significantly. In the ninth century, the roles and physical spaces of men and women began to be segregated, and the political standing of women diminished from an official perspective, as women were pushed to the periphery of the political sphere. Yet to place all the emphasis on this trend would be to ignore the foundations of power in the regency age. The power of the kokumo allowed her male relatives to use the political space in the rear palace, where the senior consorts and empress dowagers lived, and to build political offices close to that area in order to represent or assist the emperor. Moreover, the strengthening of the parental authority of the imperial house, and the confirmation of Michinaga’s line as the house of the regency, were the results of Fujiwara no Shōshi’s dominance as nyoin, in which she used her authority to select consorts. While pushed to the periphery in public spaces, women could participate in politics at its core privately inside the imperial house. This participation was based on the way in which power evolved from the ninth century, when senior members of the imperial house began to exercise their authority, initially as parents of the rulers. In fact, it was because of this removal from the spotlight outside of the immediate center of the imperial court that leading women of the imperial house were able to continue to exert a profound influence on political issues. Still, it would not be until the end of the eleventh century, from the rule of Retired Emperor Shirakawa, that the imperial lineage could establish itself as one of the elite houses (kenmon). And it was through the power of the kokumo that this autonomy came to fruition.
Notes 1. For these views, see Kobayashi Toshio, Kodai jotei no jidai; and Araki Toshio, Kanōsei toshite no jotei. For an analysis in English of this topic, see Joan Piggott, The Emergence of Japanese Kingship. 2. Satō Nagato, “Kodai tennō sei no kōzō to sono tenkai,” 45. 3. The establishment of the title of emperor (tennō) has been the topic of much debate in recent years. The current consensus points to the court of Tenmu, but for convenience, I will use the title for earlier rulers as well. 4. See Inoue Mitsusada, “Nihon no jotei.” 5. See the aforementioned works by Kobayashi and Araki (n. 1). 6. Araki, Kanōsei toshite no jotei, 5.
From Female Sovereign to Mother of the Nation | 33 7. Yoshie Meishi, “Kodai jotei ron no kako to genzai,” 34. 8. Shoku Nihongi, Tenpyō 1 (729) 8/24. 9. See Haruna Hiroaki, “Kōtaihi Abe no kōjo ni tsuite: ryōsei chūgū no kenkyū.” 10. Kitō Kiyoaki, “Kōgōgūshiki ron.” 11. See Yoshikawa Shinji, “Ritsuryō kokka no jokan.” 12. See Haruna Hiroaki, “Naishi kō”; and Katsuura Noriko, “Kodai kyūtei soshiki to seibetsu bungyō.” 13. In the early modern period (kinsei, 1600 – 1868), political fighting between the bakufu and the court led to the enthronement of two female sovereigns. See Araki, Kanōsei toshite no jotei. 14. See Misaki Hiroko, “Kisaki no miya no sonzai keitai”; Hashimoto Yoshinori, Heiangū seiritsu shi no kenkyū; idem, “Kōkyū no seiritsu.” 15. See Hashimoto, Heiangū seiritsu shi no kenkyū. 16. See Kitō, “Kōgōgūshiki ron.” 17. Nishino Yukiko, “Bokō to kōgō.” 18. See Okamura Sachiko, “Tennō shinsai saishi to kōgō.” 19. Nihon kōki, Kōnin 6 (815) 7/13. 20. See Hashimoto, “Kōkyū no seiritsu”; and Yoshikawa, “Ritsuryō kokka no nyokan.” 21. Kuribayashi Shigeru, “Kōgō juga girei no seiritsu to tenkai”; idem, “Heian chō ni okeru sangō girei ni tsuite.” 22. See Hashimoto, “Kōkyū no seiritsu.” 23. Ibid. 24. Okamura Sachiko, “Onna joi ni kansuru kisoteki kōsatsu.” 25. Translator’s note: The establishment of male secretaries may also be related to the role played by Fujiwara no Kusuko, Heizei’s favorite consort, in the insurgence against Emperor Saga. 26. Haruna, “Naishi kō.” 27. Shoku Nihon kōki, Jōwa 9 (842) 7/23. 28. Nihon sandai jitsuroku, Gangyō 3 (877) 3/22. 29. Nishino, “Kyū seiki no tennō to bokō.” 30. Shoku Nihon kōki, Jōwa 9/8/4. 31. Nihon Montoku tennō jitsuroku, Saikō 1 (854) 4/26. 32. Nihon sandai jitsuroku, Gangyō 3/3/23. 33. See Haruna, “Naishi kō.” 34. Junshi is referred to as imperial mother even though she had already become grand imperial dowager on Saikō 1 (854) 4/26. The mother of Seiwa, Fujiwara no Meishi became imperial mother on Ten’an 2 (858) 11/7. Therefore, there was no kōtaibunin on Ten’an 2/8/29, and one needs to look elsewhere to deduce the identity of this figure. In an entry for the following year, Jōgan 1 (859) 4/18, there is a reference to the kōtaigō riding in the same palanquin as the emperor on the twenty-ninth day of the eighth month the previous year, making it clear that Junshi was the grand imperial dowager. 35. Nihon sandai jitsuroku, Ten’an 2/8/29. 36. See Nishino, “Kyū seiki no tennō to bokō.” 37. See Suematsu Hiyoshi, “Sokui shiki ni okeru bokō no tōdan.” 38. Nihon sandai jitsuroku, Ten’an 2/11/7. 39. Nihon sandai jitsuroku, Jōgan 7 (865) 11/4. On 8/21, however, Emperor Seiwa temporarily moved from the Eastern Palace to one of the Grand Council’s official halls (daijōkan no zōshi) to avoid a directional taboo (hake monoimi).
34 | fukutō sanae with takeshi watanabe 40. Nihon sandai jitsuroku, Gangyō 3 (879) 3/23; Nishino, “Kyū seiki no tennō to bokō.” 41. This was common practice in the Heian age, and as long as the emperor did not accept the rejection, the appointment would be remade, as it was in Mototsune’s case. 42. Nihon sandai jitsuroku, Gangyō 5 (881) 2/21. 43. See Hashimoto, “Kōkyū no seiritsu”; Okamura Sachiko, “Shiki no mizōshi ni tsuite.” 44. I have treated this topic in more detail in my “Ōken to kokumo: Ōchō kokka no seiji to sei.” 45. See Tsunoda Bun’ei, “Taikōtaigō Fujiwara no Onshi”; idem, Heian jinbutsu shi. By contrast, Kōchi Shōsuke does not address Princess Hanshi’s role in his Kōdai seiji shi ni okeru tennō sei no ronri. 46. Kyūreki, Tenryaku 4 (950) 6/15. 47. Senso burui shō, Emperor Daigo, Kanpyō 9 (897) 7/3. 48. Nihon kiryaku, Kanpyō 9/8/9; Shōtai 1 (898) 2/13. 49. Nihon kiryaku, Kanpyō 9/7/26. 50. Nihon kiryaku, Shōtai 1/4/25. 51. Translator’s note: Untimely deaths, epidemics, and other natural events caused many at the imperial court to believe that the scorned spirit of Sugawara no Michizane, who died in exile in 903 during the reign of Emperor Daigo, had caused these calamities. For an extensive account of these events, see Robert Borgen, Sugawara no Michizane and the Early Heian Court. 52. This point is discussed in greater detail in my article “Ōken to kokumo.” 53. Satō Nagato, “Kodai tennō sei no kōzō to sono tenkai.” 54. Kyūreki, Tenryaku 4 (950) 5/26. 55. A prince born from a nyōgo took precedence over one from a kōi (female attendant). For more on this topic, see Yamamoto Kazuya, “Nihon kodai no kōgō to kisaki no joretsu.” 56. Kyūreki, Tenryaku 4 (950) 5/26. Morosuke’s deduction regarding Norihira’s selection as crown prince from a comparison of gifts is suggested by Kojima Naoko, “Monogatari no gishiki to ‘hi’ junkyo.” 57. Kyūreki, Tenryaku 4/6/15, 25, 27. 58. Satō Nagato, “Kodai tennō sei no kōzō to sono tenkai,” 45. 59. The directive issued by the Council of State on the tenth day of the eighth month shows that Princess Shōshi was born in Tenryaku 4, but the date remains unclear. However, imperial consort Kiko died on the fifth day of the fifth month, and it is believed her death may have been caused by childbirth. See Dai Nihon shiryō, 1:9, entries for Tenryaku 4/5/5 and 8/10. 60. See Fujiki Kunihiko, “Fujiwara no Onshi to sono jidai.” 61. See Tsunoda, “Taikōtaigō Fujiwara no Onshi.” 62. See the works by Tsunoda and Fujiki (nn. 60 and 61). 63. Shōyūki, Chōtoku 3 (997) 7/5. 64. Shōyūki, Eiso 1 (989) 8/23. 65. Shōyūki, Eiso 1/8/23; Shōryaku 1 (990) 8/29. 66. Gonki, Chōhō 1 (999) 12/7, Chōhō 2 (1000) 1/28, 12/7. 67. Tomita Setsuko, “Heian jidai chūki ni okeru rikkō jijō to gaiseki kankei — toku ni Michinaga no baai wo chūshin toshite.” 68. Takamatsu Momoka, “Nyoin no seiritsu — sono yōin to chii wo megutte.” 69. Ryō Susumu, “Nyoin sei no seiritsu”; Hashimoto Yoshihiko, “Nyoin no igi to enkaku.” 70. Shunki, Chōkyū 1 (1040) 6/8. 71. Fukutō, “Ōken to Kokumo”; Maki Michio, Insei jidai shi ronshū. 72. For example, see Shōyūki, Chōwa 2 (1013) 1/19. 73. Shunki, Eishō 1 (1046) 10/22.
3
d Joan R. Piggott
Court and Provinces under Regent Fujiwara no Tadahira Let Minister of the Left Lord Fujiwara no Tadahira protect and help the young lord....And while [the tennō] is not yet familiar with the plenipotentiary powers of rulership, let him [the Minister of the Left] protect and aid the sacred person and take charge of governmental affairs.
R
eceiving such royal orders from the dying Daigo Tennō (r. 897 – 930), Fujiwara no Tadahira (880 – 949) became regent (sesshō) for the sevenyear-old Suzaku Tennō (r. 930 – 946) in 930.1 Later, Tadahira would also serve as chief of staff (kanpaku) to Suzaku and as regent to Suzaku’s younger brother, Murakami Tennō (r. 946 – 967). Tadahira was not the first scion of the Northern Fujiwara family to serve as regent; he was the fourth son of Fujiwara no Mototsune (836 – 891), the adopted heir of Fujiwara no Yoshifusa (804 – 872), and both Yoshifusa and Mototsune had served as regent before Tadahira. Nevertheless, Tadahira’s tenures as sesshō and kanpaku saw the regularization of many of the precedents and protocols that came to structure these offices of court leadership, making Tadahira’s a particularly important historical moment. In this volume dedicated to better understanding the relations between center and periphery up to the late eleventh century, it is important to consider how Regent Tadahira presided over provincial administration. Well documented as his activities are in various types of records including entries from his own journal, the Teishinkō ki, it is time to revise the old view of Tadahira as an urban and aristocratic leader too alienated from and uninterested in the provinces to involve himself in their governance. The record demonstrates instead that Tadahira had long training in provincial administration before 930, which experience served him well later as court leader. But the same record shows that the court of Tadahira’s day suffered from a lack of consensus on successful provincial policy as violence in both capital and countryside was increasing. My argument here is that a major cause of contention was the increasing contradiction between two systems of management and provisioning — one more bureaucratic and the other more patrimonial — that linked rulers of the tenth-century capital with the periphery of their realm. We shall also see that court leaders such as Tadahira were deeply invested in both systems, making resolution of the contradictions difficult.
36 | joan r. piggott
The Training of a Future Regent, 895 – 930 Long before he became regent, Tadahira’s early postings as a royal chamberlain (jijū), senior controller (daiben), and then as middle councilor (chūnagon) and minister of the right (udaijin) on the Council of State (Daijōkan) provided him with broad training in court operations, including issues of provincial administration. Tracing Tadahira’s early career through the following narrative and accompanying table (table 3.1) provides substantial insights into how the court actually functioned in the early tenth century, both in terms of how it governed the provinces and how it extracted resources from them. After his coming-of-age ceremony at the age of sixteen in 895, Tadahira was promoted to the senior fifth rank lower step by the sitting monarch, Daigo Tennō. He was also given his initial posting in the court bureaucracy as a chamberlain in the Royal Affairs Ministry (Nakatsukasashō). The following year Tadahira was named provisional senior governor (gon no kami) of Bingo Province, from which posting he gained a share of provincial taxes as livelihood, although likely he never left the capital. At that point Tadahira doubtless became more aware of the challenges facing tax collection, because they affected him personally. The tenth-century polity — some researchers have called it the “court-centered polity” (ōchō kokka) in contrast to the “tennō-centered polity” of the eighth and early ninth centuries — had two separate systems for drawing provincial resources into the capital.2 On the bureaucratic side, the monarch, whose court and provincial administration was originally structured by the Chinese-style ritsuryō codes, had come to depend on a regent or chief of staff, who on the monarch’s behalf oversaw policy deliberations by the senior nobles (kugyō) of the Council of State. In turn, the senior nobles supervised provincial administration by custodial governors (zuryō) who acted as the monarch’s emissaries and tax collectors in the countryside.3 In many respects this system perpetuated the integrative hierarchies — officialdom, a network of central places, and a network of ritual centers — that had given shape to the original tennō-centered realm.4 But with the failure by the early ninth century of ritsuryō-mandated processes of cultivator registration, field distribution, and taxes based thereon, those who had lived off such arrangements — nobles, official religious institutions, and even lesser officials with the means to do so — began investing in the opening of rental fields to provide themselves with an alternative source of livelihood. One result was the proliferation of absentee interests across the Kinai and beyond, accompanied by increasing bonds of clienthood between powerful patrons in the capital and wealthy local elites (known as fugō no tomogara in contemporary sources). The latter signed up as housemen (kenin, keishi), attendants (toneri), or staff (zōnin) of capital patrons to gain tax exemptions and other favors.5 Meanwhile, sons and grandsons of governors who were taking up residence in the provinces were frequently accused of interfering with official tax collection. Soon provincials were turning to the households of their noble patrons rather than to provincial governors for an array of adjudicative and other services.6 Directives
Table 3.1. Fujiwara no Tadahira’s career Rank/post/privilege
Month/year
Monarch/comment
Senior first Chief of staff Chief of staff Equal of the three queens Prime minister Junior first Regent Court liaison for Enryakuji Crown prince preceptor Senior second, minister of the left Junior second Minister of the right (Council of Nobles) Commander of the left (Inner Palace Guards) Senior third Senior counselor (Council of Nobles) Middle counselor (Council of Nobles) Commander of the right (Inner Palace Guards) Director, Royal Secretariat Fujiwara clan leader Provisional middle counselor (Council of Nobles) Junior third Director, Royal Police Office Director, Left Gate Guards Director, Crown Prince’s Household Advisor (Council of Nobles) Provisional provincial governor, Bizen Junior fourth, upper Senior right controller Advisor (Council of Nobles) Junior fourth lower, chamberlain Provisional provincial governor, Bingo Chamberlain Permit to enter royal presence Senior fifth lower
08/949 05/946 11/941 02/939 08/936 11/932 09/930 12/929 10/925 01/924 02/916 08/914
d. Tadahira 08/14 Murakami Tennō 946 04/–>
Suzaku Tennō 930 09/–>
04/913 01/913 01/911 01/910 09/909 05/909 04/909 04/909 04/909 09/908 08/908 02/908 01/908 01/905 01/903 05/900 01–02/900 11/898 02/897 01/896 09/895 08/895
d. Fujiwara no Tokihira
marriage to Uda’s daughter Daigo Tennō 897 07/–>
38 | joan r. piggott
(kanpu) sent by the senior nobles to prohibit such practices from the mid-ninth century onward indicate the serious nature of the challenge to governors. Three examples follow. In recent years various noble households (ōshin no shoke) produce their seals and claim to be owed goods. They compete at the residences of district chiefs (gunji) and wealthy locals (fugō), where they amass a fortune in rice. When there is a quarrel with the governor [over such goods] they manufacture some excuse. This degrades not only the governor’s authority but also that of the court. (Jōwa 12 [845] 6/23)7 Whether locals or wanderers, all call themselves housemen of royals or ministers (ōshin). They fear neither the authority of the provincial governor nor that of the district chieftain. They become ever more aggressive, even resorting to violence. Moreover, they fraudulently claim that they are exempt from tax levies. (Jōgan 2 [860] 9/20)8 Residents of this [Mino] province are skilled only at collusion, resulting in loss of tax goods. When the provincial governor confiscates private goods [to make up for delinquent tax goods] and tries to transport them to an official storehouse, they immediately claim that [such goods] belong to a noble household. They request that a missive (chō) be sent to the provincial office stating that the goods [in question] are profits from private lending or repayment for loans. Sometimes [the noble household] displays a banner or erects a wooden marker. There is no crime greater than this! (Kanpyō 7 [895] 9/27)9 At the same time, staffs of the great noble houses in the capital — those serving the retired monarch, queens, crown princes, and senior nobles — continued to expand in size, in part because by the turn of the tenth century only senior nobles and those closest to the throne — such as royal secretaries (kurōdo), controllers, and royal intimates (tenjōbito) — were receiving regular salaries from government coffers. Even gifts (roku) that had once been given out at royal banquets as livelihood for officials declined because of shortages in tax goods. Lower rankers in the capital who had once received salaries could thus but turn to the great households for their sustenance. The senior nobles of the Council of State began issuing directives like those above to prohibit alliances between locals and capital patrons as early as 824, but persistent interdictions thereafter indicate that such efforts had little effect. In 896 the Council was still inveighing against noble households quarreling “over fields, houses, and other goods when presented with complaints by cultivators,” and in 905 the households were forbidden to send out agents to investigate crimes or attacks on locals.10 On the other hand, provincial governors were petitioning the court to authorize the appointment of police agents (kebiishi) in the provinces to halt crime and resistance to taxes. Thus the conflict between two different systems linking capital and provinces — one more bureaucratic and focused on the provincial governor, versus one more patrimonial and employing great house-
Court and Provinces under Fujiwara no Tadahira | 39
holds as actors — was surging when Tadahira was beginning his official career at the court of Daigo Tennō.11 As the grandson and son of two previous regents, the young Tadahira naturally enjoyed great advantages at court, but he also enjoyed the support of a powerful living patron as well. In 900 the retired Uda Tennō (r. 887 – 897) chose Tadahira to marry his daughter, Princess Nobuko, and become his son-in-law (muko).12 Likely it was also Uda’s influence that secured Tadahira’s appointment as right senior controller (udaiben) with the privileges of a royal intimate.13 Tadahira could thus function in the important role of controller-cum-royal-intimate (tenjō no ben), whose responsibility it was to be a liaison between the Controllers’ Office, the Council of State, the throne, and the retired monarch. At the time the highest-ranking minister on the Council was Tadahira’s eldest brother, Fujiwara no Tokihira (871 – 909), who was then serving as minister of the left. Controllers coordinated key elements of the tennō’s government, which meant that Tadahira’s posting as senior right controller put him at the heart of court operations. Extant entries from Tadahira’s journal begin in 907 but are sketchy for the early years. Nonetheless we know from the tenth-century Collected Commentaries on Administrative Law (Ryō no shūge) that as right senior controller Tadahira would have overseen the ministries of Defense, Justice, and the Royal Household for his bosses, the senior nobles of the Council of State.14 Senior controllers worked closely with middle controllers (chūben), junior controllers (shōben), and controller-secretaries (shi), receiving incoming petitions and reports from ministries and provinces and sending back directives promulgating policy decisions made by the senior nobles and the throne.15 Senior controllers also joined Council secretaries (geki) in preparatory sessions called katanashi to organize documents for Council discussions. When further information was required, controllers summoned knowledgeable officials for questioning. They determined where to make their report to the Council’s noble in charge (shōkei), who would then preside over Council deliberations either in the Council Secretariat (Gekichō), the Chamberlains’ Room (Jijūsho), or in one of the guard-post chambers (jin) near the royal throne room of the residential palace (map 2.2).16 After such deliberations, a memorial reporting Council members’ views (called the jin no mōshibumi) was submitted to the tennō, who then responded with his command. Were a Council directive subsequently sent out to promulgate that decision — either to official agencies or the provinces — a senior controller would also sign the repository copy (an).17 Finally, controllers also served as directors (bettō) of various offices in the residential palace, overseeing its smooth functioning in concert with members of the extracodal Royal Secretariat (Kurōdo dokoro). This description of how a senior controller worked closely with officials both in the capital and at provincial headquarters makes it clear that Tadahira would have become well acquainted with all the reasons provincial governors typically gave for their failures to pay the requisite taxes: drought, poverty, famine, plague, banditry, and interference by powerful court patrons. Tadahira would have helped draft, promulgate, and execute the directives known as the Engi
40 | joan r. piggott
reforms in the third month of 902. They called for enforcing traditional ritsuryō procedures such as requiring field registration and redistribution, higher standards for the quality of goods remitted as taxes, stricter evaluation of provincial governors’ performance, the prohibition of royal grant fields (chokushiden) that drew cultivators away from public lands, and other strictures against aristocratic land opening and cliental recruitment in the countryside.18 The year 908 was an important one in Tadahira’s career because it saw him named an advisor (sangi) on the Council of State, with concurrent appointments as commander of the Right Palace Guards and director of both the Office of Royal Police (Kebiishichō) and the Crown Prince’s Household. There can be no doubt that the triumvirate then leading the court — the retired Uda, the sitting monarch Daigo, and Minister of the Left Tokihira (Tadahira’s elder brother) — looked to Tadahira to help them solve a host of critical problems. One such was a dearth of qualified candidates for governorships: the office responsible for evaluating governors’ performances, the office of Discharge Examiners (Kageyushi), was extremely behind in its workload, leaving few candidates for reappointment on its approved list.19 Another challenge was famine: in 909, royal banquets had to be canceled, and emergency stores of rice in the capital were sold to relieve inflationary food costs and hunger resulting from widespread wasted fields.20 Adding to these difficulties, Tokihira died suddenly in the fourth month of 909. His passing left Daigo Tennō and his queen consort, Yasuko — a full sister to both Tokihira and Tadahira — more reliant than ever on Tadahira. The latter was quickly named provisional middle counselor (gon no chūnagon). In that post he could finally oversee particular court projects or events as noble in charge.21 At the same time, Tadahira was named right commander (utaishō) of the Inner Palace Guards (Konoefu) and director of the Royal Secretariat, assignments that essentially made him responsible for the operation and security of the tennō’s residential palace. Over the next few years Tadahira continued to rise rapidly in the Council and in the palace: in 911 he was named senior counselor (dainagon); in 913 he became left commander of the Inner Guards; and in 914 he was named minister of the right. The latter posting made him, in the absence of a higher-ranking minister of the left, leader of the Council (ichi no kami). But then Tadahira’s dynamic rise stalled: it took another decade before Tadahira was finally promoted to minister of the left (sadaijin) in 924. Most scholars think that Tadahira’s unusually long term as minister of the right resulted from Daigo Tennō’s willful determination to lead his own court.22 During these years Tadahira nonetheless remained busy taking matters of provincial administration before the throne. Tax shortages and the dearth of qualified appointees for governorships remained serious problems.23 In 910, for instance, Middle Counselor Tadahira noted in his journal that the directorcontroller (bettō no ben) who oversaw Council of State stores (mikuriya) had reported a shortfall of 4,132 koku of rice due from the provinces (enough rice to feed 4,132 people for a year).24 Such rice, normally collected as rent from public fields in every province, was to be used to support the activities of the Council, and replacement rice had to be located. On the advice of a member of the controllers’
Court and Provinces under Fujiwara no Tadahira | 41
staff, Tadahira and his colleagues decided that eleven provinces should send in special shipments of polished rice (shōmai) in lieu of other tax goods to cover this shortage.25 And to avoid such occurrences in future, Tadahira proposed establishing a new system of receipts by which tax payments were to be verified at the end of a governor’s term. Meanwhile, with regard to fallow fields, Tadahira memorialized the throne in 913 to report the dire news that provincial governors had reported extensive wasted fields in forty-four of sixty-six provinces. In response, he and his fellow senior nobles recommended that the Bureau of Taxation (Shuzeiryō) should decide annual limits for fallow fields in every province and that thereafter governors whose reports indicated excessive wasted fields would not be recommended for reappointment. This policy gained the tennō’s support and became law.26 Provincial administration continued in such disarray that Daigo Tennō unusually called for reform proposals directly from provincial governors in 914. A twelve-point memorial authored by the scholar and former governor Miyoshi Kiyoyuki (847 – 918) provides insight into an experienced governor’s views of current problems. Specifically, he urged Tadahira and his colleagues on the Council to assure further economies in administering the capital, larger and better-paid provincial staffs, and more attention to provincial matters.27 The fact that Miyoshi was eventually appointed an advisor on the Council suggests that Tadahira and his fellow senior nobles valued Miyoshi’s views. Tadahira was then minister of the right, and his name appears in connection with numerous directives promulgating pragmatic and innovative strategies. For instance, he ordered an unprecedented 5,400 koku of rice collected from official rental fields to pay official salaries in that same year.28 Cornelius Kiley has also argued that it was under Tadahira’s leadership about 925 when the Council of State proposed to the throne that innovative land-based taxes be levied on socalled named units (myō). Such a policy meant that wealthy members of rural society were being given increasing responsibility for overseeing the extraction of tax payments from their less-wealthy neighbors.29 Meanwhile, in the residential palace the Royal Secretariat over which Tadahira then presided was devising a new system of daily tributes (hitsugi nie) meant to supply the royal table with fowl from six provinces near the capital. These new arrangements emphasizing both geography and special benefits for royal provisioners were intended to free the palace from the tax shortages faced by other units of government. Notably here the bureaucratic system of taxation was thereby replaced with a more patrimonial one: tax payers became provisioners who looked to the palace as patron. Moreover, in another similar instance Tadahira went so far as to solicit an order from the throne that all tribute goods from the sea to be used for royal banquets known as sechie should be shipped to the Queen Consort’s Agency (Chūgūshiki) rather than the Kitchen Bureau (Ōiryō), as had been the rule in the past.30 Such a change would have made provisioners of such banquets into clients of the Agency while strengthening the role of Tadahira’s own sister staff in the residential palace. In this regard Amino Yoshihiko has argued that from Tadahira’s era onward, the Northern Fujiwara Regents House (sekkanke) became increasingly involved in
42 | joan r. piggott
marine transport via the Inland Sea as a result of its oversight of both the tennō’s Royal Secretariat and the Queen Consort’s Agency.31 Governors who abused their authority or ignored responsibilities presented challenges to which Tadahira, the Council, and the throne had to respond. In 914, for example, Senior Counselor Tadahira solicited a royal order prohibiting cruel treatment of post station workers by governors coming from and going to their provinces.32 And in 918 Tadahira again went before the throne arguing that no governor should be reappointed without fulfilling requirements for discharge from his previous appointment, including the submission of all receipts to verify tax payments.33 During these years the Office of Discharge Agents was frequently ordered to determine whether a governor’s negligence of his duties was serious enough to justify criminal charges, and the office was ever behind in its work, which meant few candidates made its cleared-for-reappointment list.34 Besides Tadahira’s involvement in activities supervising and evaluating governors’ service, there were frequent reports of bandits and pirates to be dealt with. In 914, for instance, Tadahira’s kinsman, Fujiwara no Toshihito, was sent out against bandits attacking tax shipments in the east country. Toshihito is an early example of a martial governor. He had served previously in the Bandō provinces of Shimotsuke and Kazusa where he married into a local elite family, despite Council of State directives from the ninth century onward warning against governors settling outside the capital. In 915 Toshihito reportedly defeated “a thousand massed evildoers” as a barbarian-suppressing commander (chinjufu shōgun).35 Governors had good reasons for arming themselves: they were frequently targets of violence.36 Researcher Abe Takeshi has identified fifty-nine incidents of assassination between 795 and 1104.37 Competing interests of powerful noble households also led to conflicts in the provinces. One such instance was a quarrel in 913 over ownership of rice fields at Takaniwa Estate in Inaba Province claimed by the royal temple Tōdaiji of Nara, as well as by the households of a princess and a Council member. In this case we have extant documents that recite competing claims; but with royals, aristocrats, and a royal temple involved, neither the provincial authorities in Inaba nor the Council of State found the dispute easy to resolve.38 Unfortunately, the record does not reveal the outcome, but later records suggest that Tōdaiji succeeded in keeping its property.39 In the early tenth century, high-rankers at court carefully looked after the needs of favored Buddhist institutions: members of the Council and the Controllers’ staff served as their lay directors (zoku bettō, kengyō), and represented their interests at court. Their monastic leaders also enjoyed substantial influence vis-à-vis the court. Entries in Tadahira’s journal demonstrate how Tadahira, other members of the Council, and the monarch relied on controllers and controller-secretaries as investigators, reporters, policy proposers, and liaison personnel to help decide provincial policy. As we have already seen, controllers were in contact with provincial headquarters and governors and were therefore familiar with conditions in the sixty-six provinces. According to the Benkan bunin, a register of appointees to the Controllers’ Office, excepting fifth-ranked junior controllers, most
Court and Provinces under Fujiwara no Tadahira | 43
controllers held the fourth rank.40 The following sampling of their activities in 931 provides a picture of their comings and goings between the royal residence, the Council Secretariat, the guard-post chamber of the senior nobles, and Tadahira’s own residence. • Having heard a report on pirates by the right junior controller, Tadahira ordered the report memorialized to the throne.41 Probably the report had come to Tadahira’s attention from the Council’s noble in charge, who would have received it from the Controller’s Office, who would have received it from provincial authorities. Notably, even though Suzaku Tennō was a child, Tadahira ordered the report memorialized to him, indicating that the protocols of tennō-centered government were still being carefully observed. We do not know the effects of the memorial, but it is likely that a royal order for appointment of a pursuit agent was handed down, drafted in writing, and sent out. • Tadahira ordered the left middle controller to investigate why special levies of exchange (kōeki) silk remained unpaid by certain provinces, and he asked the same official to determine which provinces should be ordered to send in polished rice needed by the Kitchen Bureau to feed officials working inside the palace precincts.42 We do not know the official’s responses, but they would later have been discussed by the Council, memorialized to the throne, and promulgated after a royal order. • Tadahira ordered the left junior controller to investigate and report concerning a complaint from the Stable Bureau that Bizen Province had failed to pay it both horses and beans.43 Presumably the complaint reached Tadahira through a report sent from that bureau to the Council by way of the Controller’s Office. Were the claim deemed correct, the Bizen governor would have faced pressure to pay what was owed. • Tadahira ordered the left senior controller-secretary (sataishi) to study precedents and report whether it was permissible to order a province to repair and clean up various facilities by the beginning of the following year.44 • Tadahira commanded a controller to bring him reports from various provinces concerning exchange items that had been levied on an extraordinary basis.45 It seems likely that in this case, like that above, Tadahira was checking up to see who was paying what, in order to chasten malingerers sooner rather than later. All of these instances show not only how busy with affairs of provincial governance Tadahira was, but also how personally involved he was in overseeing governors’ tax payments. Indeed the early tenth century witnessed a grand effort to routinize processes and rituals of court government through compilation of the Procedures of the Engi Era (Engi shiki). Beginning about 912, Tadahira is known to have worked closely with Right Senior Controller Tachibana no Sumikiyo to compile this massive handbook.46 Daigo Tennō had begun the project in 905, but compilers were beset by sickness and death and so made slow progress. Entries in Tadahira’s
44 | joan r. piggott
journal indicate that he spent considerable time and energy on this assignment, which was finally completed only in 927. The Procedures mandated proper protocols for government not only in the capital but in the provinces as well. Ten out of fifty rolls concern the propitiation of deities and shrines distributed all over the realm; and another section devoted to the Council of State details how that body was to conduct itself when receiving requests for decisions from provincial headquarters. Other groups of rubrics laid out how the court and capital offices were to be provisioned with taxes collected, stored, manufactured, and transported from every province of the realm. There were procedures for recruiting guards and palace attendants in the provinces as well as prescriptions for evaluating the service of provincial governors and organizing the provincial military. By a rough estimate, twenty-nine out of fifty rolls of the Procedures concern aspects of provincial affairs.47 As a meticulous courtier who cared deeply about following proper procedures and precedents, Tadahira must have been particularly well suited to this work of compilation. Before turning to the second half of Tadahira’s career, when he served as court leader and preeminent advisor to the throne as both regent and chief of staff, it is important to consider the development and activities of Tadahira’s noble household in the provinces. Above we have focused on Tadahira’s official (bureaucratic) involvement with the provinces, through his posts in the tennō’s court. What follows is a discussion of the alternative patrimonial system by which resources from the countryside were also drawn into the capital through the authority of noble households. Like that of other high-rankers, Tadahira’s household received revenues from prebendal sustenance households called fuko.48 As minister of the right in 920, for example, Tadahira and his family and entourage would have been supported by two thousand households in various provinces. Fuko were assigned to their holder in units of 50 ko each, representing a township.49 According to the Engi shiki, the holder received all the tax proceeds from his prebendal households.50 Although provincial authorities were charged with delivering such proceeds to the beneficiary in the capital, some noble households worked out deals to receive the proceeds directly from the district chieftains who collected them. When this happened, the authority of the provincial office was compromised, which is why in 891 provincial governors requested a royal order forbidding the agents of prebendal beneficiaries from entering a province without the governor’s permission.51 We do not know in which provinces Tadahira held prebendal households, but records of the 930s indicate that at the time he held fuko at least in the eastern province of Shinano and the Inland Sea province of Sanuki.52 How was Tadahira’s household organized? There is not much evidence, but an extant missive initiated by the household in 920 was signed by fourteen housemen including a provincial governor and others serving concurrently in various government posts. Nishibeppu Motoka has argued that missives like this began to wield increasing influence in the mid-ninth century, as the activities of royal and aristocratic households in the countryside expanded. Indicative of the thick web of relations in which Tadahira’s household was involved, this missive was
Court and Provinces under Fujiwara no Tadahira | 45
addressed to the provincial governor of Tanba Province on behalf of an official temple in the capital, the “Eastern Temple” — Tōji. Apparently the temple had sent its own letter to Tadahira complaining that the Tanba governor was confiscating temple fields at Ōyama Estate, putting the temple and its monks in economic jeopardy. Following what we assume was Tadahira’s orders, his household staff urged the governor to treat the temple generously; supporting the Buddha’s work, they opined, was critically important.53 Why might the temple have complained to Tadahira rather than to the Council of State? Perhaps it was easier, faster, and more reliable. Or perhaps the temple knew that it was on uncertain legal ground; government policy concerning rental fields, which had not been dealt with in the original ritsuryō codes, was only gradually taking shape. In any event, what is notable is that the temple had a choice as to which of two parallel command systems linking court and province it would utilize to influence the Tanba governor’s views, and it chose the unofficial one. And yet we must also note that in 920 Tadahira was the preeminent member on the Council of State, which body oversaw appointment and evaluation of governors, so the Tanba governor would certainly have held the counsel of this ranking council minister in the highest regard. Since the services of provincial governors in the countryside were so critical to the urban nobility, it is not surprising that Tadahira reportedly had at least one provincial governor’s daughter as a consort. According to the genealogical compendium Sonpi bunmyaku, the governor in question, Fujiwara no Tsunekuni, served at one point in Musashi.54 It is probably no coincidence that Tadahira reportedly held rice fields in that province; they may well have been opened when Tadahira’s ancestors served as provincial governors there at the turn of the ninth century. Some of his prebendal households might have been there as well. Another province in the east country of special interest to Tadahira was Shimōsa, home place of the rebel Taira no Masakado (? – 940). We know from the Masakado ki (alt. Shōmonki) that Masakado, whose ancestors included Kanmu Tennō (r. 781 – 806) and the prince-governor Takamochi (? – ?) who went out to Shimōsa in the mid-ninth century, spent part of his youth in Tadahira’s household in the capital. Fukuda Toyohiko has suggested that Masakado’s family may have served as wardens of a royal horse ranch in Shimōsa.55 Masakado’s experience suggests that other sons of provincial notables would also have served in Tadahira’s household. Given that the Masakado ki claims that the young Masakado was personally acquainted with Tadahira, those provincials could have provided Tadahira with firsthand information about their home places.56 Some of Tadahira’s fourteen housemen who signed the missive in 920 would have spent time outside the capital, while agents coming and going from countryside estates would also have been additional informants about the world beyond the capital. As the highest-ranking courtier in the Fujiwara clan after Tokihira’s death, Tadahira served as Fujiwara clan head (chōja), which meant that his household oversaw clan operations like the clan temple Kōfukuji in Nara. That temple in turn possessed estates (shōen) like Kada no shō in Bizen, Katakami no shō in Echizen, and Kuzuha Horse Ranch (maki) in Kawachi.57 Indeed, Tadahira’s jour-
46 | joan r. piggott
nal records that in 927 he appointed the overseer at Katakami no shō. Agents, reports, and provisions from properties such as these would have arrived at Tadahira’s residence frequently, which meant that however urban and aristocratic Tadahira was, he and other members of his household staff were certainly neither ignorant of nor cut off from the provinces. This review of Tadahira’s involvement with the provinces during his preregency career makes it clear that as a controller and then as a Council member Tadahira was deeply involved in the bureaucratic processes of appointing, supervising, and evaluating the provincial governors who served as the monarch’s emissaries to the provinces. At the same time, as head of a noble household that supported itself by marshaling resources and personnel in the countryside through cliental relations, Tadahira was also deeply invested in a patrimonial system linking capital and countryside. In the latter role, he was inclined to frown on any governor who dared to interfere with his household agents or those of a favored religious institution such as Tōji or Tōdaiji. As we shall see below, contradictions implicit in this two-track involvement in the countryside by high-ranking courtiers such as Tadahira grew noticeably worse in the second half of Tadahira’s career, as fiscal conditions at court and violence in the realm worsened.
Tadahira as Court Leader, 930 – 949 The accession of his nephew, the eight-year-old Suzaku Tennō, led to Tadahira’s appointment as regent in 930. The charter of his regental office is partially preserved in the Abstracted Records of Japan (Nihon kiryaku), which reports, “The Heavenly Sovereign [Daigo Tennō] retired from the throne, and passed it to Crown Prince Hiroakira. A royal order instructed, ‘Minister of the Left Fujiwara no Tadahira shall protect and help the young lord, and take charge (setsugyō) of governmental affairs (matsurigoto).’ ”58 An additional fragment elsewhere provides us with another part of Daigo Tennō’s command: “While he [Suzaku Tennō] is not yet familiar with the plenary powers of rulership (banki), [you, Fujiwara no Tadahira, shall] protect and aid the sacred person and take charge of governmental affairs.”59 Key terms here include setsu, meaning to grasp or steer, and banki, denoting the plenary powers of the monarch. To courtiers in the early tenth century, both terms would have brought to mind accounts in the eighthcentury Nihon shoki that describe how the famous Prince Shōtoku (574 – 622) had been charged with “overseeing (sōsetsu) the plenary powers of the throne (banki)” during Suiko’s reign at the turn of the sixth century.60 And later on in the ninth century, the same term, setsugyō, was used in royal orders given to Tadahira’s ancestors. In 866 an order empowered Seiwa Tennō’s maternal grandfather, Fujiwara no Yoshifusa, as prime minister (daijōdaijin), “to oversee governance of all under Heaven” — Tenka no matsurigoto wo setsugyō seshimeyo — in the wake of the Ōtenmon Coup (866). This order has long been regarded the first articulation of Yoshifusa’s regental powers.61 The second order, dated 876, directed that Yoshifusa’s heir (and Tadahira’s father), Minister of the Right Fujiwara no Mototsune, “take up decision making within and without the palace, and serve day
Court and Provinces under Fujiwara no Tadahira | 47
and night without cease. As the monarch’s uncle, look after and protect him, and let the young lord [Yōzei Tennō)] depend on your advice. As long as he remains unfamiliar with royal plenary powers (banki), carry them out (setsugyō) and serve nearby as the faithful and just minister.”62 Thus the royal command appointing Tadahira regent in 930 purposefully recalled both the hoary and more recent pasts.63 Thereafter Tadahira would remain regent until 941 when, some years after Suzaku Tennō had had his coming-of-age ceremony, Tadahira was appointed Suzaku’s chief of staff.64 In his new post Tadahira’s service to the monarch was shaped by precedents set down by Fujiwara no Mototsune as kanpaku for Uda Tennō. Uda’s command to Mototsune in 887 had read as follows: Being of faint virtue, I take up the regalia. Gazing at the throne, it is as though I tread on the thinnest ice, touch the dragon’s scales, or cross a fathomless pool. But for the protection of the senior minister [Mototsune], how could I proclaim my royal will throughout the realm and make straight the way of heavenly governance in the palace? He has overseen governmental affairs (matsurigoto) for three reigns, always with a loyal heart. He makes clear the sacred words of my predecessor monarchs, and so do I look up to his advice. In my youth I am a foundling, but he wields all plenary powers and all officials obey him. So should he oversee (kanpaku) every matter, and afterwards let it be memorialized and [then] promulgated as in the past.65 An oral proclamation (senmyō) of the following year clarified Mototsune’s responsibility for inspection of memorials sent to the throne and for edicts sent out from the throne, a prerogative dubbed nairan, “royal inspection.” “From this time on, aid the throne by carrying out government and leading officialdom. Oversee all that is to be memorialized [upward] and ordered [downward]. Render advice as in the past, that I might follow it in my decision making.”66 As chief of staff with this power, Mototsune became the gateway to and from the monarch. Only remnants of the royal order issued to Tadahira in 941 are extant. One in the Nihon kiryaku commands, “Let the preeminent minister fully command all royal powers, great and small, and let him superintend all officialdom. Only afterward shall anything be memorialized or ordered, as in the Ninna era (885 – 889).”67 Whereas as regent Tadahira had represented and acted for a child monarch at the behest of a retiring elder monarch, as chief of staff Tadahira was to manage an adult monarch’s court following the mandate of the sitting monarch. The two posts of regent and chief of staff were alike in their court leadership function but different in the nature and source of their charters.68 In terms of provincial administration, as court leader — whether regent or chief of staff — Tadahira was involved in appointing and supervising provincial governors, coordinating realm-protecting ritual, and overseeing efforts to keep the peace. At the same time, Tadahira’s noble household expanded and became more powerful, thereby increasing the potential for conflict between his official responsibilities and the patrimonial authority and interests of that household.
48 | joan r. piggott
Appointing and Supervising Provincial Governors At the nexus of court government was the official appointment process, which was accomplished through the Appointments Ceremony (Jimoku). Two such ceremonies were held annually, and although most provincial governors were appointed at the beginning of the year, others were appointed in the summer. The Teishinkō ki and other contemporary records provide some insight into how appointees were chosen. At the end of the first month in 931, for instance, once a list of provinces needing new governors had been drawn up, Regent Tadahira received written evaluations of the past performances of two provincial governors.69 But because he was ill, the ceremony for official deliberations known as Jimoku no gi was postponed until the third month. At that time two members of the Council — a middle counselor, who probably represented the Council as noble in charge of the appointments process that year, and an advisor who was concurrently serving as right senior controller — presented their views concerning the administrative skills of two candidates during a discussion in which Regent Tadahira represented the young tennō.70 This discussion took place following nominating meetings attended by Council members — while he himself was ranking minister on the Council, as regent (and therefore the tennō’s stand-in), Tadahira did not usually join Council deliberations.71 These meetings to determine appointees that involved Regent Tadahira and Council representatives frequently continued over three or four days, after which the decisions were formally memorialized to the throne by the throne-room report (ōmagaki).72 At that time the child monarch Suzaku Tennō would have accepted the slate of appointments as presented, but additions and changes (naoshimono) could be made even after appointments had been announced. By 945, by which time Tadahira was serving as chief of staff for the adult Suzaku Tennō, the monarch himself was participating in these discussions of candidates with Council members.73 But in 946 the journal notes that a royal emissary nevertheless delivered a large sheaf of evaluations (kōka sadame kanmon) for governorship candidates to Tadahira, and that pattern of the monarch seeking his chief of staff’s counsel continued into the reign of Murakami Tennō (926 – 967).74 During this same period, we also see controllers continuing to play critical roles, coming and going from Tadahira’s presence whether at his home or in his quarters in the back palace of the royal residence, the shiki no sōshi.75 Evaluation of provincial governor candidates’ credentials was also becoming increasingly strict; specifically, there was more auditing of a governor’s accounts and demands for performance evaluations from offices verifying tax payments.76 In 939, for instance, Tadahira noted in his journal, The governor of Ise was promoted to senior fifth rank for [past] meritorious service in Higo. He should have been promoted at the Appointments Ceremony, but he did not obtain the comprehensive receipt for one year’s taxes in kind and labor and so was not promoted. Today, however, he was promoted because there were precedents for doing so and also because he submitted a complaint.77
Court and Provinces under Fujiwara no Tadahira | 49
In 931 Regent Tadahira and his colleagues on the Council had decided that only after a governor had set out for a new appointment should he receive a promotion in rank, presumably to encourage such embarkation. Here in 939 such a promotion was actualized.78 In Nara and early Heian times the tennō himself would have sent a governor off with exhortations to govern well, but Regent Tadahira took over such tasks for the child monarch. In the fourth month of 931, for example, Tadahira personally entertained two departing governors.79 And similarly in 939 he sent a stockade commander off to Dewa Province, on the distant eastern frontier, with sake and horses.80 Alternatively, when a governor failed to leave the capital in a timely fashion the culprit was obliged to write a suitably penitent missive, after which a legal expert would give his opinion to the Council as to whether punitive measures were warranted. One such case concerned the governor of Shimōsa in 949.81 The written statement of views that emerged from Council of State deliberations over such cases, known as jin no sadame, was scrutinized by Chief of Staff Tadahira in his capacity as nairan because ultimately it was the monarch who had to make a final decision based on advice from the senior nobles included in their jin no sadame statement.82 Besides appointing and exhorting provincial governors to govern well in the tennō’s name, as court leader Tadahira was deeply involved in evaluating their performances. As mentioned earlier, the Office of Discharge Examiners had been established to evaluate governors in early Heian times, and over the course of the ninth century the focus of its concern had become increasingly fiscal: its members were to certify that governors had met all their tax-paying obligations before being cleared for reappointment to another governorship. According to Cornelius Kiley and others who have studied these developments, by the mid-ninth century the head (kami) of a team of custodial governors was required to submit a discharge report (geyu) summarizing his full performance of tax-payment responsibilities up to the end of his tenure.83 This evaluation system for governors developed dynamically from the late ninth into the early tenth century, as the court sought ways to compel good provincial administration. Materials in the mid-Heian compilation, Abstract of Government Affairs (Seiji yōryaku), indicate that when Tadahira was regent, the examiners’ office was staffed by controllers and members of the Royal Secretariat, all of whom worked closely with Tadahira and some of whom had had actual experience as governors.84 For instance, in 933 the examiners indicted the Tanba governor with failing to keep Buddhist images, ritual objects, and provincial headquarters facilities in good repair, and with neglecting to carry out mandatory field distribution (handen). They likewise found that the governor of Awaji had failed to maintain official temples or even to report their ramshackle condition.85 They also indicted the governor of Echizen for failing to pay his tax obligations from the interest earned from the public-lending (suiko) program, as was required of governors who could not pay the full taxes owed from their province.86 A penitent admission of wrongdoing could sometimes get the governor pardoned for such offenses: in 932 Tadahira remanded such a case to a legal
50 | joan r. piggott
scholar for his opinion.87 Such increasingly detailed checks made the work of the examiners more costly, which was no doubt why in 936 governors filing the mandatory report known as the fuyogeyujō to explain their failure to complete their discharge paperwork were directed to personally provide the necessary paper, ink, and other supplies.88 The trajectory toward ever more lengthy discharge reports reached a high point in 949, the year of Chief of Staff Tadahira’s death, when governors were directed to submit statements from the Ministry of Popular Affairs, the Bureau of Taxes, and the Bureau of Statistics (Shukeiryō) as part of their discharge reports.89 The high incidence of fallow and ruined fields that deprived the court government of tax income was another cause of consternation to Tadahira, and he insisted on strict measures by provincial governors to control them. Every year provincial governors were required to survey and report the total number of such fields, on which basis the Council of State would deliberate and then submit the Fallow and Ruined Fields Memorial (Fukan denden sō) to the throne. Required annual tax payments for each province were figured based on that memorial; according to his journal, in 941 Chief of Staff Tadahira inspected and approved the memorial after it was brought to him at his home by the right senior controller.90 Later it was presented to Suzaku Tennō in his royal residence, the Seiryōden, with two counselors (nagon) and two advisors from the Council present.91 There were instances when governors unable to collect the requisite taxes — Kiley thinks obstruction by local elites was the major cause of the problem — shielded themselves by using the excuse of fallow fields.92 But Tadahira’s government was unwavering in its determination to put pressure on the governors, and fourteen of them were charged with the crime of overstating the number of fallow fields in 946. Similarly, in 948 the governors of Ise, Tajima, Bingo, and Iyo were all indicted for exceeding by more than 10 percent the limit on derelict fields set for their provinces.93 If stated policy were actually followed, none of these governors should have received new appointments as governors. As another defense against governors’ plaints of fallow and therefore untaxable fields, under Tadahira’s leadership the Council of State mandated that each governor should open an increasing number of fields every year, to the extent that by the end of his term more fields should have been opened than those opened by his predecessor.94 By midcentury Tadahira as court leader was having this rule vigorously enforced. All of this evidence indicates that Tadahira was a hands-on and tough boss of provincial governors. Entries in Tadahira’s diary after he became regent indicate that a very difficult challenge was securing adequate resources to pay for the costs of government. According to the eighth-century ritsuryō codes, proceeds from two types of poll (head) taxes — the handicrafts tax (chō) and the labor tax (yō) — were to be shipped to the capital to support government activities there. But the collection and quality of both sorts of resources dwindled during the ninth century, and by Tadahira’s day other means of provisioning had to be devised. One response was to declare extraordinary levies on specific provinces. In 931, for example, the Council ordered a special levy in coin and goods to pay for the construction of
Court and Provinces under Fujiwara no Tadahira | 51
Daigoji, a royal-vow temple to be built near the Heian capital. And in 938, when an earthquake destroyed the wall around the palace precincts, the Council decided that nine provinces — the five Kinai provinces plus Ōmi, Tanba, Harima, and Mino on its periphery — were to support the rebuilding effort. A Council directive was drafted, sent on to the chief of staff for viewing and then to the tennō for signing, and finally to the governors concerned.95 Provincial authorities did not always pay these levies gladly or quickly; for instance, goods to be sent in from Sagami, such as limestone and thick paper, were not fully paid in 934, occasioning a complaint from the Repairs Office (Shurishiki) through the controllers to the Council.96 And on another occasion, following an opinion rendered by an expert in the office of the Royal Police, Tadahira ordered the head of the Left Gate Guards to confiscate lumber that should have been sent in but had not been. In that case, enforcement agents were ordered to divide the take equally between the Repairs Office and the Carpentry Bureau (Mokuryō).97 Since the court continued to have difficulty getting provincial governors to pay taxes and special levies, the court experienced dire fiscal circumstances into the 940s. In the spring of 940, for example, a Council secretary advised Tadahira that Council storehouses (mikuriya) were too depleted to provide for the customary banquet in the fourth month (shun-no-za).98 Tadahira sent the secretary to the Chamberlain’s Office (Jijūsho) in the residential palace, which apparently had adequate provisions.99 This action would seem to confirm that the royal tribute system developed by Tadahira for the residential palace — which treated provisioners as clients rather than as taxpayers — was working more effectively than was the provincial tax system. In another instance, in 945 the Prelates’ Office complained that they could not perform the usual official Buddhist rites in the precincts of the Hall of State (Chōdōin) because it was in ruins.100 Rites for the Kamo princess-priestess (sai’in), which received careful attention during Tadahira’s era, were similarly affected by shortages.101 To provision the latter, in 946 planners decided to sell items from Council stores to purchase needed goods.102 By that point it seems that the problem was greater than a lack of tax stuffs: the capital itself was starved for resources. Conditions were so bad in 947 that senior nobles offered to return a share of the take from their prebendal units, and bandit attacks in the capital were common.103 These were very difficult times indeed for the court government over which Chief of Staff Tadahira was presiding. Keeping the Deities Happy Long before Chinese-style law was adopted in Japan at the turn of the eighth century, the predecessor of the tennō, the Yamato Great King, had established himself as the preeminent ritual coordinator for worship of both buddhas and deities (kami) in the realm.104 In Tadahira’s time too the court continued to support official temples and to send regular offerings called hōhei to provincial shrines, to pray for realm protection and prosperity. From the mid-eighth century onward the list of official shrines expanded: the Engi shiki lists 3,132 shrines to which yearly offerings were to be sent to celebrate the Annual Prayer Festival (Kinensai) in the second month.105 While special agents sometimes carried these offerings,
52 | joan r. piggott
the provincial governor frequently delivered them during his regular round of visits to shrines in his province. Such shrines also received promotions in court rank on occasion; deities, like local notables, appreciated receiving a share of the tennō’s symbolic capital. There were years during Tadahira’s tenure as court leader, however, when no dispatch of offerings to provincial shrines was recorded in the extant sources.106 Ninomiya Masahiko interprets this to mean that the official shrine system, having expanded so exuberantly up to the era of the Engi shiki, was losing currency. It may have been that Tadahira and his colleagues hoped to focus what had become an overly diffuse and costly system of propitiation by concentrating on key shrines, such as that of the royal ancestress’ Great Shrine of Ise, the Kamo Shrine, and the Usa Hachiman Shrine in Kyushu. Providing substantiation for that theory, Tadahira actually makes reference to “the fifty-four great shrines” in his journal.107 And yet in 944 the governor of Chikugo was ordered to prepare a detailed register of the deities in his province so that promotions in rank could be made.108 And even if there were years without the regular dispatch of shrine offerings, there were also years when extraordinary offerings were sent out. It would therefore be wrong to claim that Tadahira’s court was lackluster about maintaining its links with provincial shrines or that a policy of focusing on fewer regional shrines was being established. There were, however, deities that Tadahira’s court found it difficult to integrate into its system of realm protection. An entry from 945 in the Annals of Our Realm (Honchō seiki), compiled from a variety of sources in the later Heian age, reports, In recent days there are rumors in the capital that various deities from the west and east are approaching. Some claim it to be a deity called Shidara; others call it the Oigasa deity; and still others call it Hachiman. There is a request from the Settsu governor for the Council’s guidance.109 Accompanying reports indicate that the district chieftain of Teshima in Settsu had reported to provincial authorities that three sacred palanquins were on the move toward the capital.110 Unfortunately, there is no record in Tadahira’s journal about this event, but Prince Shigeakira, who was Tadahira’s son-in-law and intimate, notes in his journal that one of the deities was thought to be none other than Sugawara no Michizane, the angry ghost (goryō) of the minister exiled in 901 by Daigo Tennō and Fujiwara no Tokihira.111 Such a rumor excited fear at court because Sugawara’s spirit had been blamed for numerous deaths among the family and supporters of both Daigo and Tokihira, including the death of Daigo’s first crown prince in 932 as well as for a lightning strike on the royal residence that had killed a courtier in 930.112 Even more alarming to authorities, the mysterious palanquins were accompanied by hundreds of clerics and secular people, elites and commoners, who were beating drums and dancing madly. Over several days throngs assembled, and the sacred palanquins moved to the royally patronized Iwashimizu Hachiman Shrine on the Yodo River, whence monastic authorities sent additional reports to
Court and Provinces under Fujiwara no Tadahira | 53
the court.113 They noted that the three palanquins had increased to six, while the singing and dancing throng had grown to include tens of thousands. Who was this deity from the countryside that dared to approach the royal center without invitation? A few songs (wazauta) archived in the Honchō seiki tell of a deity that promised successful land opening after properly propitiation.114 Moreover, the deity’s association with Hachiman, who in turn was associated with both the exiled Sugawara no Michizane and the rebel Taira no Masakado, marked it as a power outside the court’s control.115 It is surely no coincidence that it was about this same time that the court over which Tadahira presided decided to establish the Kitano Tenman Shrine (Tenmangū) north of the palace, thereby taking steps to propitiate the wrathful spirit and incorporate it into the official program of realm protection.116 Making Peace in the Provinces Tadahira’s tenure as regent and chief of staff witnessed a dynamic increase in the posting of armed enforcement agents in the provinces to fight bandits, pirates, and rebels. Such agents were generally appointed in response to a request by a governor, and they were called variously envoys to pursue and destroy (tsuibushi), envoys to protect (keigoshi), and envoys to protect against crimes (kebiishi). In some instances these agents were sent out from the capital; in others, local appointees were nominated by provincial authorities for confirmation by the court. The problem of pirate activity in the Inland Sea was particularly acute during the 930s and 940s. On the twenty-eighth day of the fourth month in 932 Tadahira noted in his journal that he had transmitted to the left senior controller a royal order for the Council to appoint a tsuibushi to suppress pirates in the Inland Sea province of Bizen.117 There were additional requests for keigoshi in 934, when soldiers from Musashi were dispatched to Bizen together with private guards that served great households in the capital.118 Nevertheless the governor of Tosa Province, Ki no Tsurayuki, wrote in his Tosa Diary that he was chased by pirates on his return to the capital in 935, meaning that even the significant force sent in 934 had not pacified the pirates.119 Then again in 935 the Council ordered two secretaries replaced with kebiishi at the Dazaifu in Kyushu.120 In 936 there were also complaints that sons of former governors were obstructing tax collection there, resulting in a Council directive ordering enforcers to suppress the troublemakers.121 As court leader, Tadahira played a major role in developing strategies responding to both Taira no Masakado’s uprising in the Kantō and Fujiwara no Sumitomo’s piracy in the Inland Sea.122 Of the two, the court’s strategy in the east — relying largely on martial provincial elites to suppress Masakado — was notably more successful than the hesitant and tardy campaign in the west against Sumitomo.123 But the Masakado affair is dealt with in detail elsewhere in this volume and so will not be a focus here. Why did the strategy against Fujiwara no Sumitomo’s pirates not go as well as that against Masakado? Sumitomo had been named second in command at
54 | joan r. piggott
Iyo’s provincial headquarters in 932. He was subsequently deputed as tsuibushi in 937, under the command of a martial governor, Ki no Yoshihito.124 But by the twelfth month of 939, reports of Sumitomo’s own criminal acts began to appear in Tadahira’s journal.125 By that time both Masakado and Sumitomo were in open rebellion, and the Council appointed additional tsuibushi for the Tōkai, Tōsan, and San’yō circuits. Provincial governors were also ordered to reward any who would join the court’s side in the fighting.126 At the same time, however, even into the second month of 940, Tadahira and his colleagues on the Council were trying hard to avoid all-out war by bribing Sumitomo with a promotion. The court was desperate to reestablish law and order in the Inland Sea because official stores were so low that, as we saw above, even the most important royal banquets had been canceled, and it was understood that there would be no improvement in the situation as long as pirates blocked transport from the west. Meanwhile, the debate over how many forces should be conscripted and how many enforcers would be needed continued. In the midst of these troubles, pirates succeeded in setting fire to the port at Yamazaki quite near the capital.127 Nonetheless, even faced by that blatant challenge the court took another seven months to assemble fighting forces, and fourteen months passed before Sumitomo was finally killed in the seventh month of 941. By that time, a great deal of havoc had been wreaked by the pirates from the Kinai to the Dazaifu. The burning historical question is, Why were Tadahira and his colleagues so reticent to send an army west against Sumitomo? The best answer is surely the lack of resources possessed by the court and provinces with which to recruit and provision the needed army. Another reason was that the court had no recent experience mobilizing a large army, nor was there confidence that such an army would decisively deal with the pirates, who had been an ongoing if lower-level threat for years. It is also quite possible that Sumitomo, like Masakado, had links to Tadahira’s household, making the regent hesitant to send troops against him.128 Indeed, that the problem was bigger than Sumitomo was confirmed when, even after the latter’s death in battle, violence failed to abate, and requests for kebiishi appointments continued through 948.129 Doubtless the shortage of resources at the center made it difficult to reward enforcers’ good service adequately, such that those sent to put down pirates and bandits frequently turned against the very authority they once supported. Such was the case in 942, when an official sent out to Suruga as an enforcer soon turned to stealing tax goods in that province.130 One more challenge to peace in the provinces continued to come from feuding noble households and religious institutions. In the sixth month of 945, for example, two official temples in the old Nara capital were quarreling over repair work to be undertaken at Hōryūji, a third temple some distance away. The crux of the matter was likely a dispute over Hōryūji’s status at the time: was it a branch (matsuji) of the royal Tōdaiji or of the Fujiwara clan temple, Kōfukuji? In any event, Kōfukuji’s monastic authorities sent charges to the court against Tōdaiji, probably through its secular director, asking that a royal police agent be sent to investigate. Instead, Suzaku Tennō ordered Tadahira as his chief of staff to per-
Court and Provinces under Fujiwara no Tadahira | 55
sonally adjudicate the matter and advise him later concerning its resolution.131 Unfortunately we do not know Tadahira’s decision on the matter. Tadahira’s Noble Household, 930 – 949 For the period of Tadahira’s tenure as regent and chief of staff, the activities of his noble household — which would have been the most powerful household in the realm after that of the royal residence itself — are surprisingly little in evidence in the extant record. On one occasion, however, conflict broke out between Tadahira’s staff and a provincial governor. Specifically, in the late summer of 945 the Harima second in command, Fujiwara no Narikuni, complained to the Council that an envoy from Tadahira’s household had committed violent acts in Harima. A district chief there was involved, perhaps as a collector and transporter of prebendal dues. Given his post as preeminent minister (daijōdaijin) at the time, Tadahira would have been receiving proceeds from as many as three thousand households in various provinces annually.132 The chieftain’s report concerning the episode supported Tadahira’s emissary against Narikuni’s charges.133 Then, subsequent to Narikuni’s initial complaint, there were ongoing discussions between Narikuni and members of the Royal Secretariat, who apparently urged Narikuni to drop the matter. But he would not relent, and he soon learned the results of taking on the most powerful man at court. He was punished by the loss of both his privilege of entering the throne room (shōden) and his Harima governorship.134 We do not know specifically what happened in Harima in 945, but a list of sore points inflaming relations between noble households and governors at the time was enumerated in a Council directive sent to the provinces in the summer of 947. Therein governors were accused of setting up private repositories for tax goods near the capital, selling and exchanging goods to maximize their personal profits, and conspiring to pay as little as possible of taxes and other levies owed to the court and its leaders.135 In that same year, when reports of bandits in the capital were rife and rice was scarce, there were even rumors that the governors of Bitchū and Iyo were hoarding rice; these suspicions led Chief of Staff Tadahira to send royal police agents to investigate.136 Not only was the provincial tax system functioning poorly at this point. The conflict between official interests and the nonofficial interests of both noble households and provincial governors was causing tension as well. Meanwhile, since the tax system was no longer providing adequate income, government offices and courtier households increasingly turned to opening their own rental fields. As noted earlier, such fields expanded the patrimonial sector of the economy, in which relations between patrons in the capital and provincial clients took precedence over bureaucratic ties. I have found only one extant rec ord documenting such private holdings of Tadahira’s household. It is a district report (ge), dated 949, which claims that Yoshida name-holding (Yoshida myō) and Kiyomi township (Kiyomi hō) in Musashi’s Chichibu district, comprising privately opened fields amounting to 718 chō and 893 chō respectively, had belonged to Tadahira’s family since Fujiwara no Yoshifusa’s day in the ninth cen-
56 | joan r. piggott
tury. These properties were said to have become holdings of Tadahira’s household by the 930s.137 The initiator of the report was the Chichibu district chief, who would likely have been friendly with a noble household claiming 4,736 acres (1 chō = 2.94 acres) in his jurisdiction. Still another signer of the 949 report was its addressee, the sitting governor of Musashi, Minamoto no Mitsunaka (913 – 997). This Mitsunaka is famed as the martial governor who later supported Tadahira’s sons, Saneyori (900 – 970) and Morosuke (908 – 960), in the Anna Incident of 969.138 If this document from Chichibu district is legitimate, it would mean that Minamoto no Mitsunaka was already involved with Tadahira’s household before Tadahira’s death in 949, twenty years before the Anna Incident. On the other hand, it was not always the case that locals were friendly to noble households holding property in their midst, particularly when there were several such patrons whose interests, and whose clients’ interests, might clash or compete. An incident of this sort was recorded in 949, when some servants (ge’nin) of the then-retired Suzaku Tennō attacked houses belonging to attendants of the palace guards’ headquarters (shoeifu no toneri). Then two days later several hundred such attendants took revenge by trashing the dwelling of an overseer from the Royal Affairs Ministry (Nakatsukasashō) sent out by the retired monarch to manage one of his properties.139 The sources do not tell us the venue for these clashes, but it was likely somewhere in the Kinai provinces, where royal holdings and guards’ attendants were especially numerous.140 In this case, as in Inaba in 913, the interests of two capital patrons and those of their provincial clients clashed. Tadahira was on his deathbed at the time; but his sons, Minister of the Left Saneyori and Minister of the Right Morosuke, were then coleaders of the Council of State. They and their colleagues would have been called upon to resolve the dispute and restore law and order.
Conclusion As court leader Fujiwara no Tadahira was generally successful at keeping court and realm functioning despite numerous challenges, including the deaths of Daigo Tennō and the retired Uda Tennō, rule by the sickly child monarch Suzaku Tennō, climatic disasters, accompanying fiscal woes in the capital, and rebellion in the east and the west. Entries from Tadahira’s journal indicate that even as an adult, Suzaku Tennō lacked the inclination to lead the court himself, preferring to rely heavily on Tadahira as his chief of staff. Suzaku’s little brother Murakami Tennō followed the same pattern as long as Tadahira lived. As a result, Tadahira’s two decades as court leader certainly contributed to the institutionalization of the regent and chief of staff offices as key elements of the mid-Heian monarchy. And as we have seen here, that monarchy depended on a sort of consensus system in which members of the royal family with their staff in the palace, including the sesshō or kanpaku as their deputed court leader, worked with Council members and their staff — the controllers and secretaries — to govern capital and provinces.141 Satō Sōjun thus termed the court government of Tadahira’s day as “an aristocratic alliance” with Tadahira as its leader.142
Court and Provinces under Fujiwara no Tadahira | 57
We have also seen here, however, that Tadahira’s era saw growing conflicts both within the bureaucratic system of provincial management and between the bureaucratic and the patrimonial systems of provisioning the capital. By the late ninth century, it was clear which of these systems was most vulnerable: provincial governors were constantly pleading with the office of Discharge Examiners and the Council to accept their reports concerning wasted fields, proposals for lower tax payments, and requests for more armed agents to collect taxes and keep the peace. The same governors often complained about sons of former governors allied with local notables who were claiming tax exemptions because of special relations with patrons in the capital. Governors were frequently the targets of assassins, and they attracted the suspicion of senior nobles, who regarded many of them as venal and self-serving. While some have argued that provincial governors were growing ever stronger during the course of the tenth century, my conclusion here is that it was not yet the case during Tadahira’s lifetime. As court leader in the 930s and 940s, Tadahira surely recognized the inherent conflict between the bureaucratic and the patrimonial systems of drawing resources from the provinces. And he must have found himself torn about how to deal with those conflicts inasmuch as he, like his fellow nobles who presided at court, had vested interests in both systems. As regent and chief of staff for the monarch, his objective was to uphold tennō-centered government, including the efforts of governors to collect taxes from the households and cultivators of their provinces. He wanted the larders of the Council of State and the Kitchen Bureau full, the various shops and offices of the government well provisioned, and officials duly recompensed for their labors. At the same time, however, Tadahira was a noble whose household depended for much of its livelihood on far-flung interests in the provinces. He understood well that government offices like the guards’ headquarters had become dependent on rental fields and client cultivators, and he himself had established a cliental system of provisioners for the tennō’s residential palace under the Royal Secretariat. Nor could Tadahira be unsympathetic to the plaints of an official temple such as Tōji when it found its newly opened rental fields confiscated by an aggressive provincial governor. Tadahira was angered in 945 by what he saw as inappropriate accusations against his household agents by a provincial governor, the likes of which were often thought of as greedy and dishonest. Such conflicted views made it difficult for court leader Tadahira — and his fellow nobles — to gain leverage on the contradictions besetting provincial administration during the first half of the tenth century. And yet we also know that this did not remain the situation for long. By 960 governors were reportedly doing better managing their tax payments.143 And by 988, when the well-known Owari Petition complained about the illegalities and violent measures of the Owari governor, at least some provincial governors had obviously learned to organize their tax-collecting operations in new and effective ways while arming themselves with formidable entourages.144 Ōtsu Tōru has concluded as well that provincial governors successfully reorganized their authority by the late tenth and early eleventh centuries, including their cadas-
58 | joan r. piggott
tral surveying and tax-collecting authority within districts and townships.145 In the early eleventh century, when Fujiwara no Michinaga (966 – 1027) was able to secure full leadership of Council, court, and back palace, the most powerful governors served him as housemen.146 What therefore remains to explore are the changes that came about in the decades following Tadahira’s death, when his sons led the court. Increasing specialization — based on the idea that nobles at court should concentrate on ritual while provincial governors should concentrate on provincial administration, together with the attitude that those governors capable of getting the job done should be left alone to do so — seems one developmental trajectory worthy of future investigation.
Notes 1. The text of the epigraph is an amalgam of two distinct quotes from the royal order, one cited in the late Heian-period chronicle Nihon kiryaku and the other in a missive by Tadahira in the mid-Heian compendium of belles lettres, Honchō monzui. For both, see the convenient compilation of records on the regency compiled by the Royal Household (Kūnaichō): Kōshitsu seido shiryō Sesshō, 2 vols. Specifically, see vol. 1, 22; 8 – 9. 2. Japanese scholars who presently use the periodizing rubric “court-centered polity” include Morita Tei, Sakamoto Shōzō, Sasaki Muneo, and Hotate Michihisa. Others, including Yoshikawa Shinji, argue that since substantial ritsuryō structures remained in place, the rubric “later ritsuryō polity” (kōki ritsuryō kokka) is preferable. On the debate see Morita Tei, Ōchō kokka; Sasaki Muneo, Nihon ōcho kokka ron; and Yoshikawa Shinji, “Sekkan seiji no tensei.” 3. For a detailed account of this reconfiguration of ritsuryō court and monarchy, see Yoshikawa Shinji, “Sekkan seiji no tensei,” and “Ritsuryō kanjin sei no saihen.” 4. On the ritsuryō process and these integrative hierarchies see Joan Piggott, The Emergence of Japanese Kingship, 176 – 226. 5. Such patrons included agencies such as the six guards’ headquarters (rokuefu) and the Ministry of Defense (Hyōbushō). See Ichi Hiroki, “Kyū seiki Kinai chiiki no fugōsō to inkyū ōshinke, shoshi,” 31 – 47, 52 – 53. 6. For a useful discussion of this process see Maeda Sadahiko, “Kebiishi bettō to shichō.” 7. Ruijū sandai kyaku, vol. 2, in Shintei zōhō kokushi taikei (1983), pp. 603 – 604. For a discussion of the context see Ichi, “Kyū seiki,” 48. 8. Ruijū sandai kyaku, 638. For a discussion of the context see Nishibeppu Motoka, “Ōshinke chō no seiritsu no dōkō ni tsuite,” 19. 9. Ruijū sandai kyaku, 604 – 605. See Nishibeppu, “Ōshinke chō,” 13. 10. Ruijū sandai kyaku, 605 – 606, 618. See Nishiyama Ryōhei, “Heian-kyō to shūhen nōson.” 11. I use the term “patrimonial” following Weber. By “patrimonial relations,” Weber meant a decentralized relationship in which the patron’s household has dependents dwelling on extended landholdings, and they deliver rent in kind. In Weber’s parlance it represents an attenuation of patriarchal authority. Dependents are bound to the patron by relations of loyalty and fidelity, and they expect a degree of reciprocity in terms of protection and help in case of need. Patrimonial administration, says Weber, knows no clear separation between private and official spheres. The patrimonial ruler governs his territory by personal fiat and draws on wide discretionary powers. In contrast to a bureaucratic system of authority, there are no statutory
Court and Provinces under Fujiwara no Tadahira | 59 practices or formal procedures. Personal considerations and personal connections are critical. See Max Weber, Economy and Society, 212 – 301 and 1010 – 1011. 12. On Uda Jōō, see Mezaki Tokue, “Uda jōō no in to kokusei.” An account of this marriage appears in the early thirteenth-century Kojidan. Bridegroom-beckoning marriage (mukotorikon), by which a father chose his daughter’s groom and provided a home and resources for the new couple, was quite common among courtiers in the tenth century (see Sekiguchi Hiroko, Nihon kodai kon’in shi no kenkyū, 41 – 122). 13. In 900 Tadahira was briefly granted a seat on the Council of State as advisor but resigned it. Some researchers have hypothesized that the retired Uda may have been worried that as a Council member Tadahira was more likely to be implicated in Sugawara no Michizane’s quarrel with Tadahira’s brother, Tokihira. Sugawara was exiled to Kyushu in 902. 14. The Ryō no shūge is a compendium of legal commentaries on the ritsuryō codes compiled by Koremune no Naomoto around the turn of the tenth century. The oldest commentaries date from the eighth century while others were written in the Heian Period. 15. See Ōsumi Kiyoharu, “Benkan no henshitsu to ritsuryō daijōkan sei,” 23 – 24; and Hashimoto Yoshihiko, “Kanmu-ke Ōtsuki-shi no seiritsu to sono seikaku,” 4. To get a sense of the extant documents for Tadahira’s time, consult Dai Nihon shiryō (subsequently cited as DNS). 16. Matters of lesser import were resolved in informal deliberations (gekisei) in the Secretariat, while more serious issues were discussed in formal meetings of the full Council in a guard-post chamber. See Tamai Chikara, “Jū, jūichi seiki no Nihon, sekkan seiji,” 34; and Takemitsu Makoto, “Sekkan ki no daijōkan seiji no tokushitsu,” 3 – 4. According to the ritual handbook Saikyūki, compiled in the late tenth century, participants in informal deliberations in 913 included a middle counselor and three advisors who met in the Controllers’ Office. In 925 such a discussion included a middle counselor, an advisor, a controller, and a junior counselor (shōnagon). See DNS 1:5, 665, Enchō 3 (925) 3/25. 17. See Abe Takeshi, Heian zenki seiji shi no kenkyū, 57 – 58, Ōsumi, “Benkan no henshitsu.” 18. For more detail see Cornelius Kiley, “Provincial Administration and Land Tenure in Early Heian,” 291. The directives are archived in the Ruijū sandai kyaku. Satō Sōjun (“Fujiwara no Tadahira seiken no keisei”) and Morita Tei (“Fujiwara no Tadahira seiken no dōkō”) both paint Tokihira as a traditionalist and guardian of ritsuryō structures, given the character of these reforms. 19. Saikyūki, Jimoku section, Engi 9 (909) 1/11 in DNS 1:4, 4. 20. Fusō ryakki, Engi 9/1/21 in DNS 1:4, 5. 21. On the development of the noble in charge (shōkei) system, see Tsuchida Naoshige, “Shōkei ni tsuite.” 22. It was unusual for there to be no minister of the left (sadaijin). Kurosaka Nobuo points out that while Tokihira became minister of the left at the age of twenty-eight, Tadahira reached that post only at the age of forty-five (“Fujiwara no Tadahira seiken ni taisuru ichi kōsatsu”). Tsunoda Bun’ei (in “Fujiwara no Tadahira”) argues that Daigo Tennō and even Queen Consort Yasuko expected Tokihira’s heir, Yasutada, to become court leader as his father’s heir. They did not, therefore, want to appoint Tadahira minister of the left. Tadahira’s promotion in 924 came only months after his sister Yasuko gave birth to the son who would become crown prince, and it was meant to protect the interests of both his sister and new nephew. Here we have yet another instance of the throne’s resort to an “affinal strategy,” making affines into royal ministers. See Audrey Richards, “African Kings and Their Royal Relatives”; idem, “Keeping the King Divine”; and Piggott, Emergence, 76. Japanese scholars describe a “familial monarchy” (miuchi tennōsei), which seems similar. For instance, see Kurosaka, “Fujiwara no Tadahira”; Ihara Kesao, “Sekkan, insei to tennō”; and Kuramoto Kazuhiro, Sekkan seiji to ōchō kizoku. 23. Terauchi Hiroshi, “Zuryō kōka seido no seiritsu to tenkai,” 57 – 58; and his recent book,
60 | joan r. piggott Zuryō sei no kenkyū. See also Tamai, “Jū, jūichi seiki,” 47 – 48. Tamai notes that according to records archived in the late Heian-period Nihon kiryaku (Shōtai 2 [899] 7/11), only sixteen of sixty-six provinces had paid taxes in full the previous year. 24. Concerning the original ritsuryō tax system see Piggott, Emergence, 200. 25. Seiji yōryaku, a Council of State directive (kanpu) dated Engi 10 (910) 12/27, cited in DNS 1:4, 272 – 273. This Seiji yōryaku is a compendium of Heian governmental records and processes compiled by the legal scholar Koremune no Tadasuke during the first half of the eleventh century. 26. TKK 44, Engi 13 (913) 9/9; and Seiji yōryaku, a royal order (senji) dated Engi 13/8/13, cited in DNS 1:4, 528 – 529. 27. For a good analysis of Miyoshi’s “Iken jūnikajō,” see Tokoro Isao, Miyoshi Kiyoyuki, as well as Sato’s “Fujiwara no Tadahira seiken no keisei.” There is an English abstract of Miyoshi’s memorial in David Lu, Sources of Japanese History, 1:60 – 65. 28. Seiji yōryaku, Engi 14 (914), 308 – 312. As Terauchi Hiroshi has suggested, such large shipments of grain for the capital demonstrated increasing urbanization. See Terauchi, “Kyō shinmai to tojō.” 29. Kiley, “Provincial Administration,” 302 – 304. Kiley cites a Council directive dated Enchō 3/12/14 in Seiji yōryaku, vol. 2, in Shintei zōhō kokushi taikei (1983), 503. 30. Saikyūki, a Council of State order (kanpu) dated Engi 11 (911) 12/2, cited in DNS 1:4, 347. The provinces were Yamashiro, Yamato, Kawachi, Izumi, Settsu, and Ōmi. See also the order in Besshū fusenshō dated Enchō 1 (923) 9/9, cited in DNS 1:5, 499. The Besshū fusenshō is a compendium of 130 government orders issued between 902 and 971 and categorized by topic. Its compiler and date are uncertain. 31. Amino Yoshihiko, Nihon chūsei no hinōgyōmin to tennō, esp. 133 – 135 and 246. 32. DNS 1:4, 614, a Council of State directive (daijōkanpu) dating from Engi 14/6/13. See also Fujiki Kunihiko, “Fujiwara Onshi to sono jidai.” 33. Ruijū fusenshō, Engi 13/5/22 and Engi 18 (918) 10/21, cited in DNS 1:4, 512 – 513; and 1:5, 22. The Ruijū fusenshō (alt. Sajōshō) is a compendium of more than seven hundred orders (kanpu, senji, kansenji) dating from 737 to 1093 and emanating from the Council of State. They are categorized by topic and appear together with the requests (gejō) by lower-ranking officials that resulted in each order. The compilation predates 1121. 34. Seiji yōryaku, Engi 13/11/09, Engi 17 (917), and Engi 23, all cited in DNS 1:4, 557 – 558, 957 – 958; and DNS 1:5, 400. 35. On Toshihito, see DNS 1:4, 692. On prohibitions against provincial governors “going local,” see Karl Friday’s Hired Swords, 96. Other martial governors in the early tenth century included some Southern House (nanke) Fujiwara, on which see Noguchi Minoru, “Nanke Kuromaro ryū Fujiwara shi no Shimōsa ryūju to ‘hyōoke’ ka.” 36. In 915, for instance, a local elite murdered the second-in-command (suke) at the Kōzuke provincial headquarters. See Nihon kiryaku, Engi 15 (915) 2/10, in DNS 1:4, 717. 37. See Abe Takeshi, Heian zenki seiji shi no kenkyū, 329 – 332. 38. DNS 1:4, 541 – 545. Additional documents from Tōdaiji’s archives can be found in Dai Nihon komonjo, iewake, Tōdaiji monjo, vol. 3, doc. 820; and vol. 2, docs. 568 – 570. 39. In 940 – 941, Takaniwa Estate was still in temple hands. See Tōdaiji monjo, vol. 2, docs. 536, 571. 40. For the names and ranks of controllers in any given year, see the register of appointments, Benkan bunin. Its compilation began in the late tenth century. 41. TKK 134, Jōhei 1 (931) 1/21. 42. TKK 135, Jōhei 1/1/28. 43. TKK 137, Jōhei 1/2/17.
Court and Provinces under Fujiwara no Tadahira | 61 44. TKK 135 – 136, Jōhei 1/2/1. 45. TKK 137, Jōhei 1/2/14. 46. Engi shiki, cited in DNS 1:4, 423. 47. For an overview, see Torao Toshiya, Engi shiki. 48. For an English account concerning these official households, see G.Cameron Hurst III, “Structure of the Heian Court.” 49. Yoneda Yūsuke, Fujiwara sekkanke no tanjō, 153. 50. According to the first Engi shiki section devoted to the Ministry of Popular Affairs (Minbushō jō), a residence unit (ko) was defined as comprising four adult taxpayers and one young male (chūnan) between the ages of seventeen and twenty. Each township included fifty residence units totaling approximately two hundred adult taxpayers and fifty young males. 51. See Nishibeppu, “Ōshinke chō,” 17 – 18. 52. DNS 1:6, 755, Jōhei 3 (933) n.d. There is also a note in the Kugyō bunin, a record of members of the Council of State from Nara to late medieval times, next to Tadahira’s name for Tenryaku 3 (949). See Kugyō bunin, 190. 53. See Heian ibun, vol. 1, doc. 217, Engi 20 (920) 9/11; and DNS 1:5, 235 – 236. Concerning the process by which household missives became official orders, see Nishibeppu, “Ōshinke chō.” For a useful chart tracing the development of the Sekkanke household organization beginning in Tadahira’s time, see Satō Kenji, Chūsei kenmon no seiritsu to kasei, 71. 54. Sonpi bunmyaku, 1:51. Compilation of the Sonpi bunmyaku — a massive compilation of aristocratic genealogies — began in the fourteenth century. But a modern register of provincial governors compiled by Miyazaki Yasumitsu, Kokushi bunin, does not list Tsunekuni as a governor in Musashi. 55. Fukuda Toyohiko, Taira Masakado no ran, 66 – 69. According to Tadahira’s journal, other such wardens received special treatment. See TKK 184, Tengyō 2 (939) 2/13. In English see William Wayne Farris, Heavenly Warriors, 131 – 146. 56. For the Masakado ki in English see Judith Rabinovitch’s translation: Shōmonki: The Story of Masakado’s Rebellion. As one more instance of Tadahira’s provincial interests, Mori Kimiyuki suggests that Tadahira maintained close relations with Munakata Shrine and its vicinity in Kyushu. He thinks that the Takada Horse Farm there may have been in existence in Tadahira’s time, and that someone from there might well have spent time in Tadahira’s household. See Mori Kimiyuki, “Ōshinke to gunji,” 14. 57. Concerning these properties, see Hashimoto Yoshihiko, “Fujiwara shi chōja to wata riryō,” esp. 247 – 255. Also see Fukutō Sanae, “Heian jidai no uji,” 15 – 17. An early reference to Katakami Estate is found in Tadahira’s TKK, in the entry dated Enchō 5 (927) 12/5. 58. Nihon kiryaku, Enchō 8 (930) 09/22. See it in Kōshitsu seido shiryō: Sesshō, 1:22. 59. Honchō monzui, sec. 4, cited in Kōshitsu seido shiryō, 1:8 – 9, 158 – 159. Honchō monzui is a collection of belles lettres compiled in the mid-eleventh century by the scholar-official and literatus Fujiwara no Akihira (? – 1066). 60. Kōshitsu seido shiryō, 1:4, 16. Concerning Shōtoku’s role in Suiko’s court, see Piggott, Emergence, 79 – 83. 61. Jōgan 8 (866) 08/19, in Kōshitsu seido shiryō, 1:4, 46. There is substantial new research on the Jōwa (842) and Ōtenmon (866) coups. See Yoshida Takashi, “Kyū, jū seiki no Nihon;” Tamai, “Jū, jūichi seiki”; Endō Keita, “ ‘Shoku Nihon kōki’ to Jōwa no hen”; and Kamiya Masayoshi, “Jōwa no hen to Ōtenmon no hen.” The earlier Jōwa Coup in 842 strengthened Yoshifusa’s position at court after the death of the retired Saga Tennō. In the aftermath, the succession was settled in favor of the future Montoku Tennō, Yoshifusa’s nephew. Modern historians have viewed the Jōwa Coup as a plot by Yoshifusa in league with Saga’s widowed queen consort, Tachibana no Kachiko, and supporters of the future ruler Montoku. See Satō Sōjun, “Saga
62 | joan r. piggott Tennō ron,” in Heian zenki seiji shi josetsu, 99 – 120. In English, see William McCullough, “The Heian Court, 794 – 1070,” 48 – 50. 62. Nihon sandai jitsuroku, Jōgan 18 (876) 11/29, cited in Kōshitsu seido shiryō, 1:20 – 21. Earlier this epithet, “the faithful and just minister,” had been bestowed on Yoshifusa. 63. In the twelfth month of 930, another royal order granted Regent Tadahira two inner palace attendants (udoneri) and eight outriders (zuijin) from the Inner Palace Guards, following precedents from Yoshifusa’s era. See Kōshitsu seido shiryō, 1:112 – 113. 64. For Kuroita Nobuo, a key moment came in 936, when Yasutada, Tokihira’s son, died. Only then was it clear that Tadahira’s own line would replace that of Tokihira in leading the court. See Kuroita, “Fujiwara Tadahira seiken ni taisuru ichikōsatsu,” 135 – 136. 65. Seiji yōryaku, Ninna 3 (887) 11/21, cited in Kōshitsu seido shiryō, 1:4 – 5. 66. Seiji yōryaku, Ninna 4 (888) 6/2, cited in Kōshitsu seido shiryō, 2:178 – 179. 67. Kōshitsu seido shiryō, 2:178; and DNS 1:8, 36 – 37. 68. Five years later, when the twenty-year-old Murakami Tennō succeeded to the throne, that monarch ordered Tadahira to serve as his chief of staff as well. See TKK 288, Tengyō 9 (946) 4/21. A particularly useful recent discussion of the development of the sesshō and kanpaku posts is Takinami Sadako, “Akō no fungi.” 69. TKK 135, Jōhei 1/1/29. For the drawing up of a list of provinces needing new governors, see TKK 156 – 157, Jōhei 2 (932) 6/23 and 7/3. And for discussions concerning appointments, see entries from the third month of 939, TKK 138 – 139. On 3/15 the results were memorialized to the throne by Minamoto no Kintada, concurrently a fifth-ranking royal secretary (go’i kurōdo) and right junior controller. See TKK 138 – 139. Kintada’s post in the Royal Secretariat (Kurōdo dokoro) gave him access to the throne while his controller’s post gave him access to the Council. 70. TKK 138, Jōhei 1/3/13. 71. The Teishinkō ki records such a meeting in the third month of 940, with Tadahira’s elder brother, Minister of the Left Fujiwara no Nakahira, presiding as the ranking Council member (ichinokami). Two days later, Council members met with Tadahira. At that time Nakahira became note taker while Tadahira, the senior official in the group, presided. In this case, office trumped age: Nakahira, while Tadahira’s elder brother, did not have Tadahira’s interest or talent for politics. Entries in the Teishinkō ki record that Nakahira was always leaving meetings early. For instance, see TKK 204, Tengyō 3 (940) 3/19. 72. TKK 204. See also Ōmanaribumishō, ed. Yoshida Sanae. 73. See TKK 214, Tengyō 8 (945) 3/27. In fact, the tennō had his coming-of-age ceremony in 937. But only in 941 did he accept Tadahira’s resignation as regent, and then he reappointed him chief of staff. 74. TKK 226, Tengyō 9 (946) 2/6. The discussions began on 2/3 and continued through 2/7. See also TKK 252 – 254, Tenryaku 2 (948) 1/4, 1/26, 1/29, 1/30. 75. On the Shiki no mizōshi, the shiki of which denoted the Queen Consort’s Household Agency (Chūgūshiki), see Okamura Sachiko, “Shiki no mizōshi ni tsuite.” 76. Morita and Yoshikawa argue that the full evaluation process had developed by about 945. See Morita, Zuryō, and Yoshikawa, “Sekkan seiji no tensei.” 77. TKK 189 – 190. 78. TKK 144, Jōhei 1/int. 5/5. 79. TKK 141 – 142, Jōhei 1/4/23. Also see TKK 192, Tengyō 2/8/17 (departure banquet for Mutsu governor); and DNS 1:9, 339 (Tenryaku 3 [949] 3/2), when, because Tadahira was sick, his son Minister of the Right Morosuke bestowed gifts on the departing Iyo governor in Tadahira’s stead. 80. TKK 187, Tengyō 2/4/26.
Court and Provinces under Fujiwara no Tadahira | 63 81. Nihon kiryaku, Tenryaku 3/8/2, in DNS 1:9, 390. 82. See, for instance, TKK 139, Jōhei 1/3/25. 83. Kiley, “Provincial Administration,” 238 – 239. 84. Seiji yōryaku, in DNS 1:6, 766. 85. DNS 1:6, 765 – 768. According to Terauchi, after 888 provincial governors were responsible for repairing and cleaning official facilities in their provinces, even those that had not been kept in repair by a former governor. See Terauchi, “Zuryō kōka seido no seiritsu tenkai,” 46 – 52. 86. Seiji yōryaku, cited in DNS 1:6, 817. 87. TKK 156, Jōhei 2/6/8. Tadahira was contacted about the case by Minamoto no Kintada, who was serving as the fifth-ranking secretary (go’i kurōdo) and as right junior controller (ushōben). 88. DNS 1:7, 53, Jōhei 6 (936) int. 11/5. See also DNS 1:8, 428, 442 – 443, Tengyō 8/1/6, 2/19. 89. Ruijū fusenshō, a Council order dated Tenryaku 3/6/13, cited in DNS 1:9, 375. Earlier instances can be seen in Honchō seiki, Jōhei 3/8/15, cited in DNS 1:6, 746. See also Ruijū fusenshō, Tengyō 8/3/8 in DNS 1:8, 446 – 448. The Honchō seiki is a compendium of court records assembled by Fujiwara no Michinori (alias Shinzei, 1106 – 1159), beginning about 1150. 90. TKK 139, Jōhei 1/3/22. 91. For examples, see TKK 139, Jōhei 1/3/22; and TKK 195, Tengyō 2/11/2. 92. Kiley, “Provincial Administration,” 312. 93. Taxes from the San’in circuit were particularly deficient in 943. So the court called for governors’ proposals as to how to resolve problems on 943/12/12 and again in 944. Tadahira read the responses on Tengyō 8/1/4, according to the Teishinkō ki. The proposals confirmed that no governor should be reappointed before his last tenure had been carefully evaluated (kōka sadame, lit. “deliberations as to praise or blame”) and “failings large and small” were to be considered. See Seiji yōryaku, Tenryaku 2/12/1, cited in DNS 1:9, 297; and Tengyō 8/1/6, cited in DNS 1:8, 428. 94. For an example, see TKK 143, Jōhei 1/5/25. Tadahira approved Council directives ordering the opening of new fields (tsukuda) in the provinces (TKK 143, Jōhei 1/5/25). And he instructed the left middle controller to send directives (kanpu) to the various provinces concerning the opening of new fields the following spring (TKK 149, Jōhei 1/11/17). 95. Kinai provinces were to pay in coins, but other provinces were to send rice from their emergency stores (fudōkoku). See Honchō seiki, Tengyō 1 (938) 10/17, in DNS 1:7, 289 – 290. 96. DNS 1:6, 801, Jōhei 4 (934) 8/29. 97. TKK 195, Tengyō 2/11/10. 98. These were banquets that had been held more frequently in earlier times. By the later ninth century, they were held only on the first day of the fourth and tenth months. 99. TKK 205, Tengyō 3/4/1. 100. TKK 218, Tengyō 8/8/23. 101. TKK 213, Tengyō 8/3/16 – 3/19. 102. TKK 228, Tengyō 9/4/11. 103. TKK 244, Tenryaku 1 (947) 3/29. 104. See Piggott, Emergence, 144 – 149, 208 – 231. On propitiating kami, see the Jingiryō chapter of the Yōrō Code in Inoue Mitsusada, Ritsuryō, 211 – 215. An excellent overview is Ninomiya Masahiko, “Kodai saishi seido no kōsatsu.” 105. On the annual prayer festival (kinensai) in English see Felicia Bock, “Engi-shiki”: Procedures of the Engi Era, 1:59 – 65, and 2:66 – 70, 107 – 171. 106. Ninomiya, “Kodai.” 107. TKK 172, Tengyō 1/6/22.
64 | joan r. piggott 108. The order came from Suzaku Tennō, after it was requested in a memorial by Tadahira’s son and heir, Senior Counselor Fujiwara no Saneyori. See DNS 1:8, 339 – 349, Tengyō 7 (944) 4/22. 109. Honchō seiki, Tengyō 8/7/7. 110. Kuroda Hideo traces the songs to Kyushu and the Kantō, which he thinks represent the “west” and “east” of the record. See Kuroda Hideo, “Nōgyō gijutsu to minshu ishiki,” 412 – 415. 111. Rihōō ki, Tengyō 8/8/2. 112. Rihōō ki, Tengyō 8/8/2. Concerning the lightning strike on the Seiryōden, see the entry dated Enchō 8/6/2 in the late Heian-period Buddhist chronicle Fusō ryakki. For a complete discussion of the entire Sugawara onryō phenomenon, see Tsunoda Bun’ei, “Fujiwara no Tadahira no eitatsu,” 257 – 272. 113. Honchō seiki, Tengyō 8/8/3. 114. Concerning the arrival of the Shidara deity in the vicinity of the capital, see Hayashiya Tatsusaburō, ed., Kyōto no rekishi, 1:419 – 424; Toda Yoshimi, “Atarashii kami to ‘fugō no tomogara’ ”; and Kuroda, “Nōgyō gijutsu.” A convenient abstract of the extant sources can be found in DNS 1:8, 473 – 475. 115. Rabinovitch, Shōmonki, 111 – 112. 116. For various interpretations of the Shidara deity, see Shibata Minoru, “Hachiman no kami no ichi seikaku,” Kuroda, “Nōgyō gijutsu”; and Kawane Yoshiyasu, “Ritsuryō kokka no henshitsu to bunka no tenkan.” 117. TKK 155, Jōhei 2/4/28. 118. TKK 163, Jōhei 2/12/16; DNS 1:6, 629, 679, 762, 797 – 798. 119. There are several English translations including one by G.W.Sargent in Donald Keene, Anthology of Japanese Literature, 82 – 91. 120. Besshū fusenshō, Jōhei 5 (935) 11/24, cited in DNS 1:6, 948. Also see Kintada Ason shū, Jōhei 5/12/3, in DNS 1:6, 952. The delegate’s title was karamonoshi, “Emissary to Seek Goods from China.” The record shows that another ship docked and still another delegation was sent just a year later, so this was not an isolated incident. See also DNS 1:7, Jōhei 6 (936) 7/29, 29. 121. DNS 1:8, Jōhei 6/11/17, 744. The order was authorized by Tadahira’s heir, Minister of the Right Saneyori. It would have been approved during “royal inspection” (nairan) by Regent Tadahira. 122. For instance, bandits attacked the Akita district office in Dewa Province in the summer of 939, where onlookers reported that the marauders took demonic form. See TKK 187, Tengyō 2/5/6. 123. On Sumitomo’s piracy see Fukuda Toyohiko, “Fujiwara no Sumitomo to sono ran”; Taira no Masakado shiryōshū tsuke Fujiwara no Sumitomo shiryō, esp. the documents (shiryō) section; and Matsubara Hironobu, Fujiwara no Sumitomo. 124. According to Matsubara Hironobu, Sumitomo was probably already in league with the pirates by 936, which would mean that his appointment in 937 was an attempt to lure him over to the court’s side. See Matsubara, Fujiwara no Sumitomo, 131 – 142. 125. If there were earlier entries in the TKK, they may not be extant. We have no entries for 933 through 937. See the entries of TKK for 939, passim. 126. DNS 1:7, 581 Tengyō 3/1/1. 127. TKK 202, Tengyō 3/2/26. For more references to this important transport node, see TKK 141 – 142, Jōhei 1/4/23, 12/2; and on the Yodo port, see TKK 149 – 50, Jōhei 1/12/2. 128. Kobayashi Shōji, “Fujiwara no Sumitomo no ran sairon.”
Court and Provinces under Fujiwara no Tadahira | 65 129. See DNS 1:8, passim. For instance, on Tengyō 9/12/10, kebiishi were appointed to Izumi, Tanba, and Kii to deal with violent forces in those provinces. Also see Taniguchi Akira, “Shokoku shinsei zatsuji,” 58. 130. Nihon kiryaku and Honchō seiki, Tengyō 5 (942) 6/30, in DNS 1:8, 144 – 145. 131. TKK 217, Tengyō 8/6/30. 132. See the Emoluments (Rokuryō) chapter in Inoue Mitsusada, Ritsuryō, 307. For the rec ord of Jōhei 3 (933) concerning the decreasing of the queen’s prebendal units, see DNS 1:6, 704. 133. TKK 218, 219, Tengyō 8/7/14, 8/19; DNS 1:8, 486 – 488. 134. Honchō seiki, Tengyō 8/10/14, cited in DNS 1:8, 428, 518; TKK 218, Tengyō 8/7/14, 8/19. Narikuni lacked political skills. He appears subsequently as the provisional second in command of the Right Palace Guards in 947, when he was accused of violating a royal order (Nihon kiryaku, Tenryaku 1 [947] int. 7/23, cited in DNS 1:9, 38). For a biography see Tsunoda Bun’ei, Heian jidai shi jiten, 2159 – 2160. This was not the only run-in between provincial authorities and agents of powerful houses that year. In the late spring, Tadahira received a royal missive charging that agents of his elder brother, Senior Minister of the Left Fujiwara no Nakahira, had committed crimes. In that instance, the household director (bettō) was actually jailed (TKK 215, Tengyō 8/4/13, 4/14). 135. Seiji yōryaku vol. 2, Tenryaku 1/int. 7/23, 271 – 272; Morita, “Fujiwara no Tadahira seiken no dōkō,” 270. 136. TKK 259, Tenryaku 2/6/4. 137. Heian ibun, vol. 11, fu 262, Tenryaku 3/3/n.d. It should be noted that the editor of Heian ibun, Takeuchi Rizō, found these documents suspect. 138. In 969 the senior member of the Council, Minamoto no Takaakira, was accused of plotting and exiled to the Dazaifu in Kyushu. Takaakira’s misfortune was the good fortune of Tadahira’s sons, Saneyori and Morosuke, whose leadership at court was then unrivaled. An important ally to both Saneyori and Morosuke at the time was Minamoto no Mitsunaka, a martial governor whose Seiwa Genji descendents thereafter flourished as the “claws and fangs” of the Regents Line. See Yamamoto Nobuyoshi, “Reizei chō ni okeru Ononomiyake, Kujōke wo megutte”; Oboroya Hisashi, Seiwa Genji, 42 – 84; and DNS 2:3, 29 – 44 (Chōtoku 3 [997], n.d.). 139. Nihon kiryaku, Tenryaku 3/6/6, DNS 1:9, 374. 140. The special circumstances of the Kinai are described in Ichi, “Kyūseiki Kinai.” 141. For current overviews and evaluations of Tadahira’s court leadership see Satō Sōjun, “Fujiwara no Tadahira seiken no keisei”; Morita, “Fujiwara no Tadahira seiken no dōkō”; Yoshikawa, “Sekkan seiji no tensei”; Sasaki Muneo, “Sesshō sei kanpaku sei no seiritsu”; and Ōtsu Tōru, Kodai no tennō sei. 142. See Satō Sōjun, “Fujiwara no Tadahira.” 143. See, for instance, Tamai, “Jū, jūichi seiki no Nihon, sekkan seiji,” esp. 23. In English see Bruce Batten’s “Provincial Administration in Early Japan.” 144. See Charlotte von Verschuer’s discussion in this volume. 145. Ōtsu, Kodai no tennō sei, esp. 103 – 105. 146. Indeed, Terauchi Hiroshi has traced the emergence of wealthy governors to the late tenth and early eleventh centuries. See Terauchi, “Sekkan ki no zuryō to seifu no chikusei.” And for one approach to these changes, see Ōtsu Tōru, “Sekkanki no kokka kōzō.”
4
d G.Cameron Hurst III
Kugyō and Zuryō Center and Periphery in the Era of Fujiwara no Michinaga
C
onsiderations of center and periphery in Heian Japan involve at the highest level the relationship between the central political apparatus and the hinterlands, the sixty-six provinces and two islands (Iki and Tsushima) that constituted the conceptualized polity. Indeed, the very existence of a state presupposes control over the human and material resources of the space thought to constitute the geographical area of that conceptualized state. This was never an easy task for Japanese rulers, despite the insular character of Japan that seems to lend itself to simple mapping of a discrete geopolitical entity. Recall, for example, the court’s hard-fought campaigns against the Emishi to extend its control over the northeast. Central government – provincial interaction over the entire Heian period is too broad a topic for this chapter. I have already considered this issue, at least tangentially, in several other works dealing with the late Heian period.1 And with the publication of volume 2 of The Cambridge History of Japan, other aspects of the changes the central government adopted in order to control the provinces in the early and mid-Heian eras have been treated.2 But the exact nature of centralprovincial relations in the mid-Heian period, from 967 to 1068, when the Fujiwara Regents House (sekkanke) enjoyed its greatest political influence, remains understudied. Below, however, I will focus primarily on the period considered to be the height of the regency period, the era of Fujiwara no Michinaga (966 – 1027). Given the information available in the Cambridge History, my analysis need be only partially institutional; instead, it will concentrate on the human elements. That is, I am concerned with how the regent ruled. Certainly Michinaga was the effective ruler of Japan in mid-Heian, and he ruled by controlling both politics at the center (Heian-kyō) and the flow of resources from the periphery (provinces). Below, I hope to demonstrate the nature of personal relationships between and among the various kugyō (ranking noble) houses, including the Fujiwara Regents House and the Imperial House. What was the overlap between personal and political relations? How did marital relations affect politics? How was Michinaga able to dominate the organs of state? Turning to the periphery, I will examine the relationship between Michinaga and the middle-ranking provincial governors (zuryō) who provided the resource base for his opulent lifestyle. How were appointments to governorships decided? How did governors form relationships
Kugyō and Zuryō | 67
and curry favor with Michinaga? Years ago I sketched out in outline form the way in which marital and patron-client relationships worked in the Heian period.3 In this essay I hope to flesh out that earlier work. By the mid-Heian age the Fujiwara Regents House had secured what must have seemed to others permanent control over the court, through the domination of two extracodal offices: regent (sesshō) and chancellor (kanpaku).4 The seeds of this exercise of power lay in the close marital alliance that the lineage was able to form with the Imperial House, virtually monopolizing the several positions of principal imperial consort over a long period. Imperial offspring were thus normally born of mothers from the Fujiwara regental line; and in the predominantly uxorilocal residential pattern of the period, Fujiwara kinsmen exercised considerable power over sovereigns, especially when it became common to enthrone children. Acting on behalf of child emperors, Fujiwara politicians served as regent; continuing to exercise power even after an emperor reached adulthood, they held the position of chancellor, noticeably weaker yet still most powerful among the senior nobility.5 This mid-Heian period is currently referred to in Japan as the ōchō kokka, a widely used term rendered in the Cambridge History as “the royal court State.”6 It postulates a basic change in the original ritsuryō, or statutory, system of the Nara and early Heian eras in which the court essentially delegated the details of provincial rule to local governors in return for a steady flow of income to Kyoto. By the end of the ninth century the statutory state could no longer effectively manage the provincial taxation and census mechanisms of the borrowed Tang-style codes, and the central government was at loggerheads with provincial officials on how to solve the issues. The “new” system that unfolded under the administrations of Fujiwara no Tokihira (871 – 909) and Tadahira at the beginning of the tenth century can be summarized in two main points. First, the public (taxable) lands in each province would now be organized into units called myō for the purpose of levying taxes and tribute. Second, the central government essentially ceased its involvement in the details of provincial administration in return for relegating provincial governors to little more than tax collectors. The ramifications of these changes were major, and the secondary literature on their implications is extensive. I will return to them in more detail when considering Michinaga’s relations with the periphery, but I will first examine the center of the ōchō kokka, the capital at Heian, in Michinaga’s era. The most prominent political features of the period were the establishment of the Fujiwara regency and a widespread tendency to privatize formerly public functions, as well as to monopolize certain offices within designated families.
Michinaga Controls the Center How are we to regard Fujiwara no Michinaga? Standard historical accounts are consistent. He is referred to as the “greatest statesman in the Heian period” or the most powerful of all Fujiwara regents, the “Fujiwara regent par excellence,”7 according to William McCullough, or in George Sansom’s assessment, the “greatest
68 | g. cameron hurst iii
of the regents.” But what do we mean by that, exactly? Was he the most influential man the Fujiwara ever produced? And if so, measured by what standard? Was Michinaga the wisest man of Heian times? Was he the greatest poet? The richest man? Did he institute administrative or fiscal policies that dramatically altered the course of Heian history? Did he bring peace and prosperity to the realm? Did he leave behind a physical legacy upon which subsequent generations might gaze in awe and say, “Michinaga built that”? One can only answer most of these questions in the negative. To be sure, Michinaga was intelligent, talented, and wealthy, but not markedly more so than his associates. He did not devise great political or economic programs; he did not renew the realm and propel it to new heights of power and prestige. Even such temporarily successful fiscal reforms as shōen regulations were the product of earlier or later times; none date from Michinaga’s era. In fact, it could be argued that the Heian court was not at its zenith in Michinaga’s time, at least not politically or economically. Certainly, the realm was none too pacific during his days. But Michinaga enjoyed great personal power, and could dominate the political process as few others in Heian times. In terms of office and rank — and the associated social and economic perquisites — Michinaga held sway over his world. In that sense his famous poem, composed upon the appointment of his daughter Ishi as Emperor Go-Ichijō’s principal consort in 1018, sums up the situation succinctly: This world, I think, Is indeed my world. Like the full moon I shine, Uncovered by any cloud!8 His offspring flourished, and the regency was secure within his line. Perhaps Michinaga’s main claim to fame as a great figure is that he presided over a society at its cultural zenith. As Ōe no Masafusa (1041 – 1111) noted in his Zoku honchō ōjōden, pointing to eighty-six men and women of talent in twenty different fields, the era produced an uncommon number of talented people.9 Interestingly, he did not include in this group either Sei Shōnagon or Murasaki Shikibu, probably because the writing they produced was not considered in the “public” realm. Be that as it may, one can certainly not credit Michinaga for the appearance of all these talented figures. Indeed, Masafusa does not even list Michinaga in any of these categories, and he refers collectively to them as great figures at the court of Emperor Ichijō. Compared to many civilizations, the Heian age did not produce great monuments: no pyramids like those of the pharaohs, no massive tombs like the Japanese rulers of the fifth and sixth centuries. Yet Michinaga loved extravagance and splendor, as attested to by many contemporary accounts. His Tsuchimikado mansion was certainly luxurious, appropriately so as the locus of many imperial visits. The construction of his Hōjōji temple-residence complex was a major project for its day, as “Michinaga urged the work on with floods of orders, chaf-
Kugyō and Zuryō | 69
ing at the slowness of dawn and bemoaning the gathering of shadows at night”;10 and the glory of Michinaga’s house was displayed for all to see. Yet today, nothing physical remains of his legacy, save perhaps the Byōdōin at Uji. But even that magnificent compound is more properly associated with his son Yorimichi, who, after inheriting it from his father, turned it into the historical building and grounds we now enjoy. How then are we to judge Michinaga’s “power”? How was Michinaga, in the words of William McCullough, able to “carry all before him at Kyoto, crushing his enemies, disposing of people and official posts mostly as he saw fit”?11 How did a regent such as Michinaga — and he was actually regent for just over a year after Go-Ichijō’s accession in 1016 — exercise such seemingly dictatorial powers? Despite the Ōkagami’s insistence upon his bravery and portrait of him as an excellent horseman and archer,12 Michinaga commanded no physical resources to force men to his will, in the normal sense of that term. Michinaga’s skills were civil rather than military. He was able to defeat potential enemies through marriage politics and by guile and intrigue, by control of the administrative machinery of state, and by control of the fiscal resources of the realm to achieve a dominance that few others have enjoyed in Japanese history “Family Ties”: Michinaga’s Kin Michinaga’s ability to effect decisions of state depended upon both the rank and office he achieved and his personal relationships with the Imperial House. Power and success by mid-Heian Japan had to do as much with personal relationships as anything else, as the privatization of the organs of state proceeded apace. Michinaga orchestrated things so that family members, kinsmen, and clients obtained important positions at both the center and the periphery, providing him with allies almost everywhere and no political opponents of consequence anywhere. The Ōkagami attributes Michinaga’s phenomenal success to “his boundless good fortune,”13 and indeed fate seems to have taken a hand all along the way. No one in Heian times — indeed at any time in premodern Japanese history — enjoyed the kind of family ties that allowed him to control the allocation of political, economic, and social resources as did Michinaga. Later warrior leaders enjoyed more raw personal power than Michinaga, served by legions of vassals. But that was power won by the sword. Michinaga’s power was achieved within the system, by strategic marriages, shrewd political moves, and much luck. As the fifth son of Regent and Chancellor Kaneie (whose daughter Senshi was Emperor Ichijō’s mother), Michinaga certainly must have anticipated a successful career, but he could not have dreamed of becoming the most notable politician of his age. In 995, however, several senior courtiers were carried away in a virulent epidemic or died suddenly. Included in this group were his two older brothers, Michikane and Michitaka, both of whom had served as chancellors after the death of their father Kaneie in 990 (figure 4.1). Indeed, the Ōkagami notes that “the explanation for all those deaths is simply that Michinaga is a supremely lucky man. He would never have risen so high if others had kept the offices to which they were entitled by seniority.”14 When his two brothers passed away at
70 | g. cameron hurst iii
ages forty-two and thirty-four respectively, Michinaga was appointed document examiner, minister of the right, and Fujiwara clan chieftain when he was only twenty-nine.15 The next year, he vaulted over all other courtiers when he became minister of the left, a post he held for the next twenty-one years until his appointment as prime minister in 1017.16 If Michinaga was lucky that the death of many other nobles opened for him an unimpeded path to success, his success nonetheless depended more upon the distaff members of his family. He married three daughters to successive emperors and made three more consorts to crown princes, and thus for much of his career Michinaga was closely related to the imperial family as uncle, father-inlaw, and grandfather of sovereigns. Michinaga was truly blessed in his family relations, especially in his selection of wives. He married Minamoto no Rinshi in 987, when he was twenty-one years old and had just become a kugyō. Rinshi was the daughter of the senior court noble, Minister of the Left Minamoto no Masanobu, grandson of former Emperor Uda and son of Prince Atsuzane. Rinshi’s mother was the daughter of the minister of the right in the late Daigo era, Sadakata. According to the Eiga monogatari, Masanobu was not initially pleased with Michinaga’s proposal of marriage.17 Perhaps the deciding factor in his accepting Michinaga’s proposal was that his elder brother Michitsuna was already married to another of Masanobu’s daughters. But Michinaga’s desire for a bride from this powerful Uda Genji lineage was fully in the tradition of Fujiwara Regents House males seeking brides from princely houses. Yet just a year later Michinaga married another powerful Minamoto family woman, Meishi. She was the daughter of the late minister of the left Takaakira, victim of a Fujiwara plot in which he was exiled to Kyushu in the Anna Incident of 969. Michinaga’s elder sister Senshi became the guardian of Meishi, and according to Ōkagami, she rejected the petitions of her two younger brothers Michitaka and Michikane and ensured that her favorite younger brother Michinaga would marry Meishi.18 These two women provided Michinaga with twelve children in all, six males and six females, almost as though they were “competing in the production of offspring.”19 The daughters were of course the key to his future political success, since all but one became a principal imperial consort or a consort to a crown prince. His senior wife Rinshi presented Michinaga with six children, four daughters and two sons. Both sons, Yorimichi and Norimichi, became chancellors; but it was the daughters who were really responsible for Michinaga’s long period of dominance at court. The eldest daughter was Shōshi, principal consort of Ichijō and the mother of two sovereigns — Go-Ichijō and Go-Suzaku. Better known by her honorary title Jōtōmon’in, Shōshi was raised to the position of senior grand imperial dowager and was the most notable lady in the realm. Kenshi, Michinaga’s second daughter, became Sanjō’s bride while he was crown prince and was raised to principal imperial consort upon his accession. Although she had no sons, she did bear an imperial princess and also enjoyed considerable status at court. The third daughter, Ishi, married Go-Ichijō just after his capping ceremony
Kugyō and Zuryō | 71
Figure 4.1. Sekkanke genealogy during Michinaga’s times. Source: Adapted from Mikael Adolphson, The Gates of Power (Honolulu: University of Hawai‘i Press, 2000).
and bore him a princess. Finally, Kishi, the fourth daughter, became the consort of Go-Suzaku while he was still crown prince and was the mother of Go-Reizei. Rinshi not only provided Michinaga with very successful children: she herself was so revered that she was accorded status equivalent to that of an empress. According to the Ōkagami, she was at the time of its writing “mother of three principal imperial consorts, the crown prince’s consort, the regent and the palace minister (to say nothing of her being the grandmother of the emperor and the crown prince). She is certainly the mother of the nation.”20 These children were produced over a long period, so that there was a significant age difference between elder and younger siblings. Thus Michinaga married Shōshi, his eldest daughter by Rinshi, to Emperor Ichijō; and she bore him two male offspring who became emperors as well: Go-Ichijō and Go-Suzaku. In turn, these emperors were married to Shōshi’s younger sisters Ishi and Kishi respectively, making Michinaga grandfather of both emperor and crown prince. Moreover, when seventeen-year-old Ishi married her eight-year-old nephew Emperor Go-Ichijō in 1018, Michinaga was simultaneously the emperor’s grandfather and father-in-law, as complex a web of marriage relationships that intertwined imperial and regental houses as ever existed in Japanese history. By contrast, Meishi’s offspring were less successful. Although her sons Yorimune, Yoshinobu, and Nagaie all became major court figures, none rose to regent or chancellor. Another son, Akinobu, became a monk. Moreover, Yorimichi succeeded his father and served as chancellor for almost a half century during the reigns of his nephews Go-Ichijō and Go-Suzaku and his grandnephew Go-Reizei, thus limiting the advancement of his half brothers. Likewise, Meishi’s daughters
72 | g. cameron hurst iii
never became mothers of emperors, but Kanshi did marry Ko-Ichijō’in, a son of Emperor Sanjō who was driven from the position of crown prince by Michinaga and replaced with Michinaga’s own grandson Atsuyoshi (future Emperor GoSuzaku) — who was then married to Michinaga’s daughter Kishi. This review of the complex marital relationships illustrates how closely intertwined Michinaga was with the Imperial House. The politics of Heian marriage were explained by the late William McCullough,21 but it behooves us here to review a few of the features, since it is the key to understanding Michinaga’s ability to effect decisions of state. Although marital institutions were changing in midHeian Japan, uxorilocal practices were still prevalent. Whether a major court figure resided with his wife (wives) or lived alone and visited them alternately,22 women most often remained at their parents’ house after marrying, and thus children were raised by their maternal kinsmen rather than their paternal ones. Marriage was arranged to secure a bridegroom for the daughters of ranking families, and the Japanese term — mukotori — literally means “bridegroom welcoming.” In mid-Heian times, it was most common for a bridegroom to move into the residence of his in-laws, who provided the couple with rooms, furnishings, food, and so on. In fact, the wife’s parents were usually far more solicitous of their new son-in-law than of their own sons, who were sent off to other families as bridegrooms themselves. Financially, it meant that in the first half of married life, a man would normally be supported by his wife’s family; and only on their death might he be obligated to provide some financial support for his wife — although it was common for the daughter to inherit her parents’ residence in which the couple may have been living.23 Children of the marriage resided with their mothers at the home of their maternal grandparents, who had the primary responsibility for their upbringing. A Heian man, then, may have depended upon his father for rank and office in his public life; but in his private life, much more influence lay with the families of his mother and his wife, or wives. Understanding Heian marriage patterns helps to explain the significant influence that regents and chancellors had as grandfathers, uncles, and fathers-in-law of emperors of the time. It was based upon the maternal connection (gaiseki). It should be noted, however, that the Imperial House operated somewhat differently as far as marriage was concerned. Daughters of higher nobles such as Michinaga were introduced into the Imperial Palace as consorts. Such was the case with his daughter Shōshi, for example, where Michinaga provided all the furnishing for her apartments within the palace and to which the emperor then went for the formal “marriage,” thus maintaining the form at least of its being a mukotori situation.24 However, we need to pursue the matter further. There is a tendency toward impreciseness in claiming that a Fujiwara lord was able to become regent because he had established a gaiseki relationship with an emperor, which means little more than that he was a maternal relative of the emperor. Another term often used to refer to the regent-emperor connection is miuchi, which is also vague, meaning that one is an extremely close relative of the emperor. Such relations to a sovereign could take a number of different patterns, and the proximity of
Kugyō and Zuryō | 73
that miuchi bond made each relationship between regent and emperor different, depending on a number of factors: whether both, one, or neither of the emperor’s parents survived; the relationship of those parents to the regent; the relationship of emperor and the main consort; and, of course, the key relationship between regent and emperor.25 Naturally, throughout the Heian period the nature of this complex web of relationships varied for each regent, and the sovereign(s) those regents served differed. It was clearly most advantageous from the point of view of the Fujiwara leader to be the grandfather of a very young emperor. But even then, it mattered a great deal what other miuchi were still around. When he first came to power under Emperor Ichijō, for example, Michinaga enjoyed reasonably close relations as the emperor’s uncle. The emperor was only fifteen when Michinaga received the right of document examination. Ichijō died in 1011 at only thirty-one, ending his career at almost the same age that Michinaga began his. So Ichijō was only in his teens and twenties during his association with Michinaga. His mother Senshi, Michinaga’s doting elder sister, was still alive; but his father, Retired Emperor En’yū, had passed away in 991. Thus Ichijō naturally looked to his mother and her favorite brother, his uncle Michinaga, for support. The regent-emperor relationship seems to have been especially close. After Ichijō’s death, Michinaga faced a radically different situation when thirtysix-year-old Emperor Sanjō ascended the throne. He was the son of the late Reizei (who also died in 1011) and Michinaga’s sister Chōshi. Again, as uncle of the sovereign, Michinaga was a close kinsman; but his sister Chōshi was also long since dead. Moreover, Sanjō had four sons by his wife, who was only a daughter of Michinaga’s cousin Naritoki. Only later, right before Sanjō’s accession, did Michinaga arrange to have his daughter Kenshi married to the sovereign and soon raised to the honorific position of principal imperial consort, although she had produced no male heir. Since he had more difficulty dominating politics in the reign of Sanjō, Michinaga put considerable pressure on the sovereign to abdicate. Fortune again cooperated: Sanjō suffered from failing eyesight, and Michinaga was able to persuade him to abdicate after only five years on the throne. He died only a year later. Michinaga further was able to coax Sanjō’s son Crown Prince Atsuakira to yield that post in favor of Shōshi’s son Prince Atsuyoshi. Thus Michinaga’s grandson Atsuyoshi, as Go-Ichijō, succeeded Sanjō, and his daughter Ishi was raised to grand imperial consort, occasioning the “full moon” poem quoted earlier. As grandfather and father-in-law of the eight-year-old emperor, Michinaga was far more readily able to exercise control over the imperial position than he had been with Sanjō. Confident of his powers, he resigned the regency in favor of his son Yorimichi. Thus the core of Michinaga’s power lay in his close relationship with several successive emperors. Thanks were due in large part to the role played by his sister Senshi in getting Michinaga properly married and brokering his relationship with Ichijō; and even greater credit must be given to his daughter Shōshi as consort and mother of two emperors. Here, too, Michinaga was extremely lucky. Early in Ichijō’s reign, Shōshi was far too young to become a consort, and Michinaga could only stand by as his brothers and others sent their daughters off to the pal-
74 | g. cameron hurst iii
ace. Yet none of the consorts was able to produce a future emperor. So although Michinaga’s eldest daughter was quite a bit younger than Ichijō’s other consorts, matters transpired to favor her as grand imperial consort and ultimately mother of two emperors herself. This was the key to Michinaga’s success. Michinaga’s Domination of the Court After the deaths of his brothers, Michinaga found himself in power, as minister of the left and with the right of document examination (nairan) for his nephew Emperor Ichijō, when he was only thirty years of age. Even from the outset he had no political rivals of importance. In fact, the only rival he had known — his nephew Korechika — had virtually self-destructed at about the time Michinaga was coming to power. While Korechika’s father Michitaka was alive, Korechika’s star was in the ascendancy; and he had even temporarily been granted the right of document examination during his father’s illness in 995, when he was only twenty-two years old. He and Michinaga were rivals as young men. The famous incident in 994 recorded in the Ōkagami, in which Michinaga bested Korechika in an archery match, highlights their rivalry.26 And Sanesuke records an incident in 995 when the two of them erupted into a verbal argument so loud that their retainers heard it from the other room.27 Only a few days later, Michinaga’s retainers and those of Korechika’s younger brother Takaie clashed on Shichijō Street.28 And just after that one of Takaie’s retainers killed a follower of Michinaga, and Takaie was forced to surrender the culprit to the authorities.29 But Korechika and Takaie were banished from the capital by their own rash act in 996. Korechika mistakenly believed that Retired Emperor Kazan was a rival for the affections of his own paramour, Senshin no Onkata, a famous beauty of the day. Takaie and his retainers accosted the retired emperor one night and shot an arrow through his sleeve.30 Shooting at a former sovereign was a serious matter; but because of the delicate nature of the affair, no immediate action was taken. On the eleventh of the next month, however, the nobles were ordered to debate the guilt of Takaie and Korechika; but it was late, and they retired with no decision.31 More than two months later, the nobles gathered in council and appointed Korechika the provisional governor-general of Dazaifu and Takaie the provisional governor of Izumo; and an imperial edict was issued announcing their exile. The charges against them were three: the incident involving Retired Emperor Kazan, having cast a curse on Higashi Sanjō’in (Ichijō’s mother), and having carried out a very special Buddhist ceremony (the daigen no hō) forbidden to subjects.32 Thus politically Michinaga was also very lucky. Korechika might have remained a rival, especially after Higashi Sanjō’in passed away. But his fate was sealed; and although his exile was brief, Korechika was never a political factor after his return to the capital, and he died in 1010 while still in his thirties. Michinaga seems to have been uninvolved in the fall of Korechika and behaved with a good deal of restraint throughout the affair. Unlike earlier Fujiwara regents, who plotted to remove rivals, Michinaga did not have to resort to trickery to best his. Indeed, the Ōkagami claims that after Takaie returned to court, Michinaga dis-
Kugyō and Zuryō | 75
cussed the exile with him on a visit to the Kamo Shrine. In the story, Michinaga addresses Takaie: “Everyone says I was the person who proposed your exile and pushed it through....But that wasn’t what happened. Do you believe I could visit Kamo Shrine like this if I had added a word to what His Majesty said?”33 Michinaga may have been disingenuous, but there is no evidence of his having concocted anything here. Once again, luck favored Michinaga. Michinaga’s luck was not unqualified. For example, his health was seriously compromised. He was often ill, and sometimes suffered debilitating effects of the symptoms of diabetes that periodically afflicted him.34 Still, he was able to have his way for much of this time, and his ability to secure the appointment of kinsmen and clients at court and in the provinces was greater than that of any other Heian statesman. Since his era also is that of great cultural florescence, represented by the works of such redoubtable authors as Sei Shōnagon, Murasaki Shikibu, and Akazome emon; calligraphers such as Michinaga’s kinsman Yukinari (Kōzei); and poets such as Fujiwara no Sanekata (and in fact, Michinaga himself), there is a tendency to assume that all was well in the realm. But that would be a misjudgment. There was in fact considerable unrest. A brief perusal of the Nihon kiryaku, for example, confirms this state of affairs. Two years after Michinaga’s birth, Fujiwara no Chitsune rebelled in Shinano Province, presaging a solid record of unrest throughout the provinces. In 974 the protests of residents of Owari led to the replacement of the governor there. And two years later the Imperial Palace was largely destroyed by fire, and a process of removal to temporary palaces began that was to continue throughout the period. In 982 the repaired palace burned again, and armed forces were sent against pirates. In 988 the farmers and officials in Owari presented their well-known petition against the misdeeds of yet another governor, Fujiwara no Motonaga, as discussed in Charlotte von Verschuer’s chapter. In 997, “foreigners” attacked Tsushima and Iki; and in 999 the Awaji governor was dismissed because the local residents complained of his actions. More palace fires occurred in 1001, 1005, 1009, 1014, and 1015, before larger, citywide fires ravaged Kyoto in 1016 and 1024. As we shall see in detail below, provincial unrest continued unabated: complaints lodged against governors are recorded in Settsu in 1003, Owari (again) in 1008, Yamato and Kaga in 1012, and Tanba in 1017. A major invasion of northern Kyushu by the Jurchen pirates in 1019 was followed just over a year later by pirates attacking Satsuma as well (see Bruce Batten’s chapter herein). In 1028, only a year after Michinaga passed away, came the major uprising in several eastern provinces of Taira no Tadatsune, discussed in the chapter by Karl Friday; this uprising caused serious, long-lasting damage to the productive capacity of the region. Besides fires, the capital was also beset by numerous incidents of murder and armed attacks on the street. Breaking and entering appears to have been rampant; targeted were not only the houses of the nobility but also the Imperial Palace and the several temporary palaces to which the emperor moved while the palace was being repaired. Not even Michinaga was immune from this crime wave. Although he simply mentions it in a straightforward account in his diary, Michinaga was robbed of some clothing and kitchen items on two successive eve-
76 | g. cameron hurst iii
nings in 1011.35 More egregious was the theft of two thousand ryō of gold money from his storehouse in 1017.36 Less than a month later, Fujiwara no Sueyoshi, who lived just to the west of Michinaga, was attacked and wounded in his house by an intruder.37 And just two months after that, someone was apprehended breaking into Minister of the Right Kinsue’s stable at night.38 From just this random selection of occurrences, it should be obvious that Michinaga’s era was not particularly peaceful, and perhaps ought not be considered the high point of good government in the Heian era either. It was notable for the lavish lifestyle enjoyed by a small circle of courtiers and their immediate followers in the capital, supported by access to provincial wealth — largely made possible by Michinaga’s appointment of kinsmen, associates, and clients to lucrative provincial posts. How exactly was Michinaga able to dominate politics? Despite McCullough’s statement, Michinaga was not a classic dictator who could crush enemies at will. That was not his style, and besides, there were institutional and familial restraints on his exercise of power. How then did he exercise control? Most important was the crucial familial relationship with the Imperial House upon which the Fujiwara regency was predicated. Although his diary is called Midō kanpaku ki, Michinaga was never chancellor, and he actually served as regent for just over a year. Thus the post itself was not that important. Access to power depended upon familial relations; and as historian Kuramoto Kazuhiro has explained, several factors determined the degree of power that a regent or chancellor could exercise. Kuramoto has constructed a number of informative charts that graphically display the nature of blood and marital ties between successive emperors and regents.39 Essentially, the ability of the regent to exercise power was normally greater when both sets of ties were close. But Kuramoto has also factored in other elements of the immediate familial situation of the emperor, such as whether the emperor was an adult with political abilities or interests, whether either his father or mother or both were still alive, whether there was another kugyō with close marital relations, and so forth. Basing his construct on those factors, he creates sixteen different models of the power structure over several periods in the Heian era. Expanding our analysis in this way allows us to discern several patterns in Michinaga’s career as major figure at court. For example, from 995, when he was named document examiner at the death of Michikane, until Ichijō’s death in 1011, Michinaga held that position and was senior noble (ichi no kami) as minister of the left. But his access to power changed depending on the configuration of the Imperial House. As I have noted elsewhere40 but Kuramoto has explained in far greater detail, it was possible for there to be parental authority constraining the emperor from both the imperial family side — usually in the form of a senior retired emperor (in), but often from the emperor’s mother, especially if she had been given an in designation — as well as the Fujiwara Regents House. On occasion, there could be animosity between a retired emperor and a regent, but normally there was a tendency to work in concert.
Kugyō and Zuryō | 77
From that point of view, Michinaga’s situation differed quite a bit from that of his father, Kaneie, and his brother Michitaka. When Ichijō came to the throne, Kaneie was first regent during his minority, and then became chancellor after his genpuku (coming-of-age) ceremony in 990. He was father of En’yū’s principal consort, Senshi, and thus grandfather of the young Ichijō (see figure 4.1). En’yū, however, was still alive and exercised some degree of parental authority, so that Kaneie as regent and En’yū apparently shared power on the basis of considerable consensus while Ichijō was a youth.41 When Michitaka took over from his father, he faced an adult Ichijō beginning to assert his will in politics. En’yū still exercised a degree of parental authority, and Michitaka was both his uncle and after his daughter Teishi entered the palace, his father-in-law. En’yū was ill for much of the time, so Michitaka appears to have been the most powerful figure. But once En’yū died, Michitaka apparently enjoyed even more power because Senshi, his younger sister, was the ranking member of the Imperial House with the title Higashi Sanjō’in (see Fukutō Sanae’s chapter herein). When Michinaga, who was Senshi’s younger brother, took over, Senshi herself became ever more active in court politics. As mother of the nation (kokumo) and ranking member of the Imperial House, she now began to assert her will in the political realm.42 Indeed, it was supposedly at her instigation that Michinaga was given the designation as document examiner in the first place.43 There appears to have been a reasonable political consensus among the three — Senshi, Ichijō, and Michinaga — in determining state policies.44 Although both retired emperors, Kazan and Reizei, were still alive, neither was closely enough related to Ichijō to exercise any parental guidance, and neither seems to have intruded in politics whatsoever. When Higashi Sanjō’in passed away in 1001, Michinaga and his nephew Ichijō formed the nexus of political power; and their relationship was sufficiently close — Teishi had also died, and Michinaga’s daughter Shōshi became Ichijō’s sole surviving consort — to allow Michinaga to control the direction of state affairs. But despite that close relationship with Ichijō, how did Michinaga control politics? He was not regent and thus could not sign imperial documents, nor did he have influence that a formal chancellor enjoyed. He was a co-participant in ruling through the right of document examination, but that seems to have been mostly a formality. Received wisdom for many years, including my own writings, has tended to emphasize that not only during the Fujiwara regency period, but in the insei era as well, court decisions were largely made in the kugyō (senior nobles) council, most frequently in the jin no sadame process.45 The Saigūki, Hokuzanshō, and other works of court procedure explain the process as follows. The senior ranking noble present (called the shōkei), in accord with the emperor’s instructions, has one of the secretaries (geki) inform the nobles to assemble. Once they have assembled, the senior noble transmits the matter to be decided. Then, beginning with the lowest-ranking member of the group, each noble states his opinion. The senior noble has one of the imperial advisers (sangi, the junior-most among the members) draft a decision (sadamebumi) summarizing the opinions,
78 | g. cameron hurst iii
which the head chamberlain (kurōdo no tō) reports to the emperor. Then, a decision ought to be proclaimed in an imperial edit or edict of the Grand Council of State. There are, however, problems in evaluating the validity of the process. For one thing, most documentary references to council decisions, such as the jin no sadame, are greatly abbreviated. They do not describe the process or the differences of opinion, but simply report that “the council decided on several matters submitted from the provinces” or the like. Just exactly how the decision was made remains, in other words, unclear. We know that in some cases, council meetings were canceled because so few nobles appeared, or that a decision was made by simply the few in attendance, suggesting that the kugyō themselves may not have regarded council meetings as the most important of state affairs. There are also references to matters that were decided at the private residence of the regent and to appointments and other matters of import that were decided by the regent alone, or by the emperor, or by the two in tandem.46 In fact, for those reasons, scholars have begun to question the importance of this noble council form of decision making, suggesting that such decisions were often no more than an ex post facto approval of a decision that had already been made by the core power holders, normally emperor and regent. The only historian to study these types of decisions in detail is Kuramoto, who has examined all the decisions he could identify over the twenty-five years of Ichijō’s reign.47 He has identified 388, which cover more than eighty separate categories of issues, major and minor, ceremonial and substantive. But of the 388, there are only 14 cases for which we actually have the whole picture before us: who assembled, what their various opinions were, and what action the court actually took subsequent to the council meeting. Let us examine a few council meetings in the diaries of Michinaga, Sanesuke, and Yukinari to see if we can determine the importance of the meetings, the decisions they reached, their role in shaping official actions, and perhaps Michinaga’s role. Recall that during Ichijō’s reign, Michinaga was neither regent nor chancellor, but only had the right of document examination prior to their submission to the sovereign. This may have actually worked in his favor, since the sekkan, as extracodal officers, did not sit at the council meetings. But Michinaga did, as minister of the left, attend them, unless illness or something else otherwise prevented him. Although there are many references to meetings in Michinaga’s diary, they are usually brief, stating that a meeting took place or simply noting the decision. Other courtiers, however, often record the process, the outcome, and sometimes the role of Michinaga himself. One decision about which we have considerable internal information is the one in 997 to pardon Fujiwara no Korechika, Michinaga’s nephew and former rival, and Takaie, after their exile, discussed above.48 In the third month, a general amnesty was declared in order to alleviate the illness of Imperial Dowager Higashi Sanjō’in. The courtiers debated whether Korechika and Takaie should fall under this amnesty. As Sanesuke describes the process, Emperor Ichijō summoned Michinaga, as senior noble, and instructed him to call a council meeting, which
Kugyō and Zuryō | 79
he did, presenting to the nobles the sovereign’s instructions: they should decide (sadamemōsu) whether the amnesty ought to apply to the two men, whether they ought to be recalled to the capital, or even if they were pardoned, whether they ought to remain in their current posts. Kugyō opinions varied widely. Two other courtiers agreed with Narinobu that Korechika and Takaie should be pardoned but that the court ought to seek the opinion of legal specialists regarding their return to the capital. Kinsue and Koretada also thought that they should be pardoned but that precedent ought to be consulted regarding their recall. The dominant opinion, offered first by Minamoto no Toshikata and joined by three more members, was that Korechika and Takaie ought to be pardoned but that their recall should be left to the emperor’s decision. (Sanesuke himself favored that position.) Fujiwara no Sukeyoshi suggested the option of a pardon but having them remain in their current posts. Michinaga expressed no opinion, but he reported the outcome himself to Ichijō (according to Sanesuke, Michinaga simply “committed the purport of the opinions to memory” and reported them). Michinaga reported back to the nobles the emperor’s decision that if there were past examples of exiles having been brought back at the will of the sovereign, then they should be recalled. Michinaga then called a secretary and ordered him to have messengers dispatched to recall the exiles. In this case, the nobles’ opinions were diverse; there were four different opinions, with the “votes” split 4, 3, 2, and 1. Thus, there was no consensus on what action to take, other than that the two men ought to be pardoned. But the emperor, after discussion with Michinaga, seems both to have pardoned and recalled them. Takaie was even reported to have arrived back in the capital on the night of the twenty-first.49 Kuramoto reads much into the series of actions involved here. First, by ordering the nobles to decide one of three specific courses of action, the emperor seems to have had considerable influence in shaping the decisions of the noble council. Second, the emperor ordered the action taken on the basis of a consensus among Ichijō, his mother, and Michinaga. Thus, the fact that Michinaga did not offer a formal opinion on the issue, and in fact did not even present a written decision of the nobles’ opinions to the emperor but delivered it orally from memory, suggests that the decision had already been made elsewhere and that the jin no sadame in this case was no more than a procedural formality.50 Michinaga’s behavior in the matter was unusual, at least if we accept the description of proper procedure in such sources as Hokuzanshō. Not only did Michinaga not report the nobles’ decision in written form — only “unimportant matters” were to be transmitted to the sovereign orally — but the reporting itself should have been done by a head chamberlain. Another jin no sadame for which there is some detail is one in 1004 involving Usa Hachiman Shrine, the major shrine in Kyushu and perhaps second only to Ise Shrine in national importance. In this case, the shrine had been protesting the harsh rule of the senior assistant governor-general of Dazaifu, Taira no Korenaka. The nobles were to debate whether to send a special envoy to investigate charges against Korenaka.51 Their decision was to send a messenger, and when
80 | g. cameron hurst iii
Michinaga duly reported it to the emperor, Ichijō asked Michinaga’s opinion. Michinaga suggested that sending an envoy was fine but that they first needed to select one and then reply to the shrine by imperial edict. The two of them — a head chamberlain was also in attendance — selected the envoy without any input from the nobles, and the edict, with the envoy’s name, was duly sent. Kuramoto therefore argues that the noble council did not really make the final decision and that Michinaga’s ability to get his way was less as leader of the council than in his capacity as document examiner and adviser to the emperor.52 Another well-documented case dealt with foreign trade. A Nihon kiryaku notation for the second month of 1005 states simply that Dazaifu reported the arrival of Song merchant Zeng Lingwen.53 Michinaga’s diary is perfunctory as well. He arrived at midday at the palace, where there was a recitation of the Sūtra of the Golden Light attended by most senior nobles; the most interesting notation, it would appear, is that Fujiwara no Tametomo presented him with two horses,54 a frequent occurrence, as we shall see. Michinaga takes no note of the Chinese trader Zeng, but he was mentioned in the Nihon kiryaku before, and thus the compilers deemed his arrival noteworthy. As is often the case, Michinaga’s kinsman and sometime critic, Fujiwara no Sanesuke, does go into the incident in some detail.55 Apparently, Zeng’s arrival occasioned considerable debate among the kugyō because he violated rules governing Chinese traders decreed in an edict a year earlier. The debate was resolved in favor of Michinaga, who displayed a tactic of handling the noble council that appears to have been common for him — and to have provoked Sanesuke. Most of the entry for the day refers to the Buddhist sūtra reading held at the palace, with a long discussion of what Buddhist images were placed where and who the various notable prelates in attendance were. Sanesuke records that at the end of the discussion Michinaga raised the matter of the arrival of the Song merchant (no name given) and requested a decision on the matter, but it was postponed because of the lateness of the hour. On the twenty-first the nobles gathered in council at the Office of the Palace Guards to decide how to deal with Zeng. The two options presented to the nobles appear to have been to allow him to stay or to send him back immediately. Sanesuke argued that because Zeng came earlier than the date set in the imperial edict, he should be sent back to China. And the ten courtiers besides Michinaga all concurred that he ought to be sent back. Michinaga asked if Zeng really ought to be sent back immediately simply because he had come earlier than stipulated. Perhaps, he suggested, Zeng ought to be allowed to remain in Japan while awaiting a favorable wind. In which case, Michinaga argued, was it necessary to have imperial permission? Two of the group agreed with Michinaga, but the rest still felt that Zeng ought to be sent back. Three days later, an imperial edict was issued allowing Zeng to remain. Emperor Ichijō was in agreement with Michinaga. Kuramoto notes that ever since the arrival of the message from Kyushu, Michinaga had visited the palace on the fourteenth and fifteenth, the eighteenth through the twenty-first, and the day of the edict itself. Although there is no record as to what was discussed between
Kugyō and Zuryō | 81
Michinaga and the emperor, he reads this as Michinaga’s and the emperor’s arriving at a “consensus” as to the action taken.56 At any rate, it seems clear that Michinaga and Ichijō were able to have their way, despite the views of the council. Kuramoto goes so far as to suggest that the two had probably arrived at a decision before the meeting: thus when Michinaga brought up an alternate solution to the matter, a few courtiers switched their opinion, most likely because they assumed that Michinaga was going to get his way in the end anyway.57 We will see later why Michinaga might show favorable treatment to a Chinese trader. That same month the noble council faced another interesting issue, one that struck at the heart of the whole complex of ideology involving imperial rule. It captures nicely the degree of disorder in the realm, and it also illustrates well how decisions appear to have been made during Michinaga’s regency. One night near the end of 1005, another terrible fire consumed much of the palace.58 It had been an eventful day at the palace for Michinaga; he and the other kugyō had begun the day by debating the guilt of the Dazaifu official Nagamine no Tadanori, who was the subject of a complaint by Usa Hachiman Shrine. There were archery matches and much other business to dispense with. Michinaga returned home, viewed the lunar eclipse that evening, then retired. But he was awakened in the middle of the night with news of a fire to the west, the location of which he soon learned was the palace. Rushing there, he met several other courtiers. They determined that both the emperor and Michinaga’s daughter Shōshi were safe. Michinaga felt that they ought to protect the Kashikodokoro, where the sacred mirror was kept, but they were too late. The fire had destroyed the building. When they searched the ruins of the building, they found the mirror, or what was left of it, among the ashes atop a roof tile. The various accounts refer to it being “burned up” or having lost its “circular shape.” Exactly what shape it was in is hard to glean from the sources, but it must have melted and congealed into a blob of bronze no longer resembling its original form. This was of course the famous Yata no kagami, one of the three Imperial Regalia symbolic of the legitimate sovereign. Recounted in the Kojiki as having been passed on to Ninigi no mikoto by Amaterasu, the regalia were the most sacred emblems of state, and the “loss” of the mirror was as grave a matter as the courtiers could face. What to do? The next day Michinaga told Sanesuke to call the kugyō together to decide what was to be done. When Sanesuke arrived at Michinaga’s, he found a number of other courtiers discussing the matter and seemingly having reached a decision that another mirror ought to be cast. Then on the seventeenth, they assembled for a formal jin no sadame. The topic of discussion was announced as “whether the mirror ought to be recast.” Sanesuke records his opinion that they consult various specialists (shodō, presumably meaning Shinto, Buddhist, and Taoist adepts and legal scholars) before taking any action. He felt that they ought not take regular bronze and forge it with the remains of the original mirror, which was of course considered sacred. (The unspoken fear is that such an action might seriously impair the sacred quality of the mirror.) If they wanted to preserve the shape of the mirror, he argued, they ought to forge a new one and enshrine it
82 | g. cameron hurst iii
along with the damaged original. Still, he said, they ought to consult everyone, and certainly do nothing before informing Ise Shrine; and the recasting ought to be done only after proper divination. The other courtiers agreed. Michinaga had that decision relayed verbally to Ichijō, who in turn ordered the consultations. More than half a year later the opinions came in. Michinaga and Yukinari both record the gozen no sadame — a council meeting in front of the emperor — that was called for the third day of the seventh month of 1006.59 The emperor had Michinaga, as senior noble, announce the topic to be decided. In turn, Michinaga had Yukinari read the various opinions of legal scholars and yin-yang specialists. Based upon their recommendations, Yukinari offered the opinion that the mirror ought not to be recast. Although it no longer retained its original shape, he argued, it was still a sacred item, and when enshrined, it would retain its sacred splendor. Several courtiers agreed, until Kinsue offered a different opinion: that tortoise-shell divinations be conducted and shamans consulted, and based on the outcome, another meeting called to reach a decision. Korechika and Michinaga joined in that view. The emperor was displeased that there was no consensus and ordered them to debate the issue again. A second discussion yielded no change in the opinions of the two groups, so that it remained seven in favor of not recasting the mirror and three wanting further clarification from divine sources. Michinaga opined that the matter was difficult to decide, and the courtiers departed the meeting. Yet about ten days later, a divination was ordered,60 in accord with the minority opinion favored by Michinaga. No further notations are found, and it appears that no divination altered the decision to cast a new mirror and enshrine both of them. Finally, in 1006, there is an interesting decision, recorded in some detail in both Gonki and Midō kanpaku ki, dealing with the appointment of the new abbot (bettō) at Tōdaiji.61 When Saishin resigned the position, a council meeting was held; Ichijō told Michinaga to have the council decide a successor. Two monks were recommended: Chōshin (933 – 1014) and Seiju (959 – 1016). When he got the report, the emperor had the message relayed back to the council to choose one of the two. The nobles were split, with six choosing Seiju, three, Chōshin. Michinaga, noting that both were qualified men, suggested that messengers be sent to their respective temples to find out who was the more competent and then appoint that one. It was this lone opinion that was reported to Ichijō, who ordered the dispatch of the messengers in accord with Michinaga’s advice. Almost three months later, Chōshin, who was apparently a close associate of Michinaga, was appointed to the position.62 It appears that even though the nobles as a whole favored Seiju, Michinaga engineered the appointment of Chōshin, suggesting once again that Michinaga enjoyed power that transcended that of the kugyō as a deliberative body. More examples could be provided, but suffice it to say that careful consideration of extant council decisions leads one to question the centrality of this process. It seems that Michinaga was able to get his way because the core authority figures — emperor and regent — were in fact not bound by the opinions expressed in the council. The court seems to have valued the meetings as an important
Kugyō and Zuryō | 83
format for expressing the individual and collective opinions of the nobility. But it seems obvious that discussions between Ichijō and Michinaga permitted them to structure issues to preclude unwanted decisions. Moreover, there were apparently informal discussions between Michinaga and some of the courtiers before the meetings were held, perhaps the functional equivalent of what the Japanese today call nemawashi (informal consultations). Moreover, for most of the era the noble council was composed of close kinsmen of Michinaga from the Fujiwara clan, including his own sons (Yorimichi became a kugyō in 1006 at age fifteen); other Fujiwara clan members of different lineages but often with close ties to Michinaga, such as Yukinari; and Minamoto clan members related by marriage. Thus at the center in Heian, Fujiwara no Michinaga enjoyed extensive personal ties with the Imperial House that allowed him control over what we may call the private, authoritative aspects of imperial rule. At the same time he dominated the bureaucratic structure that represented the public aspects of the state. In the latter sphere, there were few real opponents likely to propose policies or decisions inimical to Michinaga’s interests, and in the former sphere he and Ichijō saw eye to eye on most issues. He had only minor problems during the short reign of his cousin Sanjō (both his father, Reizei, and his mother, Michinaga’s elder sister Chōshi, were dead and unable to exercise any parental authority). And of course, when his grandson Go-Ichijō came to the throne at age ten, Michinaga was in total control.
Michinaga and the Periphery Provincial Governorships: The Appointment Process Having seen how thoroughly Michinaga was able to orchestrate the organs of the central government to his and his family’s advantage, we turn now to see how he controlled the periphery, more precisely how he was able to guarantee the flow of resources from the countryside to the capital. As noted at the beginning of this chapter, the ōchō kokka was predicated upon a major restructuring of the way the provinces were governed. First, the central government relaxed its micromanagement of the provinces. The state did not totally neglect the provinces; but governors were given almost blanket authority to administer their provinces, in return for an agreed-upon tax remittance. Second, there was a shift from the imposition of taxes on persons to units of land called myō. The ritsuryō allotment system (kubunden) was predicated upon centrally “owned” land that was allotted to families according to size and gender differentiation and taxed accordingly. As the size of families rose or fell, the taxes and tribute levied upon the families were supposed to rise or fall accordingly. But the process of periodic census readjustment was unmanageable, and the populace found various ways to evade the system. Thus from the tenth century various holdings were lumped together into myō, a prominent cultivator was assigned to represent the unit, and taxes were assessed on that unit. The system was far from perfect because the central government relied upon provincial registers (kokuzu) produced in the tenth century
84 | g. cameron hurst iii
to calculate the taxable lands in each province, and the provincial governor was responsible for producing a tax quota based upon that assessment. The taxable land in each province thus became fixed, and the governor was responsible for meeting the tax quota assigned to his province. Provinces were divided into four categories according to productivity, and they ranged widely — from Mutsu, with over 51,000 tan of taxable lands and Hitachi calculated at slightly over 40,000, all the way down to Izu, which registered only a bit over 2,110 tan. (The islands of Iki, Tsushima, and Oki had the smallest amount of taxable lands, about 400 – 600 tan, and the island province of Awaji ranked just above Izu at 2,650.) The size of provinces, or more accurately, the rankings according to the productive capacities of these provinces, mattered greatly to the Heian courtiers. Applicants for governorships sought the most prestigious, or wealthiest, and they could be devastated when they found themselves appointed to “minor” provinces.63 One well-known story from the Kojidan, for example, involves Fujiwara no Tametoki, the father of Murasaki Shikibu, who in 996, after ten years without an official posting, was appointed governor of Awaji.64 Disappointed because he had sought appointment to the then open and far more prestigious Echizen, the despondent Tametoki composed a Chinese poem expressing his emotion and had a court lady show it to Emperor Ichijō. A scholar of some repute, Tametoki so touched the emperor with his verse (“Bitter study on winter nights brought blood-red tears to soak my sleeves; but in the spring morning on Appointments Day my hopes were high in the blue heaven”65) that the young sovereign even shed tears over Tametoki’s fate. When Michinaga learned what was troubling Ichijō, he took decisive action, even though the new appointments had already been posted. He summoned Minamoto no Kunimori, who had already celebrated his good fortune at having been made governor of Echizen, and had him write a letter of resignation. Michinaga then appointed Tametoki to the post, and he and his daughter set off for Echizen later that year. Poor Kunimori fell ill; and although he was appointed governor of Harima in the fall round of appointments in the same year, he died without being able to take up the post.66 True or not, the story underscores the decisive influence that Michinaga could have in the appointments process, an influence that goes a long way to explaining the nature of the relationship he enjoyed with provincial governors. That Michinaga personally intervened to overturn the appointments, after the kugyō had their deliberations and Michinaga and Ichijō had issued the final postings, is fully in line with the argument developed above, that power lay with the regent (or in Michinaga’s case document examiner) and emperor, with the opinions of the noble council considered nonbinding. In fact, it is probably true that decisions regarding appointments, especially those for governorships, were more important than others as far as Michinaga was concerned inasmuch as it was through them that he maintained control over center and periphery alike. At any rate, the court had an expectation of income based upon the assessed value of paddy fields in the provinces; and the governors served as tax collectors, answering to the senior nobility in Heian. Although the technical term for
Kugyō and Zuryō | 85
governor was still kami, as in Echizen no kami, the post Tametoki sought, the generic term for governors during this period was zuryō (custodian), its widespread adoption attesting to the primary role now played by provincial governors. Zuryō is a nominal form of what was originally a verb, in fact part of a couplet of verbs that described the process of discharging a retiring governor. That is, under the ritsuryō stipulations when a new governor was appointed, he was to proceed to the province, where he met with the departing governor in an official banquet at the border of the province. The former governor turned over the documents related to provincial affairs, most important an accounting of tax receipts and the tax-grain fund held in provincial granaries. This process involved the transfer (bunsuke) of the “stuff” of office — the provincial land register, storehouse and granary keys, the provincial governor’s and other official seals, a registry of monks and nuns, and other documents67 — and its receipt into custody of the new governor. Thus the term originally referred to the new governor’s taking actual custody of what was needed to carry out his job and was applied specifically to those who actually journeyed to the province to serve there. This is the important meaning that zuryō maintained into the mid-Heian period: it referred to resident governors (in “custody”) in contradistinction to absentee governors, who might send a deputy in their place. Commonly, such people were called yōnin governors — literally, “appointed at a distance” — that is, they held important official posts at court and thus served as governors only as concurrent (ken) appointments. Thus it was possible that there might be two governors for a particular province, a zuryō actually serving in the province and another, also middle-ranking courtier, perhaps one appointed in the Chamberlain’s Office, resident in the capital and serving concurrently (for income purposes) as governor. Thus zuryō means “custodial governor” in the sense that such a governor journeyed to the province and took custody of the articles of office and served there — even though, of course, he might make regular trips between Heian-kyō and the province of appointment, depending upon its proximity to the capital.68 The competition for office and rank was fierce at the Heian court, as one’s entire social, political, and economic life depended upon the place one occupied in this highly stratified society. Although many offices at the lower levels of the bureaucracy had seen their emoluments and stipends drastically reduced or even in some cases eliminated, higher nobles continued to receive theirs. At any rate, the only hope for furthering one’s fortunes lay in securing appointment to office. The excitement and apprehension as appointments approached, as well as the often desperate nature of the measures taken to obtain posts, especially lucrative governorships, was captured in the writings of court ladies of the time, even if in their pampered existences they minimized the importance of the outcome to the participants. This is strange, given the fact that many, including the author below, were daughters of courtiers who themselves lived and died by the economic assets derived from their official appointments as governors. Sei Shōnagon’s father, Kiyowara no Motosuke, was governor of several provinces, including Suo, to which post she accompanied him in 971 when she must have
86 | g. cameron hurst iii
been about ten years old. Below, Sei Shōnagon comments on an upcoming round of official appointments: It is fascinating to see what happens during the period of appointments. However snowy and icy it may be candidates of the fourth and fifth ranks come to the palace with their official requests. Those who are still young and merry seem full of confidence. For those candidates who are old and white-haired things do not go so smoothly. Such men have to apply for help from people with influence at Court; some of them even visit ladies-in-waiting in their quarters and go to great lengths in pointing out their own merits. If young women happen to be present, they are greatly amused. As soon as the candidates have left, they mimic and deride them — something that the old men cannot possibly suspect as they scurry from one part of the Palace to another, begging everyone, ‘Please present my petition favorably to the emperor’ and ‘Pray inform Her majesty about me.’ It is not so bad if they finally succeed, but it really is rather pathetic when all their efforts prove in vain.69 Again, when she is cataloging “depressing things,” she notes with somewhat greater compassion, Most depressing is the household of some hopeful candidate who fails to receive a post during the period of official appointments. Hearing that the gentleman was bound to be successful, several people have gathered in his house for the occasion; among them are a number of retainers who served him in the past but who since have either been engaged elsewhere or moved to some remote province. Now they are all eager to accompany their former master on his visit to the shrines and temples, and their carriages pass to and fro in the courtyard. Indoors there is great commotion as the hangers-on help themselves to food and drink. Yet the dawn of the last day of the appointments arrives and still no one has knocked at the gate. The people in the house are nervous and prick up their ears....‘Tell us,’ they say, ‘what appointment did His Excellency receive?’ ‘Indeed,’ murmur the servants, ‘His Excellency was governor of such-and-such a province.’ Everyone was counting on his receiving a new appointment, and is desolated by this failure.70 On yet another occasion, the lady shows much greater compassion for those who actually fared well in the process, and she tells us that I enjoy watching the officials when they come to thank the Emperor for their new appointments. As they stand facing His majesty with their batons in their hands, the trains of their robes trail along the floor. Then they make obeisance and begin their ceremonial movements with great animation.71 Doubtless upon occasion some of the newly appointed officials were those very “white-haired” ones that the ladies-in-waiting had so ridiculed, rushing about, leaving no stone unturned to attain their goal. The selection of governors was part of the larger official appointment process, jimoku no gi, with one round held in the spring — always in the first month — and
Kugyō and Zuryō | 87
another late in the fall. There could also be special appointments at other times during the year. In the diaries of courtiers such as Michinaga, the process of selecting governors is referred to as zuryō kōka no sadame, “decisions on the merits of provincial governors,” because for the most part, it required the nobles to assess the records of those currently appointed governors before reappointing them or transferring them to a new province. Applicants for an initial provincial appointment had their previous bureaucratic record closely scrutinized, with special attention to seniority; but apparently they were at a disadvantage vis-àvis applicants with prior experience as governors. As noted in The Pillow Book, applicants for positions were required by midHeian times to submit “official requests.” This is Ivan Morris’ translation of the term mōshibumi, an official petition in which an applicant stated his desire to be appointed to a vacant position, usually specifying exactly what post he sought. Everyone at court knew what offices were vacant at each round of appointments, both by word of mouth and because of the creation of a record called the ōma or ōmagaki (great space), so called because the various posts were listed on a large sheet of paper with blank spaces after those that were vacant so that the newly appointed officials’ names could be written in. Individual nobles, on their copies, would write in their nominations (zuryō no kyo) for the vacant posts, which were then considered and debated. The selections were then presented to emperor and regent for final decision, after which they were announced. Petitions of courtiers seeking official positions are filled with phrases of frustration over the inability to obtain appointment. Tachibana no Naomoto, then professor of literature (monjō no hakase) and submitting a mōshibumi for a concurrent appointment in the Ministry of Civil Affairs, lists case after case of similar concurrent appointments in the past, but complains that men his junior have passed him by to enjoy far greater imperial favor.72 There are more than a dozen extant petitions seeking new provincial appointments, reappointment to the same post, or transfer to another one. Most are artificially elegant, dripping with hoary Chinese phrases trying to catch the attention of the nobles sitting in judgment; and often a petition might be polished by a courtier more skilled in Chinese composition than the submitter himself. Historian Murai Yasuhiko divides these petitions into two types, which might be rendered the pitiful and the boastful.73 The former, including that from Naomoto, seek favorable action by appealing to the sentiments of the nobles. Several lament unrewarded but diligent service at court while others received lucrative provincial appointments. Ōe no Masahiro, for example, laments that his dilapidated house no longer keeps out the wind and rain and that his aged mother is in pitiable condition. Listing relevant precedents of men in his post receiving concurrent governorships, Masahiro asks that in consideration for past service, he be made governor of either Echizen or Owari.74 Such appeals obviously worked. In 980 Sugawara no Fumitoki, provisional governor of Owari, petitions for appointment to the junior third rank, which would represent entry into the kugyō.75 He lists his long years of service as Confucian scholar-official and points to several men his junior who have been appointed before him. Recalling his illustrious
88 | g. cameron hurst iii
grandfather Michizane, Fumitoki notes his advanced age (he was eighty-two at the time); although little time was left to him, he wistfully observes that it would be splendid if he could be raised to the third rank. Sure enough, the Kugyō bunin notes that next year he was granted his wish and became senior noble — just eight months before his death.76 The second type is more positive, the applicants choosing instead to boast of their accomplishments in past or current posts to make their case for appointment. Perhaps the quintessential boastful petition is that of former Mino Governor Minamoto no Tamenori, who was seeking appointment as governor of Mino, Kaga, or any other vacant province in 1014.77 Favorably comparing himself to all other eight governors appointed at the same time as himself, he claims that when he took over his former Province, it was in total decline. Whereas the former governor could only hand over to him 1,200-plus tan in rice fields, Tamenori boasts that his administration expanded that amount to more than 3,500 tan. The extant petitions do not break down so readily into the two types; many mix descriptions of the woeful state of the applicant at the moment while stressing his stellar service in the past.78 Michinaga and Zuryō This sojourn into the appointment process barely scratches the surface of a complex process, involving officials from several bureaus, a large number of documents, boxes in which the petitions were kept, and the like. As we have seen from literary works, the tension was high, the celebrations for success were triumphant in nature, and the dejections at failure deeply felt. The competition was also fierce. Of course, the number of vacant posts varied annually, depending upon the number of governors completing their term or ones who died while in office, or others who may have been dismissed, a not uncommon occurrence. Just to pick one year for which the ōmagaki remains, in 996 there were only ten provincial governorships open, and there appear to have been perhaps one hundred qualified applicants.79 Michinaga, as the highest-ranking kugyō and co-participant in the emperorship of three successive sovereigns, was responsible for most decisions at court for the three decades of his career, including those involving the appointment of provincial governors. What was Michinaga’s attitude toward governors, their appointment, and their behavior? To state my conclusion at the outset, it was passive, reactive, and ultimately supportive of governors, no matter their behavior. The mutual interdependence of Michinaga and the zuryō seems to have outweighed all other considerations. First, there was the economic consideration on both sides. Since bureaucratic emoluments continued to decline over the Heian period (although kugyō such as Michinaga continued to receive their income regularly), appointment to office with the possibility for economic gain was most eagerly sought. Thus on virtually every appointment involving provincial positions, activity around the houses of nobles such as Michinaga and Sanesuke was intense, with the comings and goings of governors or would-be governors. It was common for the visitors
Kugyō and Zuryō | 89
to bestow presents (kokorozashi) on the nobility as tokens of appreciation of favors rendered or expected. A reading of the major diaries reveals that the most commonly mentioned present was horses. For example, in the tenth month of 1016, the first year of Michinaga’s grandson Go-Ichijō’s reign, several governors visited Michinaga’s Tsuchimikado residence. Minamoto no Tadataka, governor of Suruga, arrived on the fifth with ten horses.80 On the nineteenth, Hitachi ViceGovernor Taira no Koretoki brought forty horses for Michinaga, twenty male and twenty female.81 Michinaga received ten more horses on the twenty-second, courtesy of Mutsu Governor Fujiwara no Sadanaka.82 Two days later the governor of Kaga brought two more horses, and Michinaga remarks frankly that lately people tend to bring horses as kokorozashi.83 The parade continues in the next month, with Chinjufu Shōgun Taira no Koreyoshi bringing five horses on the sixth and Izu Governor Sukenori four more on the ninth.84 Thus, in just two months we know that Michinaga received at least sixty-nine horses as a form of tribute. Other than the brief mention of the frequency of horses as kokorozashi, Michinaga makes no comments on why these horses ought to be donated to him. There was a minor round of provincial appointments on the sixth, and then a major round of appointments from the twenty-third through the twenty-sixth of the eleventh month, so it could well have been in anticipation of favorable consideration.85 From just the extant entries in Midō kanpaku ki alone, Murai Yasuhiko has identified 301 horses and 29 cows donated to Michinaga.86 Since there are many missing entries in the diary, it is likely that he received far more than this. In fact, Murai missed the five horses brought by Koreyoshi referred to above, bringing the total to at least 306. One wonders what Michinaga did with all these animals. He owned several large properties, and there were stables within his mansions. But the numbers are staggering: sixty-nine in one month is a substantial herd to be kept in Kyoto. One is led inevitably into contemplation of the practicalities of stabling and cleaning up after so many animals. However, it seems that “recycling” was an established practice. Michinaga donated some of these horses to his nephew Takaie, and he also notes that lately people have taken to requesting horses from him.87 There are numerous other references in his diary to Michinaga’s offering horses as gifts, so it appears that, perhaps in much the same way that people today recycle New Year’s and summer presents, Michinaga and other kugyō redistributed any number of those horses, and other goods, that they received as gifts from those seeking or returning favor.88 The exchange of such gifts appears to have been an important social and political lubricant of the ōchō kokka. Zuryō contributed in several ways to the lives of the higher nobility. First, as essentially tax managers, they forwarded to Kyoto the basic tax receipts that formed the core of nobles’ stipends, allotted on the basis of both rank and office. This was their primary function. Second, they presented various kinds of payments. One form was the above-mentioned kokorozashi, or special gifts in anticipation of or in thanks for favors rendered, usually assistance in the appointment process. Another was a more direct shipment of goods from the provinces once appointed to various nobles, usually brought by messengers while the governor
90 | g. cameron hurst iii
was still in province. (See, for example, the chart Murai has constructed on the various goods received from zuryō by Sanesuke just in the extant months of his diary for 1025.)89 By Michinaga’s time, however, the zuryō were relied upon to do far more than forward their regular tax receipts. The various fires that racked the city of Kyoto, either accidental or the result of arson, required a constant round of building and rebuilding. The rebuilding, public and private, was handled in several ways. One was simply to allot the construction, especially of the Imperial Palace, to various governors, who would then assign the work to men from their province. After the major fire in the fall of 1005, for example, the nobles met to allot the various construction tasks to governors — the governor of Ōmi to construct the Bifuku Gate, the Tanba governor to build the Burakuin, et cetera — at the direction of Michinaga.90 Sanesuke records his displeasure that the cost of all this reconstruction is ruining the provinces. Such reconstruction was considered a normal public function of governors when disasters befell the state. But there was a “private” aspect to this construction as well, through the practice of jōgō, by which wealthy governors made private contributions to the construction of public buildings. It was an institutionalized form of purchase of office, which was common in the Heian period, becoming ubiquitous in the insei era. In the discussion above, Sanesuke records in his diary that the Harima Governor Fujiwara no Munemasa had submitted a petition in which he agreed to rebuild the Jōneiden and Sen’yōden in return for being reappointed governor of Harima. The nobles decided the issue, and an edict was issued confirming Munemasa’s reappointment.91 Michinaga records that there was considerable disagreement over whether to approve Munemasa’s reward of extended appointment inasmuch as the actual work had yet to completed. But they agreed, because it was deemed to be a public benefit (kōeki).92 The term is not used here, but it is common to read in public and private documents of the time reference to governors being appointed, or having their appointments extended, for having made “special contributions” (bekkō) as in the example above. In 1010 a round of appointments was made late in the third month, and Michinaga lists many of the governors who were appointed. Two governors had died while in office, and so their places were now filled — the Owari governorship by a man Michinaga called the “biggest office buyer [in the land]” (jōgō daiichi no mono nari).93 This type of reward for special contributions appears to have been well accepted by the nobles at the time, but when it came to such contributions and levies for more private purposes, then there was disagreement. In the reconstruction of his Tsuchimikado mansion, for example, Michinaga allocated the construction and furnishings to different provincial governors,94 and the governor of Iyo, Minamoto Yorimitsu, according to Eiga monogatari, “provided the interior furnishings for the entire establishment, supplying everything that could possibly be needed by the three personages — to say nothing of blinds, mats, jugs, basins, and other furnishings for the ladies’ apartments, and equipment for the offices occupied by retainers, chamberlains, and escorts.”95 Sanesuke confirms
Kugyō and Zuryō | 91
virtually everything claimed, noting that this allocation among zuryō was “unheard of,” and he is critical of the excess of opulence. He carries on at great length about the contributions made by Yorimitsu — “the rarest of the rare” — noting that citizens are reputedly copying down the list of goods donated by Yorimitsu. Once again, he repeats his shock that the sum total of Yorimitsu’s contributions was “unheard of.”96 The mutually dependent relationship of the kugyō and the zuryō is obvious from the above. The court depended upon the zuryō for their livelihood — taxes, construction projects, and contributions both public and private — while the governors themselves relied upon the nobles, especially the regent, for appointments to the posts that guaranteed access to those resources. Both parties seemed to appreciate this dependency. For the nobility, however, there was the problem that governors by the midHeian period had developed a reputation for gouging the provincial populace. The avarice of governors is underscored by the numerous complaints lodged against them by the farmers and lower officials of the province, as in the case of the famous complaint against Owari Governor Fujiwara no Motonaga mentioned earlier. There are certainly other cases mentioned in diaries and histories, such as the thirty-two articles of complaint against Kaga Governor Minamoto no Masamoto in 1012, the complaint in twenty-four articles against the governor of Tanba, Fujiwara no Yoritō, in 1019, and others; but in none do we have the level of detail that Motonaga’s case provides. We can glean from all these documents that governors enjoyed a relatively free hand in managing the provinces. Motonaga seems to be simply an extreme case, since most of complaints allege not that his deeds were unheard of, just excessive. The former tax rates set by the ritsuryō codes were no longer in force, and governors raised and lowered taxes in accord with local custom, harvest conditions, and what they could get away with. But Motonaga exceeded customary bounds. The attitude of the court in Michinaga’s era seems ambivalent, but in the long run, the nobles did little to alleviate governors’ abuses in the provinces. In the regency period, there are records of sixteen petitions of complaints against governors, either forwarded from the provinces or more commonly brought up to Kyoto by representatives of the populace and the lower officials.97 Early on, the action of the nobles under Michinaga, and his immediate predecessors, was to terminate the governor’s appointment, as in Motonaga’s case. In one case, that of Fujiwara no Yoritō in 1019, the governor himself used mounted troops to drive away the complainants when they went up to Kyoto to deliver their petition.98 Although both Michinaga and Regent Yorimichi rebuked him a few days later, Yoritō was not dismissed from office. And since the complainants were driven away, the case was never heard.99 The historian Sakamoto Shōzō argues that the nobility was largely unaware that the bitter complaints presaged changes in the provinces and that they simply felt obligated to dismiss governors who behaved in such a lawless manner. Even Michinaga, whom he sees as being aware of the severity of the situation, seems to have adopted this attitude.100 Later on, however, the Heian authorities seem to have adopted a more lenient
92 | g. cameron hurst iii
view of the actions of provincial governors, covering at least, if not condoning, their actions. Sakamoto attributed this not to the will of the nobility in general but of Michinaga personally.101 Several examples of court action demonstrate this change. In 1023, for example, a group of seven Tajima officials went to the capital to protest the actions of Governor Fujiwara no Sanetsune, Yukinari’s son.102 They were held hostage in Yukinari’s house, for which action both father and son “lost face” according to Sanesuke.103 Over the course of the next few days, the Tajima officials were questioned, and Yukinari had to report to Michinaga what he knew of the affair. It was clearly very difficult for Yukinari: Sanesuke reports that he was under tremendous stress (shinrō kiwamari naku) over the Tajima governor affair.104 Sanetsune was questioned several times over the next few months; a written account of the affair was presented to the emperor; and finally in the sixth month Sanetsune was determined to be guilty of misdeeds and dismissed.105 But just several months later, Sanetsune’s dismissal was lifted, and he was once again able to resume his duties as governor of Tajima.106 Sanesuke makes only a perfunctory reference to this action, and there seems to be agreement that it was probably the work of Michinaga.107 Yukinari was a long-time associate of Michinaga, that year serving as provisional major councillor and holding the second rank.108 This was an exceptional case: it was the first example faced by Michinaga and the rest of the nobles of one of their own class serving as governor — and the subject of complaints as well. Sakamoto notes an incident in which Sanesuke’s adopted son Sukeyori, then governor of Hōki, was the subject of an informal complaint in the form of a note (rakugaki) that was passed to Michinaga. After some passage of time and frequent discussion, Michinaga dismissed the matter.109 While it is true that both of these cases involved close associates among the kugyō whose sons now held zuryō posts, it apparently was more than just Michinaga’s desire to please friends or protect the nobles as a whole. Sakamoto argues persuasively that Michinaga’s attitude had changed, from one of dismissing governors accused of excessive exploitation to one of responding with force to drive away the complaining farmers and lower officials. He sees Michinaga as now adopting an attitude of covering for the offending governors.110 Perhaps we should not be surprised at this attitudinal change. With his two major construction projects disrupting life in the capital, drawing off considerable resources from the provinces and earning criticism from other nobles, Michinaga was hardly unaware that all was not well in the realm. In fact, a group of senior nobles paid him a visit at the Hōjōji residence, then still under construction in 1021, and Major Councillor Fujiwara no Narinobu confronted Michinaga with the fact that “society is unsettled.” But Michinaga refused to engage him on the issue, preferring to watch the horse races he was hosting there.111 Perhaps Michinaga, having reached the heights of power with three daughters as principal imperial consorts, a son as chancellor, and two grandsons as emperor and crown prince, was concerned simply to protect the realm as he knew it. He was already suffering frequently from effects of his diabetic condition and must have been
Kugyō and Zuryō | 93
aware that he was not likely to live to a ripe old age (he was fifty-six in 1021). The illness, I have argued elsewhere, was affecting his behavior, often leading him to bouts of anger and irrationality that the younger Michinaga had not displayed.112 Besides reviling the senior courtiers, dismissing two secretaries, and scolding his own retainers for nonattendance at court,113 Michinaga also lost his temper with his son Yorimichi, bawling him out in front other courtiers.114 At any rate, Sakamoto too detects a change in Michinaga that led him to ignore the complaints of provincial officials and farmers and instead to support the zuryō. No one can argue that it was not in Michinaga’s best interests. Under him, the Regents House had developed a firm network of support in the provinces from governors who were kinsmen, friends, and clients. The historian Tamai Chikara argues that there was a pattern to Regents House zuryō appointments in Michinaga’s era.115 His analysis of such appointments shows a large number of household retainers (keshi) appointed as governors.116 Tamai notes the importance of the provinces controlled by such appointments: in the capital region, Settsu; in Tōkaidō, Owari and Kai; in Tōsandō, Ōmi, Shinano, and Mutsu; in the San’in, Tanba, Tajima, and Hōki; in San’yōdō, Harima, Mimasaka, Bitchū, Bingo, and Suō; and in Nankai, Awa, Sanuki, Iyo, and Tosa. Tamai relies upon the work of Tsuchida Naoshige, who classified all these provinces and found that they represented six out of the seven top-ranked provinces (in terms of productivity). From this evidence, it is clear that the Regents House under Michinaga monopolized the most fertile and wealthy of Heian Japan’s provinces through zuryō appointments. The locations of these Fujiwara-controlled provinces are also instructive. That is, they include almost the entire Inland Sea area, suggesting how the Regents House under Michinaga and Yorimichi was able to control the flow of trade goods from China and Koryŏ on the Korean Peninsula. Earlier I discussed a jin no sadame in 1005 in which Michinaga engineered a favorable decision on behalf of the Song trader Zeng Lingwen. It was not the first time Zeng had arrived in Japanese waters; in fact, he may even have been a resident. Along with another intrepid trader, Zhu Rencong, Zeng seems to have arrived in Japan several times and moved around the country trading various goods. As far back as 987, Zhu and Zeng appear in many notations in the diaries of Yukinari and Sanesuke, as well as in annalistic histories such as Nihon kiryaku. In the year 1000, for example, Yukinari reported directly to Emperor Ichijō regarding a Dazaifu communication on Zeng and his goods.117 He reported at length on the negotiations for an appropriate exchange rate for Zeng’s trade goods. The matter was reported to Michinaga, and the rate was set the next day. There are many other notations, but a telling notation in Nihon kiryaku for 995 records that more than seventy Chinese went to Wakasa from Echizen.118 Later notations by Yukinari and Sanesuke have Zeng among the group. In all the communications, Michinaga is involved, although he makes no mention of the dealings in his own diary. At any rate, it is very likely that Michinaga was closely connected with the Chinese trade, and the provinces his men controlled were well placed to facilitate the transport of such goods.
94 | g. cameron hurst iii
Not only that, but internal transportation was largely in the hands of Michinaga’s supporters as well: ports on Lake Biwa; Kai and Shinano, well-known for horse ranches; and Mutsu, noted for gold as well as horses. Thus the Regents House controlled the tribute routes from the western provinces, as well as those from Hokuriku, which provided gold and horses.119 This all of course was dependent upon the development of a relatively sophisticated system of transportation and specialized handlers of goods; but through that system the Regents House seems to have developed a broad network of zuryō who managed to supply Michinaga with all the resources he required. Michinaga and Shōen Another important issue of land management that directly addresses the issue of control of the periphery during the height of the Fujiwara regency is the matter of shōen (estate) holdings. This has been an issue long debated by Japanese scholars from an earlier generation that saw the realm as uncomplicated: the Fujiwara as a private entity ruled through the family’s administrative office, and its economic base was located in vast shōen throughout the country. Indeed, Sanesuke once lamented that all the land was under the control of Michinaga’s house, so that no public land remained.120 The next generation of scholars turned that explanation on its head, however, arguing that the Fujiwara power base in the mid-Heian period was in public lands rather than shōen. The situation seems to have stabilized somewhat at a middle ground, with the importance of shōen recognized but not exaggerated. Indeed, in perusing the diaries of Michinaga, Sanesuke, and Yukinari, one finds scattered references to shōen, but they are few and far between. For example, the occasion of Sanesuke’s above lament was an incident involving an act of violence by one of the residents of the Yamashiro estate of Michinaga’s son Yorinobu. Sanesuke does not even provide the name of the estate: it is the behavior of the residents that he first criticizes. Then, in a sort of indignant manner, he laments the sad state of the world in which all the land is in Michinaga’s hands. It is, of course, a great exaggeration; and Sanesuke himself was a significant shōen holder, so we should not take the statement literally. Perhaps it is little more than one of the many barbs aimed at his associate and rival. But what can we say about Michinaga’s attitude toward shōen? For one thing, I am unaware of any statements about shōen either by him or attributed to him by anyone else. Thus it is hard to get a clear picture. The Heian ibun is also of little help since there are few documents relating to Fujiwara Regents House estates there. Still, there can be no doubt that Michinaga was a substantial estate holder for the period,121 and it is clearly during this era that the expansion of shōen in the form of commendations by local cultivators becomes noticeable. At this time, local lords (zaichi ryōshū), in order to avoid being organized into the state’s new myō system, were beginning to commend their lands to nobles and to temples and shrines to gain certain exemptions as shōen. As the most powerful figure in the land, Michinaga was quite naturally the focus of such acts of commendation.122 The evidence on whether in the Michinaga era there was any deliberate attempt
Kugyō and Zuryō | 95
to protect estates or not is mixed: in some cases the regent took action against a governor who had attempted to collect unpaid taxes on a sekkanke estate; in others he was boycotted by the monks of the Kōfukuji for not having taken action against a governor who confiscated one of their estates. I mentioned earlier as well that there were no ordinances regulating shōen during this period either. The evidence suggests a passive attitude on the part of Michinaga. If he did not aggressively seek to accumulate estates, he neither sought to control their spread. In addition, for all his lamentations, Sanesuke seems never to have championed any legislation himself. But that Michinaga became the focus of commendation from local cultivators and gathered a number of shōen that proved to be economically beneficial to the Regents House is undeniable. Apparently, the public income that Michinaga and his family received from all the public emoluments and gifts from various clients was sufficient to underwrite the elegant lifestyle they enjoyed; shōen increased their opulent lifestyle but seem not to have defined it. That would not be true of the late Heian period by any means, nor, as Sakamoto reminds us, does that mean that lower-ranking, non-sekkanke courtiers did not rely far more heavily upon income from estates than did Michinaga.123
Conclusion I have reviewed the career of Fujiwara no Michinaga, generally conceded to be the greatest and most powerful of the Heian-period Fujiwara regents, even though he served only briefly as regent. Above all, Michinaga was born in the right place at the right time, into the right family, and with the right genetic makeup. He began his political life with considerable help from his influential sister Senshi, who favored him above two elder brothers, Michikane and Michitaka, who held power briefly before him. Although the deleterious effects of his diabetes compromised his health, he lived to a relatively advanced age for a Heian courtier, far outliving his brothers. More important, he married two very comely and clever Minamoto women who provided him with abundant sons and daughters over a long time, again unlike his brothers. His daughters proved to be his most valuable assets, with three of them becoming empresses and producing heirs to the throne. Blessed with excellent family ties, Michinaga was shrewd at playing marriage politics, which allowed him unparalleled ability to share in the exercise of imperial authority. As I pointed out many years ago, the Japanese imperial institution thrived by turning an apparent weakness into strength.124 That is, authority and power tend to be bifurcated in the Japanese case, with the sacred authority of the imperial position the most important asset of the Imperial House, a sacerdotal role in state ceremonies that could be performed only by members of the Sun line, however young they might be. Power, on the other hand, was most often decoupled from sacred authority, and others, both members of the imperial kin group (imperial consorts, prince regents, retired sovereigns, and ladies [in]), as well as members not of the imperial line (regents, shogun) were able to exercise power, usually in cooperation with the reigning sovereign but sometimes without it. Michinaga, and the Fujiwara Regents House as a whole, enjoyed such close re-
96 | g. cameron hurst iii
lationships with the imperial kin group through a complex web of intermarriage that it is difficult for the outsider to discern what separates the two houses. The secondary literature often presents this as a nefarious Fujiwara scheme designed to wrest “legitimate” power away from the Imperial House, but it was not a new strategy, just a more successful one than had been attempted even earlier by the Soga, for example. At some point in time, this bifurcation was creatively established, and it protected the Imperial House from destruction by other powerhungry actors. Totally distancing themselves from other kin groups, by such devices as the Amaterasu descent myth and its physical embodiment in the Ise Shrine, as well as the intentional dropping (or never adopting) a surname like all other groups, the Imperial House reserved to itself the exercise of sacerdotal authority but left open the exercise of power by others. Once thus established, this tradition of “outsourcing” of the exercise of effective power proved unshakeable, and it guaranteed the survival of the Japanese ruling line in ways unavailable to its counterparts in China or Korea. That even men as powerful as Minamoto no Yoritomo, Toyotomi Hideyoshi, or Tokugawa Ieyasu never seem to have harbored designs on casting out this tradition and totally remaking the realm with themselves at the core proves the genius of this institution of corulership. But Michinaga must be credited as the most skillful manipulator of this system. As uncle, father-in-law, and grandfather of sovereigns, he was able to cooperate when necessary to effect political decisions favorable to his immediate family or to oppose when necessary to get his way. But he rarely saw decisions go against his will. Moreover, he successfully played the political game of his day, which represented a sort of compromise between imperial authority and noble power. That is, authority for state decisions legally resided with the emperor and the co-participants in that authority (his parents or Fujiwara maternal relatives), while a certain amount of power was delegated to the noble council, representing the pinnacle of nonimperial society. Formal council meetings were frequently held to debate major issues of state, but the emperor and regent were not bound to accept them. There seemed to be a great deal of deference to the opinions of the nobles, and it was certainly a measure of their status in society to be able to voice opinions on such issues. While the form of such meetings, representing a shared ideology of rule, was carefully maintained, as we have seen, they were not binding on the actual decision makers, often no more than de facto rationalizations for decisions that had already been made — or whose outcome was a foregone conclusion. If Michinaga was able to control the imperial part of this equation, he also held sway over the noble council as well. The members were largely relatives from the regental house or other close lineages of the Fujiwara, or Minamoto relatives of his wives, or clients. The kugyō were associates who occasionally disagreed with Michinaga; but for the most part, he was able to maintain control. At the next level of the bureaucracy, the chamberlains and secretaries who controlled access to the emperor and handled the important documents of state, Michinaga also managed to insert kinsmen and clients. And, finally, many of the zuryō in the provinces were related to or beholden to Michinaga in some fashion. Other
Kugyō and Zuryō | 97
regents enjoyed similar relationships and associations and shared the outlines of the political networks of Michinaga, but none were able to attain the same heights. Thus Michinaga had control over a wide and deep network of personal and material resources that allowed him to enjoy a true life of splendor, in which his poetic exclamation “This world is indeed my world!” was as much a statement of fact as it was a boast.
Notes The notes include several abbreviations: GK for Gonki, HI for Heian ibun, MKK for Midō kanpaku ki, and SYK for Shōyūki. I have used a number of works of literature from the Heian period. Since most of them have been translated into English, I have referenced the English translation where available. 1. G.Cameron Hurst III, Insei: Abdicated Sovereigns in the Politics of Late Heian Japan. 2. Cornelius Kiley, “Provincial Administration and Land Tenure in Early Heian”; and Dana Morris, “Land and Society.” 3. See Hurst, Insei, chap. 2; and idem, “The Structure of the Heian Court.” 4. Another title — nairan, or document examiner — was used, predominantly by Michinaga, as a titular office that allowed its holder to influence the direction of court decisions. About that, more below. 5. The ability of the regent or chancellor to act in the political realm was predicated entirely upon the marital alliance with the Imperial House. The system worked best for the Fujiwara when a young emperor born of a Fujiwara woman came to the throne while her father was still alive. As grandfather of the new sovereign, father of the principal consort, and likely head of the Fujiwara clan, a Fujiwara lord such as Michinaga was in a position at the head of the government, his power almost unimpeachable. But many regents did not enjoy such close relationship, and were often only uncles of the sovereign, or even further removed. For the complexity of this relationship, see Kuramoto Kazuhiro, Sekkan seiji to ōchō kizoku, esp. 2 – 41. See also Fukutō Sanae’s chapter in this volume. 6. Many secondary sources discuss the concept of ōchō kokka. See, for example, Sakamoto Shōzō, Nihon ōchō kokka taisei ron, or his more accessible Nihon no rekishi, vol. 6: Sekkan jidai. 7. William H. McCullough, “The Heian Court, 794 – 1070,” 68. 8. SYK, Kannin 2 (1018) 10/16. The translation is from Ivan Morris, The World of the Shining Prince, 60 – 61. McCullough translated it a bit more tersely: “No waning in the glory of the full moon — this world is indeed my world!” (McCullough, “The Heian Court,” 70). 9. See Tsuchida Naoshige, Ōchō no kizoku, 163 – 164. 10. Helen Craig McCullough and William H. McCullough, A Tale of Flowering Fortunes, 499 – 503, covers the construction of this complex in detail. 11. W.McCullough, “The Heian Court,” 71. 12. Helen Craig McCullough, trans., Ōkagami, 195 – 197. 13. Ibid., 190. 14. H.McCullough, Ōkagami, 185. 15. Kugyō bunin, 1:241 – 242 (Shōryaku 6/Chōtoku 1 [995] — the era name changed in the second month on the deaths of eight men of rank of shōnagon and above). The previous year Michinaga had been provisional major councillor, the sixth in rank among the kugyō. Now he stood at the pinnacle of the court after the death of Minamoto no Shigenobu, the seventyfour-year-old minister of the left.
98 | g. cameron hurst iii 16. Ibid., 265. 17. McCullough and McCullough, Flowering Fortunes, 141 – 142. See also H.McCullough, Ōkagami, 39 – 40 and 198. 18. H.McCullough, Ōkagami, 187. 19. Two other women — the daughter of Minamoto no Shigemitsu and another whose name remains unknown — also produced a child each by Michinaga, a boy and a girl respectively. Kitayama Shigeo, Fujiwara no Michinaga, 37. 20. Ibid. 21. William McCullough, “Japanese Marriage Institutions in the Heian Period,” 103 – 167. 22. Michinaga’s father, for example, had at least six wives, including the famous “mother of Michitsuna,” author of Kagerō nikki. He did not live constantly with any of them but visited them in no discernible order or regularity; his children were largely reared at their mothers’ residences, and none of his wives ever seems to have lived with him at his Higashi Sanjō mansion. Michinaga, by contrast, settled down mostly with his wife Rinshi but also spent much time at the residence of Meishi as well. Thus Michinaga lived more of what might be called a settled life with his two wives as married couples. Even though Michinaga seems to have been devoted to Rinshi and indeed spent more time with her, scholars argue that it would be a mistake to designate her as a “principal wife,” with Meishi as somehow lesser in stature. See Tsuchida, Ōchō no kizoku, 85 – 98. 23. Michinaga’s magnificent Tsuchimikado mansion was in fact the property of his fatherin-law, Minamoto no Masanobu. 24. Tsuchida, Ōchō no kizoku, 104. 25. Kuramoto, Sekkan seiji to ōchō kizoku, esp. 2 – 42. Kuramoto has extensively examined the relationships between all the Heian regent and chancellor figures and the Imperial House and been able to document many different patterns in the pages above. 26. H.McCullough, Ōkagami, 196 – 197. 27. SYK, Chōtoku 1 (995) 7/24. 28. Ibid., Chōtoku 1/7/27. 29. Ibid., Chōtoku 1/8/3. 30. Ibid., Chōtoku 2 (996) 1/16. 31. Ibid., Chōtoku 2/2/11. The Imperial Police had already searched the house of one of Korechika’s retainers on the fifth of the second month (ibid). 32. Ibid., Chōtoku 2/4/24. 33. H.McCullough, Ōkagami, 175 – 176. 34. G.Cameron Hurst, “Michinaga’s Maladies: A Medical Report on Fujiwara no Michinaga,” 101 – 112. 35. MKK, Kankō 8 (1011) 12/8 – 9. 36. Ibid., Kannin 1 (1017) 5/27. 37. SYK, Kannin 1/1/22. 38. Ibid., Kannin 1/3/19. (The break-in occurred on the previous night.) 39. Kuramoto, Sekkan seiji, esp. the chart on 13. 40. Hurst, Insei, 576 – 583. 41. Kuramoto, Sekkan seiji, 30. 42. In reviewing the shifts in the locus of power over several reigns, Sanesuke notes that several ministers have controlled affairs and that the “imperial mother” now dominates the court (SYK, Chōtoku 3 [997] 7/5). See also the essay by Fukutō Sanae in this volume. 43. Honchō monzui, Chōtoku 4 (998) 3/12. In this document, where Michinaga tries to give up his office as minister of the left and return some of his prerogatives of office because of his
Kugyō and Zuryō | 99 illness, he mentions his own lack of merit and virtue, attributing his success, among other things, to the support of his sister Senshi. 44. Kuramoto, Sekkan seiji, 30. 45. Hurst, Insei, esp. chap. 2. See also McCullough and McCullough, Flowering Fortunes, 796. 46. Kuramoto, Sekkan seiji, 49. 47. Ibid., 42 – 77. 48. SYK, Chōtoku 3/4/5. See Kuramoto, Sekkan seiji, 54 – 56, for a discussion of the case. 49. SYK, Chōtoku 3/4/22. 50. Kuramoto, Sekkan seiji, 55 – 56. 51. MKK, Kankō 1/4/28. See also Kuramoto, Sekkan seiji, 60. 52. Kuramoto, Sekkan seiji. 53. Nihon kiryaku, Kankō 2 (1005) 8/14. 54. MKK, Kankō 2/8/14. 55. SYK, Kankō 2/14 – 24. Kuramoto, Sekkan seiji, 62 – 64. 56. Kuramoto, Sekkan seiji, 64. 57. Ibid. 58. See MKK, Kankō 2/11/15 – 17; SYK, same dates; and GK, Kankō 2/11/15. Kuramoto, Sekkan seiji, 64 – 65, and 79 – 85. 59. MKK, Kankō 3 (1006) 7/3; GK, same date. 60. GK, Kankō 3/7/15. Michinaga makes no mention of it in his diary. 61. GK, Kankō 3/12/29; MKK, same date. Kuramoto, Sekkan seiji, 69 – 70. 62. GK, Kankō 4 (1007) 3/28. Kuramoto notes that he appears on some twelve different occasions in Michinaga’s diary performing various Buddhist rites on Michinaga’s behalf, whereas Seijū is mentioned only three times (Sekkan seiji, 76, n. 32). 63. On the other hand, sometimes a candidate eager for a job might be satisfied with even a minor province. In 973, as an example, Fujiwara no Atsushige submitted a petition for appointment to the governorship of Awaji, which he noted was a minor province, and thus there was no need to appoint an experienced governor, suggesting that he would do just fine (Honchō monzui, 134, Tenroku 4 [973] 1/15). 64. Kojidan, ed. Kobayashi Yasaharu, 18:8 – 9. 65. The extant lines of the poem — the text indicates there was more — have been translated by Richard Bowring, The Diary of Lady Murasaki, xxxiv. 66. The story is likely true. We have the ōmagaki for the series of appointments in the spring of Chōtoku 2, which clearly shows Kunimori appointed to the vacant governorship of Echizen (697 – 698) and Tametoki named governor of Awaji (701). See Chōtoku ninen Ōmagaki. Sanesuke makes no mention of the incident, and there is no extant record of that year in either Michinaga’s or Yukinari’s diary. But we do know that Tametoki did serve as governor of Echizen, returning to the capital in 1000. Sonpi bunmyaku, vol. 2, 53, lists Tametoki as having served as governor of Echizen. Likewise, the outline of the incident appears in Nihon kiryaku, Chōtoku 2/1/28. The text reads, “The minister of the right arrived at the palace and suddenly canceled the Echizen governorship of Kunimori. He appointed the Awaji Governor Tametoki to that post.” 67. There is a detailed description of all the paraphernalia exchanged between old and new governors in Kokumu jōjō no koto, in Chōya gunsai, 517 – 525. This catalogue of provincial administrative matters (kokumu) gives all the details about how a new governor was supposed to act, from appointment to assumption of duties, including all the various procedures for receipt of the symbols of office, visits to local shrines, and so forth. The document is undated.
100 | g. cameron hurst iii 68. Among the famous series of thirty-one complaints against Owari governor Fujiwara no Motonaga, for example, was the accusation that he spent too much time in the capital. Even when he was in the province, the complainants allege, Motonaga neglected his provincial duties, such as hearing the suits of locals (“Owari no kuni gunji hyakuseira no ge,” in HI, 2:473 – 485, Eien 2 [988] 11/8). 69. Ivan Morris, trans., The Pillow Book of Sei Shōnagon, 23 – 24. 70. Ibid., 42 – 43. 71. Ibid., 34. 72. Naomoto’s petition is collected in section 6 of Honchō monzui (see n. 40), Tenryaku 8 (954) 8/9. 73. Murai Yasuhiko, Heian kizoku no sekai, 260 – 262. 74. Honchō monzui, Chōtoku 2/1/15. 75. Ibid., Tengen 3 (980) 1/5. 76. Kugyō bunin, 1:230, Tengen 4 [981]. 77. Ibid., 145 – 146, Chōwa 3 (1014) 1/23. 78. Murai, Heian kizoku no sekai, 262. 79. Chōtoku ninen Ōmagaki, 689 – 706. The speculation that there were one hundred applicants is that of Abe Takeshi, Sekkan seiji, 121. 80. MKK, Chōwa 5 (1016) 10/15. 81. Ibid., Chōwa 5/10/19. 82. Ibid., Chōwa 5/10/22. 83. Ibid., Chōwa 5/10/24. 84. Ibid., Chōwa 5/11/6-9. 85. Ibid., Chōwa 5/11/7. 86. Murai, Heian kizoku no sekai, 337 (totals from chart). Not surprisingly, Murai notes that most of the horses came from eastern Japan, where there were public and private horse “ranches” (maki), whereas the cows were largely from more agriculturally advanced areas west of Kyoto (338). 87. MKK, Chōwa 5/11/10. 88. Ibid., Kankō 7 (1010) 3/13. Michinaga is visited by three zuryō on their way to their provinces, and he gives a horse to each of them. 89. Murai, Heian kizoku no sekai, 275. Murai provides a chart of the various goods Sanesuke received from provincial governors in the extant months of his diary for 1025. 90. SYK, Kankō 2 (1005) 12/21. 91. Ibid. 92. MKK, Kankō 2/12/21. 93. Ibid., Kankō 7/3/23. 94. SYK, Kannin 2 (1018) 6/20. Michinaga makes no mention of the matter. 95. McCullough and McCullough, Flowering Fortunes, 2:485. 96. SYK, Kannin 2/6/20. 97. See the chart in Sakamoto, Nihon no rekishi, 6:309. 98. SYK, Kannin 3 (1019) 6/20. The struggle between the Tanba folks and Yoritō’s warriors was apparently quite heated, and they caused a good deal of commotion within the palace grounds. 99. Ibid., Kannin 3/6/21, records the confrontation with Yoritō, noting that Michinaga and Yorimichi’ s reprimand was “extremely severe” and that the governor’s actions were “unthinkable” (ryogai no koto nari). 100. Sakamoto, Nihon no rekishi, 6:316. 101. Ibid., 316 – 317.
Kugyō and Zuryō | 101 102. SYK, Jian 3 (1023) 3/18. The notation of their arrival comes on the eighteenth, but it is discovered that they were held at Yukinari’s only on the twenty-first. 103. Ibid., Jian 3/4/21. 104. Ibid., Jian 3/4/29. 105. Ibid., Jian 3/5/1 – 6/2. Rather than saying he was dismissed (genin), Sanesuke says that his duties (rimu) were terminated. 106. Ibid., Jian 3/7/3. 107. Sakamoto, Nihon no rekishi, 6:317. 108. Kugyō bunin, 1:272, Jian 3. 109. SYK, Jian 3/11/3, 5, 8, 17. Sakamoto, Nihon no rekishi, 6:317. 110. Sakamoto, Nihon no rekishi, 6:317 – 318. 111. SYK, Jian 1 (1021) 3/19. 112. Hurst, “Michinaga’s Maladies,” 111 – 112. 113. SYK, Kannin 4 (1020) 9/11 – 13. 114. Ibid., Jian 3/6/19. 115. Tamai Chikara, “Jū, jūichi seiki no Nihon, sekkan seiji,” 51 – 52. 116. In many ways, the keishi zuryō of this era are the forerunners of the inshi zuryō of the late Heian period, discussed in Hurst, Insei, 237 – 253. 117. GK, Chōhō 2 (1000) 7/13 – 14. 118. Nihon kiryaku, Chōtoku 1/9/6. 119. Tamai, Heian kizoku to tennō, 52. 120. SYK, Manjū 2 (1025) 7/11. 121. Sakamoto, Nihon no rekishi, 6:263. 122. Ibid., 268. 123. Ibid., 273 – 274. 124. Hurst, Insei, 217 – 218, n. 1. See also Hurst, Insei, 576 – 583.
part ii
Shifting Categories in Literature and the Arts
5
d Ivo Smits
The Way of the Literati Chinese Learning and Literary Practice in Mid-Heian Japan
U
nlike those who study Japanese history, scholars of Japan’s literature have long been reluctant to seriously take into account texts written in Chinese, or Sino-Japanese. While this peripheral position of Chinese texts is shifting, it is necessary to restate the obvious: insofar as the written word is concerned, premodern and early modern Japan was a bilingual country. The marginalization of Chinese some two centuries ago resulted in a fading awareness of a large cultural heritage. With the rise of kokugaku (national learning) in the late eighteenth century, the bias against Chinese grew steadily and was consolidated in the late nineteenth century with the distinction made between kangaku (Chinese studies) as an academic field devoted to texts from China and kokubungaku (national literature) studies as a field dealing with texts in Japanese. This has resulted in an institutional neglect of kanbun, or Chinese written by Japanese, as a language of Japan’s cultural and literary heritage.1 This neglect, as well as the relative inaccessibility of texts written in Chinese to those of us trained mainly in reading Japanese, partially explains why modern scholars have maintained a long-standing bias against the corpus of kanbun. The situation has changed in the past two decades or so, especially in Japan, and the importance of both literary and documentary kanbun texts for any assessment of the Heian period’s cultural production is generally acknowledged even if many of them remain unread. This essay explores the ideological centrality of Chinese writing in Heian cultural practice and suggests some areas in which kanbun texts could shed light on hitherto unexplored textual terrains of literature in Japanese (kana bungaku), namely what I would call literature of the social fringe, and eroticist parody; both in their own way constitute a textual periphery. I hope to restore kanbun to its proper place alongside kana literature and to consider its centrality in modern literary histories of the Heian period. This task implies more than a mere recognition of Heian kanbun as a factor in shaping Japan’s cultural past; it also involves an understanding of the texts themselves. Such comprehensive views of literary history are an important step toward revising our preconceived notions of Japan’s literature during the Heian period.
106 | ivo smits
Uncomfortable Balances: Literature in Chinese as the Ideological Center One reason for Heian kanbun’s ugly duckling’s position has to do with its nature as a perceived antithesis of poetry and narrative in the vernacular. In the traditional grand narrative of Japanese literary history, the Heian period is presented as the age in which poetry in Japanese (waka) finally gained status as one of the highest literary arts through the first imperial waka anthology, the Kokinshū (Collection of Poems Ancient and Modern, 905). It is also seen as a period in which narrative fiction and memoir literature came of age with such masterpieces as the Genji monogatari (The Tale of Genji) and the Makura no sōshi (The Pillow Book). However, despite the prominence of these now famous works, they counted for little on the surface of the largely male-oriented world of Heian Japan. In his study of Japanese canon formation, Haruo Shirane, for example, mentions a Buddhist text of about 1176, Genji ippon kyō (A Genji Offering with Sūtra Chapters) by the monk Chōken (1126 – 1203), that makes the genre hierarchy very clear: Buddhist and Confucian texts (in Chinese) and Chinese histories and poetry collections ranked above waka, and at the bottom are narrative texts in kana.2 That Chōken’s text accompanied an offering ceremony (kuyō) to counter the “bad influence” of fiction such as The Tale of Genji and its author’s suffering for it in hell only reinforces the textual hierarchy. While The Tale of Genji was canonized within two centuries after it was written and became a cultural icon worth stealing, as in the case of Fujiwara no Teika’s (1162 – 1241) collated edition,3 by contrast the library of the statesman, scholar, and poet of Chinese verse (kanshi) Fujiwara no Michinori (1106 – 1159) contained not one book in the Japanese vernacular script.4 This was also likely the case regarding the collection of Fujiwara no Kin’aki, who died sometime after 1133. When his brother, the monk Renzen (1082? – ?), visited Kin’aki’s house shortly after his death, he recorded in a poem, “His books and scrolls lie uselessly thrown about under the moonlit window. (All the books that had been handed down through time, both those from Japan as well as from China [wakan]: there was no one to organize them. That is why I write this.)”5 Although wakan might mean “[books] in Japanese and Chinese,” here it more likely means “[books] from Japan and China” and one should be careful in assuming that this necessarily implies that any of the books from Japan in Kin’aki’s library were written in Japanese. Compared to prose, the situation was somewhat different for poetry. During the ninth century, the time of the three imperial kanshi anthologies Ryōunshū (Cloud-topping Collection, 814), Bunka shūreishū (Collection of Beauties among the Literary Flowers, 818), and Keikokushū (Collection for Governing the State, 827), and personal kanshi collections offered to the throne, such as Sugawara no Michizane’s (845 – 903) Kanke bunsō (Michizane’s Writings, 900), bunjin or literati scholar-poets were cultural heroes who potentially embodied forces that countered the rising Fujiwara Regents House. The Confucian ideal honored learned men who stood by their monarch. Suga-
The Way of the Literati | 107
wara no Michizane is a prime example of a courtier from a family of limited influence who, through his usefulness as a scholar-bureaucrat, rose to a prominent position by the emperor’s side, from which he could threaten Fujiwara dominance.6 Following Michizane’s exile in 901, we find the emergence of a new aristocratic ideal, described by Joshua Mostow as being “antiprofessional, antimeritocratic, and, to some extent, anti-intellectual.”7 The rise of the Regents House was made possible largely by a system that brokered power through marriages and therefore the political center shifted toward the supposedly private quarters of the palace complex, where the prevailing script was kana and the cultural emphasis rested with writings in vernacular Japanese.8 Indeed, as tenth-century emperors no longer seemed to care for kanshi collections, the Regents House gradually overtook the role as its patrons and simultaneously coerced the literati into underwriting the new power relations. Fujiwara marriage politics and support of women’s writing went hand in hand with more traditional use of kanbun production. From the tenth century onward, public kanshi collections as a rule were sponsored by the Regents House. Kanbun’s saving grace rested largely with its role in matters of ceremony and state. Official documents, law codes and petitions, poetry at state banquets, seemingly private texts such as journals or religious dedications (ganmon), and public yet unofficial texts that hovered somewhere between state edict and personal essays (encyclopedias or manuals on ceremonial) were to be composed in kanbun. Given this convention, it is tempting to assume that kanbun might have been ideologically central but peripheral in practice and that the political and cultural elites of Heian-kyō paid only lip service to kanbun’s role in the textual landscape. This is far from the case. The Heian government evolved around rituals of which the language was Chinese. This realization in particular helps to explain why so many Heian kanbun texts have remained.9 It would also be misleading to assume a marginal role for kanbun in Heian literary history because it was seldom practiced in court circles. In Shikyōki (Rec ord of the Poetic Realm), a minihistory of Chinese poetry written in kanbun, the scholar, bilingual poet, and raconteur Ōe no Masafusa (1041 – 1111) wrote, In our country Chinese poetry originated in the Kōnin and Shōwa eras [810 – 847]. It was at its peak during the Jōgan and Engi eras [859 – 922], was revived during the Shōhei and Tenryaku eras [931 – 956], and prospered again in the Chōhō and Kankō eras [999 – 1011]. Broadly speaking, there were about thirty-odd poets. If I narrow it down to the truly talented of these periods, then we are left with no more than six or seven people.10 The number of poets may not impress us partly because Masafusa probably limited himself to those whom he perceived to be truly outstanding poets. In fact, the actual number of kanshi poets for these periods is much higher. In either case, there can be no doubt that the court of Emperor Ichijō (980 – 1011; r. 986 – 1011) was a period of kanbun revival. While we now tend to think of The Tale of Genji and The Pillow Book as classic masterworks of this age, Masafusa had in mind such kanbun collections as Nikkanshū (Collection of Japanese Views, ca. 930s;
108 | ivo smits
compiled by Ōe no Koretoki, 888 – 963) in twenty books,11 Fusōshū (Japanese Collection, ca. 995 – 998; compiled by Ki no Tadana, 957 – 999) in sixteen books,12 Honchō reisō (Beautiful Poems from Our Court, ca. 1010; compiled by Takashina no Moriyoshi, ? – 1014) in two books,13 and Honchō monzui (Literary Essence of Our Court, ca. 1058; compiled by Fujiwara no Akihira, 989? – 1066) in fourteen books.
The Construction of a Canon: Indigenous Chinese The mid-Heian period did not merely witness a revival of kanbun writing after the dominance of waka throughout the tenth century. It also ushered in a period in which Japanese literati began building a new kanbun tradition and with it a new canon of poetry in Chinese. This new canon focused on Japanese authors, not Chinese poets, with the eventual result that by the late-Heian period (1086 – 1185) kanshi became an indigenized notion.14 This tendency is also evident in kanbun manuals of the late-Heian and very early Kamakura periods such as Sakumon daitai (Basics of Composition, 1108), compiled by Fujiwara no Munetada (1062 – 1141), and Tekkinshō (Throwing Metal Notes, ca. 1206 – 1210).15 Practically all the examples given in these two manuals are by Japanese poets. One reason for the focus on Japanese examples was the widespread habit of composing versetopics (kudai; see below). Extensive rules existed for the breakdown of the versetopic into the poem’s first couplet and ways to work up the theme in the remaining three couplets. While kudaishi, or verse-topic poems, were not unknown in China, this particular genre with its specific compositional rules was very much a Japanese phenomenon; consequently available examples were Japanese. This development suggests that mid-Heian literati were engaged mostly in a dialogue with earlier Japanese literati, much less so with authors from the Asian mainland. However, not everyone was given a voice within this dialogue. It is striking to see that Japan’s very first anthology of Chinese poetry, Kaifūsō (Fond Recollections, 751), contemporaneous with Japan’s oldest waka collection, the Man’yōshū (Collection of Ten Thousand Leaves), was as good as ignored throughout the Heian and Kamakura periods. The near denial of this anthology in later centuries is perhaps symbolic of the clear break that Heian culture and society seems to have made from the previous era. While not necessarily a conscious program, Heian literary and cultural life appears to have been bent on reinventing itself through a process in which the ninth century functioned as an important episode. This century witnessed the compilation, in quick succession, of three imperial kanshi anthologies that were relatively vague about the history of Chinese verse in Japan. A past was acknowledged, but in terms that obfuscated a need to explicitly pay homage to a previous court.16 This feeling was reinforced by later Heian kanshi histories that associated the beginning of kanshi composition in Japan with the court of Emperor Saga (786 – 842; r. 809 – 823).17 The ninth century was also the period when waka poets created a new body of poems that would officially be sanctified with the compilation of Kokinshū. Its preface selected early Heian poets for inclusion in its hall of fame, the rokkasen or six poetic
The Way of the Literati | 109
immortals; and while Man’yōshū poetry generally is a clear point of reference, the only Man’yōshū poets similarly identified are Kakinomoto no Hitomaro (? – 710?) and Yamabe no Akahito (? – 736?).18 The three imperial kanshi anthologies of the early ninth century, although quickly recognized as an important milestone in literary history, also seemed to have very little actual impact on later generations. In fact, this seems symptomatic of kanshi histories in general. Where it is relatively easy to write a history of waka that suggests a strong continuity and historical awareness, the history of kanshi writing in Japanese seems to be one of fits and restarts. It is only in the Edo period (1600 – 1868) that attempts were made to be comprehensive in constructing a kanshi canon. In Heian Japan, on the other hand, the history of kanshi apparently was the history of Chinese poetry only after it regained a place on the map following the imperial patronage of waka. This view was still endorsed by the Kyoto scholar and kanshi poet Emura Hokkai (1713 – 1788), who in his Nihon shi shi (History of Chinese Poetry in Japan) of 1771 devotes curiously little space to Kaifūsō, skims over the three imperial anthologies, but dwells considerably on post-900 kanshi production. Adaptation of its Chinese cultural heritage is a recurrent theme in Japan’s history. Chinese notions of the forms and functions of literature were not borrowed wholesale; instead Japan selected what seemed useful, and soon “China” became a construct that was only tangentially related to the state across the East China Sea. This construction can be aptly described as “China within Japan,” or even multiple Chinas within Japan. It suggests Japanese perceptions of China and their role in incorporating elements of that culture in molding a tradition to fit contemporary ideas of literature. China was indeed the model for many facets of Japanese cultural life, but this archetypal China within Japan was one of Japanese making. The result was an integration of Japanese and Chinese elements that created a dynamic blend that, while not always harmonious, extended to the domain of kana literature as well.19 The continuous appropriation between Japanese and Chinese elements is often indicated by the term wakan, “Japan and China” (or “Yamato and Han/Kara”). The term itself suggests that it is not very productive to think in binary oppositions between Japan and China. Rather, the process resulted in what Thomas LaMarre has called “a binary machine that could synthesize and organize multiple forms of expression and production: the Yamato-Han or ‘wa-kan’ assemblage.”20 LaMarre’s study focuses on script and the ways in which its varieties, from mana (or what we now would call kanji or Chinese characters) to kana, are best seen as different modes of calligraphic performance, that is, how they may visually represent a text. In doing so, he attempts to steer our focus away from obsessing about the language of texts in favor of blurring the notion of a distinctly “Japanese” language and culture, as opposed to Chinese cultural and linguistic dominance. Indeed, as Tomiko Yoda suggests, it is more fruitful to allow for “the possibility that multiple cultural values and logic may have operated in Heian court society without necessarily constituting a sharp dichotomy.”21 In this context, it is simply inadequate to suggest that gender identifications
110 | ivo smits
of script — “male writing” (otoko moji, that is, mana) and “the female hand” (onnade, that is, kana) — equate to mutually exclusive ideas about public and private. “China” could be seen to invade the private quarters of women when Fujiwara no Michinaga’s (966 – 1027) daughter Shōshi (or Akiko, 988 – 1074) moved into the palace in 999. She made a tremendous social impact with the screens she brought with her. Such screens were regularly ornamented with cartouches inscribed with poetry that was not necessarily limited to waka, for it appears that women owned screens filled with Chinese texts, set in Chinese designs, and even trimmed with Chinese brocade.22 Shōshi, who would become a long-lived and influential imperial dowager, showed an active interest in things Chinese, it seems. Well known is the instance when she asked Murasaki Shikibu (973? – 1014?) to teach her to read the “new ballads” (xinyuefu) of the famous Chinese poet Bai Juyi (772 – 846).23 Similarly, literati could slip in and out of different languages at will, composing kanshi and waka, as well as Chinese prefaces to poems in Japanese. Within the realm of kanshi composition itself they might use idiosyncratic forms of kanbun, and even, on occasion, refer openly not to Chinese classics or even earlier kanshi, but to poetry in Japanese (waka) as the orientation points for kanshi composition. While active participation in kanbun production was an exclusively male prerogative throughout the Heian period,24 there were quite a few women at court who had varying degrees of understanding of how to read Chinese texts. In this sense women should be taken into account as producers of meaning of these texts despite their exclusion from authorship. Historical records for this period abound with meetings to compose kanshi to which women were never invited. Such educational opportunities were available only to male literati. This situation seems to compare to the development of schools or “houses” (ie) in the field of medieval waka, where women were active as poets but were excluded from positions of influence.
Training Poets: Chinese Learning as a Way The building of a new kanbun canon went hand in hand with yet another development, namely that of Chinese studies as a “way” or michi. This in turn was accompanied by the emergence of a tradition of family learning (kagaku), in which different scholar families or houses attempted to gain monopoly over certain types of scholarship. By the early tenth century, Ōe no Asatsuna (886 – 957) spoke of “the way of scholarship” (gakumon no michi) in the preface to his kanbun manual.25 And when the scholar-poet Ōe no Masahira (953 – 1012) died in 1012, Fujiwara no Sanekane lamented that “the way of literature (bundō) has vanished.”26 Ever since Konishi Jin’ichi’s studies of the concept, michi has been recognized as a leading principle of Japanese medieval arts, centering on the dedicated pursuit of an art or technique and valuing anyone who committed to it.27 This theoretically egalitarian idea valued craft over breeding and was antithetical to the Heian code of miyabi (courtly refinement) for the arts and family connections for politics. Chinese studies was the only art for which a training trajectory and
The Way of the Literati | 111
dedication to specialization was accepted, but this very fact seems to have prevented scholars from rising very high in the echelons of power and the court circles of what Ivan Morris once called the “good people.” Consequently, unlike waka poets, kanshi poets may at first sight seem to have started out as “medieval” poets, by which I mean that they experienced poetry as an art form that required tremendous dedication and not a mere social expression.28 For waka poets, this notion of poetry as an art or a way took shape from about 1100 onward. In that respect the literati’s situation in the mid-Heian period did not significantly differ from that of waka poets from the twelfth century onward: expertise had always been at the core of their existence, both artistically and socially as well as economically. It is partly for such reasons that when Sei Shōnagon (966? – after 1017) listed scholars as splendid things (medetaki mono) in her Pillow Book, there was a certain ambivalence in her praise. I need hardly say how splendid I find a learned professor (hakase). He may be lowly of appearance, and of course he is of low rank (gerō), but he is free to approach the most eminent members of the emperor’s family, and he is consulted about all sorts of special matters, serving as an imperial tutor (onfumi no shi), so that the world at large regards him as an impressive figure. When he has composed one of his religious dedications (ganmon) for the emperor, a petition to the throne (hyō), or some poem preface (mono no jo), he becomes the object of universal praise.29 While learning was an essential prop for state matters, Sei Shōnagon took issue with “learned professors,” for their value lay in their learning despite their lack of general sophistication and their unimpressive family background. Her perception of literati as “of low rank” is accurate enough: although they came from court families, the world of the elite nobility was highly restricted. The famous mid-Heian scholar and poet Fujiwara no Akihira, for instance, who had fulfilled many distinguished academic positions, was promoted to junior fourth lower rank only at the advanced age of seventy. Sei Shōnagon’s ambivalence toward the cult of miyabi and its problematic relation to the low-ranking class that produced the learned professors likely stems from her own upbringing in such a family of scholars. Her father, Kiyohara no Motosuke (908 – 990), was acting governor of Kawachi and governor of Suō and Higo provinces. While best known as a waka poet and the cocompiler of the second imperial anthology, Gosenshū (Later Collection, mid-tenth century), he was also a man of Chinese learning and an early scholar of the Man’yōshū. A closer look at the family background of literati reveals that several other women whom we associate with the building of narrative fiction in kana and the creation of court-oriented literature were born to families of literati serving in the provinces. The father of Murasaki Shikibu, Fujiwara no Tametoki (dates unknown, active 997 – 1018), was a student at the State Academy (Daigakuryō) and served at the Ministry of Ceremonies (Shikibushō) before his appointment as governor of Echizen Province, where he exchanged poems with the Chinese merchant Zhou
112 | ivo smits
(or Qiang) Shichang.30 The author of Sarashina nikki (The Sarashina Memoirs, ca. 1060), the daughter of Sugawara no Takasue (973 – ?), was born to a family descended from the illustrious Michizane with a well-established tradition of Chinese learning and served as assistant governor (suke) in Kazusa and Hitachi provinces. In their youths, these women accompanied their fathers to the provinces where they spent what historian Tanahashi Mitsuo calls “their impressionable girlhood,” implying that the “female hand” wrote in a wakan world from an early age.31 To a large degree, the association of each of these women with her father’s tradition of Chinese learning seems to be at odds with the court culture of miyabi, which is said to characterize women’s writing in the mid-Heian period. As Sei Shōnagon points out, it was not unusual for literati to regularly meet with emperors because of their expertise in kanbun literature. In creating and shaping a way of Chinese learning, comments on texts and poets became the basis from which a new canon of kanbun was built, the pattern typical of Chinese scholarship transmitted throughout the Heian period.32 In this sense, the anecdotal quality of many remarks in the Gōdanshō (The Ōe Conversations, early twelfth century), and many prefaces to both kanshi and waka should probably be regarded as an essential element in transmitting an established attitude toward texts and literature. These anecdotal remarks, similar to Chinese poetry talks (shihua), constitute one form of commentary and were intended to be instructive as well as entertaining. Again this is similar to waka, where one sees the emergence of the anecdotal remarks as a common feature of poetic treatises (kuden, zuinō) only after poets began to formulate their art as a way and started to withdraw into schools. As most observations about poets in Japan are based on developments in the field of waka, it is worthwhile to ask how the worlds of kanshi and waka differed. For practitioners of Chinese poetry, formal training was acquired through an institution. Unlike waka poets, who learned their craft from their fathers or through private teachers but had no formal curriculum in a recognized institute, most kanshi poets received their education at either the State Academy or related clan-based colleges such as the Kangakuin.33 Founded about 670, the State Academy offered courses (or “ways”) in Confucian classics (myōkyō), literature (monjō) and history (kiden), law (myōhō), arithmetic (san), calligraphy (sho), and pronunciation (on). At first these courses were taught with unvarying attention, despite an inequality of status within the curriculum. However, when the academy’s literature program increased in popularity and prestige, differences between the programs in literature and history grew less distinct, and eventually the two were joined together. Knowledge of Chinese literature was a prerequisite for the composition of documents in Chinese, the language of state documents, and the sixth-century Chinese anthology Wenxuan (Selections of Refined Literature; J. Monzen) was a standard item on the reading list for examinations, as eventually were the collected works of the famous Tang poet Bai Juyi. Chinese poetry was an important aspect of court ritual, and the academy catered to that need by teaching composition.34 Kanbun manuals and other primers facilitated the educational trajectory for
The Way of the Literati | 113
Chinese learning.35 Typically, a boy of the Heian upper class would start learning to read and write characters at about age six. The most important textbooks for beginners became known as the four primers (shibu no dokusho). One was Mengqiu (Youth Inquires; J. Mōgyū; early eighth century), a biographical dictio nary of Chinese cultural history. The text consists of nearly six hundred minibiographies arranged in pairs, all with headings of four characters, to make their contents easy to remember.36 The next two texts were Qianzi wen (Thousand Characters Text, J. Senji mon; ca. 500), and Baiyong (Hundred Compositions; J. Hyakuei; early eighth century).37 From the eleventh century onward, a Japanese text, Wakan rōeishū (Japanese and Chinese Poems to Sing, early eleventh century) was added to the list, thereby introducing couplets by Japanese kanshi poets into the curriculum.38 From this point on, students began reading basic texts from China, all from the four major categories of the Chinese writing canon: the classics (jing, J. kyō), the philosophers (zi; J. shi), the histories (shi) and the collections (ji; J. shū). To facilitate studies, Japanese scholars developed a plethora of compendia on the composition of texts in Chinese. The post of professor (hakase) and head (kami) in the State Academy tended to remain within a limited number of families specializing in Chinese studies, and many kanshi poets were somehow related by blood or marriage. These literati had acquired a thorough formal training, including examinations, in the reading and composition of Chinese poetry. In fact, many remained affiliated with the State Academy or clan institutions, either as head, professor, or acting assistant professor (tokugōshō), which allowed them to receive a salary or stipend in recognition of their academic skills. In short, long before other arts developed into creative skills practiced principally by recognized experts from different schools, Chinese learning was from the beginning a possible career track. The unease over Chinese scholar-poets that Sei Shōnagon experienced is indicative of the unique position held by literati in a court society that needed at least another century before it would more broadly acknowledge artistic talent and training as a form of cultural capital.
The Wasp Waist Controversy A debate in 997 provides a glimpse of how scholarly reputations were made and broken and how academic knowledge was an essential tool in crafting poetry. Identifying incorrect details was essential in establishing one’s position in the poetic pecking order. Judging from disputes at poetry contests, this also holds true in the case of waka discourse. The overriding impression is that poets seldom won a round in poetry contests (utaawase) because their poetry was good, but rather because their adversaries lost on points. Utaawase and State Academy examination disputes constituted formal settings or “places” (ba) where literary precedent and a painstaking observance of rules formed the backbone of the debates. Discussions of work submitted for an academic degree echo the double notion of the wenchang of Tang China, a term that could refer both to the examination grounds and to the arena of letters in a much broader sense.39
114 | ivo smits
An examination at the Ministry of Ceremonies (shōshi) was an essential step in climbing the ladder of the State Academy and also provided access to stipends. From there one could advance to the status of acting assistant professor and could even be appointed as a provincial governor.40 For those who sought to develop a career as Chinese scholars, composing poems for this examination was a vital skill. It was within this context that the ministry examination of the seventeenth day of the seventh month of 997 gave rise to a scholarly controversy that managed to divide the State Academy into two camps around two leading scholar-poets of the time, Ōe no Masahira (952 – 1012) and his nemesis, Ki no Tadana (also read Tokina).41 The examination candidate was the provisional graduate student (gimonjōshō) Ōe no Tokimune (? – after 1057),42 and the submitted text was a fivecharacter poem on a topic taken from the Shijing (The Book of Odes).43 Tadana wanted to fail Tokimune, but Masahira objected. Tokimune was Masahira’s adopted son, and in this case there was more at stake than a simple difference of opinion between examiners. There was another examiner involved, but Tadana and Masahira took center stage. Both were scholars with impressive reputations. Masahira could take pride in being the grandson of the famous Koretoki, who compiled the Nikkanshū and the Senzai kaku (Fine Couplets for a Thousand Years, first half of the tenth century), an anthology of more than a thousand couplets by Chinese poets, some of them almost contemporary with Koretoki. Furthermore, Masahira was building a career as scholar and bureaucrat and would later serve as provincial governor; as such he is the typical representative of the class of provincial governor literati. In addition, Masahira was also a waka poet and is considered one of the “thirtysix poetic geniuses of the classic age.”44 Tadana was also a gifted bureaucrat, who as major private secretary (dainaiki) was selected together with Masahira to draft the court’s reply to a Chinese monk who had arrived the previous year.45 Tadana’s early death cut short what might have been a fruitful administrative career, but before his death in 999 he finished the Fusōshū, which probably was intended for the Regents House or even the emperor.46 Together with Ōe no Yukitoki (955 – 1010), Masahira and Tadana were regarded as important kanshi poets of the Ichijō court. Tadana’s critique of the examination piece centered on two points. One was what he considered a poetic defect known as wasp waist (hōyō), which stipulated that the second and fifth character in a five-character line of verse should not have the same tone. Masahira’s refutation argued that this rule applied only to the first line of a couplet and not the second, as was the case with Tokimune’s poem. He based his position on the Wenbi shi (Code of Style), an anonymous Chinese manual that, ironically, Tadana had also referred to but apparently not read as thoroughly. The second point related to what Tadana believed to be a series of flaws (kakin) and even a grave error (kyogai). For example, in his fifth and sixth couplets Tokimune twice used the character sō (grass). Known in handbooks as repetition defect (nennibyō), this was not exactly a fatal slip, but it was generally considered more elegant to avoid homonyms or outright repetitions.
The Way of the Literati | 115
Three days after the exam Masahira filed a report to Michinaga, then minister of the left (sadaijin), an action that suggests just how firmly the Regents House had established itself as central authority in the world of Chinese learning. Masahira called Tadana a nitpicker (“he is blowing away the animal’s fur to look for wounds”)47 whose attitude “harms the way of literature.” This report prompted Tadana to write a rebuttal of considerable length, in which he took Tokimune’s poem apart line by line to substantiate his points of criticism. Masahira then retorted with an essay that was three times longer than Tadana’s already impressive if tedious epistle. He refuted every objection raised by Tadana and listed several precedents for Tokimune’s phrasings. Tadana remained unmoved and maintained that there were still grounds for failing Tokimune; he pointed out that there was a collection of old examination poems available, Ryūmonshū (Dragon Gate Collection),48 which the student should have consulted. After more than six weeks, the bombardments with textbook classics seemed to abate, but the strained relations between the two literati continued. In the end Tokimune survived the incident and would eventually serve as head of the State Academy before being appointed as governor of Dewa and other provinces. The quarrel may seem exceedingly trivial to modern ears, and Tadana descended into such petty detail that it earned him a reputation as nitpicker, acting, in the words of Ōsone Shōsuke, “childish for such a distinguished scholar.”49 Nevertheless, Tadana’s insistence helped set a precedent for the evaluation of later examination poems. The quarrel itself was elevated to the level of academic jurisprudence by inclusion not only in the eleventh-century Honchō monzui, but also in an early twelfth-century compendium, Chōya gunsai (Collections of Our Court and People, 1116).50 It was through this incident that later scholars understood the standards set by earlier kanshi poets. This is evident in a comment on Tadana’s poetry by Masahira’s great-grandson Masafusa: “Tadana would rely solely on the perceived gist of old collections. He had no new ideas to offer whatsoever. All his parallel prose and his every verse are snippets of old lines. That is why his style is so classical. But when he is not making use of the old masters, he is incapable of surprising you, because he lacked any new ideas.”51 If Chinese learning, with poetry composition as one of its key elements, was a “way,” then the literati community was also a battleground for different houses that all tried to make their mark. Formal settings, such as examinations, could give expression to that competition.
Poetic Genre as a Social Code In Heian Japan, poetry was gradually becoming an art, but it also remained a social act. One must distinguish between audiences and situations; the more formal the situation, the more formal the poem. This protocol centered on the idea that a good poem was to be suitable to the atmosphere of its occasion. The poem’s setting is traditionally described as the ba or “place.” Scholars revert to this traditional term to indicate the audience in an almost physical sense with the emphasis on where, rather than to whom, the poem was read. While the
116 | ivo smits
audience at different ba may be the same, it is understood that they will react differently according to the situation. A ba expresses a sense of focus on audience and moment. From the Heian period, Japanese court poetry and its moments of composition were divided into two categories, hare (formal) and ke (informal). The latter indicated poetry composed for everyday situations with low “visibility,” such as casual exchanges, love poems, and letters. The former consisted of poems that were expressly intended for such public occasions as poetry contests, gatherings at the Imperial Palace, banquets, or folding screens. The distinction was not absolute, and it might be better to speak of degrees of formality, as a ba was either formal or less formal.52 This division into degrees of formality held true for both waka and kanshi and by no means implied a value judgment, but instead emphasized the importance of the occasion for which a poem was composed. In addition to theme and diction, the format of kanshi was also determined by the characteristics of the occasion. In his Chōya gunsai, Miyoshi no Tameyasu (1049 – 1139) distinguishes several situations, or ba, that require distinctly different styles of poetry (shoshitai).53 His breakdown follows a social hierarchy that ranges from poetry composed at gatherings organized by emperors and retired emperors and gatherings hosted by high-ranking court nobles to outings at shrines and temples, rituals for the worship of Confucius (sekiten),54 study groups,55 and individual poetry exchanges. The first and socially most important categories that Tameyasu distinguishes are at imperial command (or “in response to an imperial poem,” ōsei), in response to an emperor’s direct family member (ōrei), and at request (ōkyō). Ōsei is further broken down into the categories sovereigns (teiō) and retired emperors (daijōkō), ōrei into crown princes (taishi) and principal imperial consorts (kōgū); the ōkyō categories consist of princes (shinnō) and noble families (kugyōke). Further stylistic categories are the Kitano Shrine and the mausoleum at the Kichijōin (Kitano, Kichijōin byō), that is, poetry gatherings in honor of Sugawara no Michizane;56 Confucius worship; Literature Hall (Monjōin, that is, gatherings of State Academy and Kangakuin students); Imperial Library (Goshodokoro, which was a meeting place for palace courtiers and scholars); gatherings to encourage learning (kangakue)57 and visits to mountain monasteries (yū sanji); examination at the ministry (shōshi) and examination while circumambulating the island (hōtōshi), that is, the examination at the Ministry of Ceremonies, such as Tokimune’s discussed earlier, as well as at the Suzaku-in when the emperor was present;58 and finally, reading (dokusho), specifically referring to the composition of poems about texts that just have been read with a group, such as the History of the Later Han. Not explicitly mentioned in this section of the Chōya gunsai are such categories as poem exchanges (zōtō) between individuals and solitary composition (dokugin). It will be noticed that Tameyasu mentions only group activities, thereby illustrating that the communal nature of poetic practice in East Asia is far more pronounced than in the West. In the case of kanshi, as was increasingly the case with waka, the act of poetry composition underlined a poet’s association with the community of the cultured elite whose recognition and esteem could be won mainly through artistic and intellectual prowess. However, such prowess needed
The Way of the Literati | 117
to be seen in action, as it were; the stage for that action consisted of the numerous poetry gatherings, not one’s study, and lasting esteem was won by inclusion in one of the many anthologies, less so the private collections of individual poets. Tameyasu’s differentiation of situations was in practice accompanied by different forms of poetry. For the poems’ format, the basic distinction was between verse-topic poem (kudaishi) and non-verse-topic poem (mudaishi). Literally, mudaishi meant that the poem had no assigned verse-topic, or kudai. Ever since the middle of the tenth century, lines of Chinese poetry, usually of five characters, regularly served as a topic for either Chinese (kudaishi) or Japanese poetry (kudai waka). In fact, the use of a verse-topic had become such a habit that kanshi came to mean verse-topic poem, and extensive rules existed for the different ways in which the verse-topic could be incorporated in the poem. Any Chinese poetry that was not a verse-topic poem was given a new name: mudaishi.59 The Chōya gunsai and records of poetry gatherings make it clear that during such occasions of a fairly formal nature, verse-topic poetry was the prevailing form. For example, roughly two-thirds of Honchō reisō, the Ichijō court’s kanshi anthology, consists of verse-topic poetry. Exceptional situations, which tended to have a less visible and rather informal profile, were honored with mudaishi. Places — such as outings to mountain monasteries, study groups, or poem exchanges — did not require assigned verse-topics and were therefore much freer in their choice of themes and wordings. The titles of such poems, often an impromptu (sokuji) or a declaration of emotion (jukkai, genshi, or kan ari), indicate the conscious clearing of poetic and thematic space.
Poetic Freedom and the Social Fringe Narrative accounts (ki) and poem prefaces (shijo, wakajo) were two specific poetic genres wherein the author was allowed a certain degree of latitude. The poem preface was a very important genre, as Sei Shōnagon already indicated, for of the 432 items in Honchō monzui, 139 are kanshi prefaces, with an additional eleven waka prefaces. In 1108, Fujiwara no Munetada ranked miscellaneous prefaces (zatsujo, or zōjo) as the first prose genre in his kanbun manual.60 These and the non-verse-topic poems came close to forming a literary free zone, a set of prose and poetry genres that allowed tremendous thematic variety and freedom of vocabulary. Although not without models of their own, their themes could cover almost anything, from “unlucky” subjects such as sickness to local folklore or itinerant entertainers. What we now consider to be interesting or innovative poetry was all composed on well-defined occasions only and through a given format: mudaishi composed at informal gatherings. It is in these literary free zones that one gets an idea of how Japanese kanbun texts could deviate from Chinese standards and explore topics that were innate to the interests of Heian court culture. The historian Tanahashi Mitsuo suggests that Heian court culture consisted of two “streams” of literature.61 The first was represented by the introspective tales (monogatari) such as the Genji monogatari; the other consisted of a type of documentary literature exemplified by certain kanbun works in the nar-
118 | ivo smits
rative account genre as well as anecdote (setsuwa) collections such as the Konjaku monogatarishū (Tales of Times Now Past; ca. 1120) and Uji shūi monogatari (Tales from Uji, ca. 1180s or later). Although Tanahashi’s remarks refer to prose, the observation that there is a strong preference for descriptions of the world outside the court holds true also for non-verse-topic poetry. In their reflections of a world beyond court society, these texts build toward what I would like to call the literature of the social fringe or texts of the periphery. One example is the Shin sarugaku ki (A New Account of Sarugaku, ca. 1052) by Fujiwara no Akihira, the founder of the Umakai Ceremonial Branch of Fujiwara scholars. Although it is risky to attempt psychological stereotyping, Akihira gives the impression of having been a willful man who was not particularly impressed with social conventions. In one curious incident in the winter of 1034, Akihira and Fujiwara no Sanenori (dates unknown) appeared outside the examination hall and prompted students on the proper characters and pronunciation of the examination poems.62 This prompting created an outrage among the academic community. Akihira and Sanenori were interrogated and punished, but received pardon half a year later with a general amnesty intended to counter a drought. This particular incident did not prevent Akihira from eventually serving as tutor (jikō) to the future regent Yorimichi and later becoming head of the State Academy (daigakuryō no kami), lecturer to the crown prince (tōgū no gakushi), and professor of literature (monjō hakase). Akihira’s penchant for the unconventional led him to explore aspects of Heian society that lay outside the court’s realm. Shin sarugaku ki is a description of a fictitious family that covers most social groups that in one way or another constitute the periphery in Heian culture.63 The story centers on a certain Uemon no jō (var. Emon no jō), resident of the semiurban western part of Kyoto, who views sarugaku performances in the capital in the company of his three wives, sixteen daughters, and nine sons. The large number of children, including his sons-inlaw, provides a glimpse into the variety of professions practiced by townspeople. All excel in their line of work whether as gambler, warrior, farm overseer, shaman, scholar, wrestler, courtesan, painter, et cetera. These are groups that Akihira might have encountered while employed in the Police Bureau (Kebiishichō) in his early forties. What is perhaps even more remarkable than Akihira’s choice of subject matter is his balance between description and irony. Shin sarugaku ki seems to have been in part intended as a model book, displaying the stylistic possibilities of Chinese writing, but it is also an exhaustive experiment in charting a social sphere that was close in geography yet distant in the social order. When Akihira wrote his New Account of Sarugaku, the periphery of Heian society had already served as subject matter for the Hinjo no gin (Song of the Poor Woman) and Hakuchoō shijo (Poem Preface to “The Old Man [Selling] White Chopsticks”) by Ki no Haseo (845 – 912) and for Ōe no Yukitoki’s Yūjo o miru shijo (Poem Preface to “On Seeing Courtesans,” 996?).64 There also is a ten-poem sequence that Sugawara no Michizane wrote when he was governor of Sanuki Province.65 In it, he describes the plight of the poor, suffering during the austere winter. Such “documentary” nature in narrative accounts and poem preface
The Way of the Literati | 119
genres also extends to correspondence model books (ōraimono), another genre that allowed for a tremendous topical range. There, too, Akihira made his mark. His correspondence became a model book known as Unshū shōsoku (The Izumo Letters, ca. 1058).66 Akihira was appointed governor of Izumo Province in 1057, and it was during this time that he wrote 209 letters on a variety of topics. Two letters, for example, describe a party in Uji attended by courtesans singing imayō (modern-style song) and plans for a visit to the pleasure quarters of Eguchi. One source of inspiration for exploring the periphery in Heian society was undoubtedly the poetry of Chinese literati such as Bai Juyi and Yuan Zhen (779 – 831), who with their new ballads created a genre that centered on the marginal in Chinese culture. These ballads were long poems in a relatively free form, and in a simple tone they criticized social and political wrongs or described other lamentable situations in society. That Japanese nobility should develop an interest for the xinyuefu is an aspect of Heian culture that has been hitherto overlooked. In other words, why would Kyoto nobles who emulated much of Chinese culture but disdained its politics become enamored of a genre that was invented for political purposes? Puzzling as it may be, the new ballads were nevertheless a great hit with the Heian nobles. Together with verses from Bai’s Song of Everlasting Sorrow (Changhen ge, J. Chōgon ka), their lines were singled out for recitation. Women, too, enjoyed reciting Bai Juyi’s poetry, as several passages in The Pillow Book indicate. As we have seen, even though the practice and study of Chinese literature was considered a male prerogative, Murasaki Shikibu taught Imperial Dowager Shōshi to read Bai Juyi’s new ballads. Again and again, when excerpts of Bai Juyi’s poetry were singled out for copying, it was practically always the new ballads that were chosen.67 The courtiers even organized study sessions to discuss the xinyuefu.68 In fact, when Heian nobles mentioned ballads they invariably meant Bai Juyi’s new ballads. The simple language of the new ballads had much to do with their popularity in Heian Japan. Bai Juyi himself emphasizes in his preface that clarity of meaning is far more important than literary style in new ballads. It is also clear that the Heian courtiers did not care much for the political implications of the Chinese examples, but they used and imitated the descriptive passages in their poems. They depoliticized the contents of the new ballads in order to appreciate them within a Japanese context. As a result, Heian descriptive poems and narrative accounts are void of the moral messages that are found in the Chinese texts that were used as models. Not everybody was happy about this drift toward what they believed to be the seamier side of the Heian world. In reference to Ōe no Masafusa’s penchant for gossip and his accounts of courtesans, itinerant entertainers, unruly city mobs, and fox spirits, Fujiwara no Munetada noted in his diary, A certain person has told me that in the last two or three years Governor Ōe has been unable to walk and no longer goes out to attend his official duties. But whenever someone comes to call he writes down all the miscellaneous gossip of the town. There are numerous errors, and much passing of judgment about the merits of others. Such indiscriminate scribbling about worldly mat-
120 | ivo smits
ters is surely the height of impropriety! Recording from hearsay what one has not witnessed nor come to know with certainty is an outrage through and through. Great Confucian he may be, but the world can hardly approve of his behavior.69 Munetada disapproved, in other words, of Masafusa’s recordings of unusual practices and people on the periphery of Heian culture. However, it was exactly in its more peripheral aspects that kanbun sowed the seeds of later developments in both kanshi and waka. The use of marginal groups in Heian society as subject matter for literary texts was consolidated in the late eleventh and twelfth centuries when more kanshi were composed on itinerant entertainers (kugutsu) and charcoal burners, and it was through such a setting that they found their way into waka literature. In fact, courtesans, kugutsu, and female divers even became a set topic for such formal settings as waka contests in the late twelfth century.70 Classical East Asian poetics have since early on operated along imitative principles. Generic genre codes and the pervasive literariness of the acts of reading and writing forced a continuous reference to past texts. However, Heian literati incorporated functional conventions of Chinese poetry to create a new genre of the social periphery in Japanese literature.71 Calls for social reform in the Chinese model were transformed into observations of exotic subject matters in the Japanese emulations. Such acts of misreading the Chinese model, deliberate or culturally determined, helped clear poetic space and broaden the field of subject matter; and by putting the Heian social fringe on the textual map, such misreading brought the periphery into closer contact with its center.
The Periphery of Decorum: Eroticist Parody There were more peripheries on the textual map than just the social fringe. Another uncharted terrain of kanbun writing, small but significant, is eroticist and pornographic parody. It is a genre not quickly associated with Heian literature and therefore worth exploring, as it pointedly shows us how insights in the possibilities of kanbun writing may affect our old views of what constituted court literature. The ironic and technical skills of the scholar-poet Fujiwara no Akihira extended to parody. While this is already apparent in A New Account of Sarugaku, it is also evident in Tettsui den (The Biography of an Iron Hammer), which he wrote under the pseudonym Dick Large (Ra Tai; Ch. Luo Tai). Not overly subtle, this work counts as one of Japan’s earliest works of pornography. Although it was part of the Honchō monzui, probably the most important anthology of Heian kanbun writing, it has been largely neglected and marginalized by later scholars.72 The text is an arrangement of many variant descriptions of the male organ and its uses. Interestingly, it is written in a style that adheres to the rules of biographical genre in traditional Chinese dynastic histories, and the diversity of erotic terminology seems in part derived from Chinese manuals of medicine.73 The iron hammer metaphor also appears in Shin sarugaku ki’s description of the nit-
The Way of the Literati | 121
wit husband of the fourteenth daughter, whose huge penis seems to be his only asset. By informing his readers in the introduction that he hopes “to make you chuckle,”74 Dick Large shows that his aim is parody. The question of why Akihira included it in Honchō monzui, an anthology that aimed to present its readers with exemplary models, is puzzling. The fact that the only other biography included is that of a priest, albeit a mythical one,75 makes its inclusion all the more curious. One could simply see this as yet another manifestation of Akihira’s eccentricity, but one may also regard it as indicative of his antagonism toward the stricter attitudes of a Confucian worldview. From this perspective, both the Shin sarugaku ki and Tettsui den function to open up new territories on the map of kanbun texts, with their models showing not only a variety of stylistic possibilities but also a richness of subject matter that went beyond traditional notions of decorum.76 Akihira was not the first to write about the periphery of Heian culture, as indicated above, nor was he the only Heian scholar to try his hand at eroticist parody. The best-known example of the latter is Danjo kon’in no fu (A Poetic Exposition on the Marriage of Man and Woman), by the early tenth-century scholar Ōe no Asatsuna, which mixes literary language with relatively explicit descriptions of physical passion.77 Put together, the rhyme words of this long poem spell out the phrase “Love and emotion bring mutual response, and afterward the body becomes pregnant,” which neatly sums up the theme of this piece. The late Heian Honchō zoku monzui (Literary Essence of Our Court, Continued, ca. 1140s) contains in the section Hymns (san) another parody full of sexual innuendo, presumably inserted as an echo of Akihira’s Tettsui den. It is titled Insha no san (Praise of the Penis) and attributed to “the Captain’s Upright Glans in Vaginal Juices.”78 Unlike Tettsui den, this piece is not anonymous; at the end it is signed by Fujiwara no Suetsuna (? – before 1102) and dated the twelfth month of Kahō 1 (1094). A kanshi poet and scholar who later became head of the State Academy, Suetsuna was a son of the idiosyncratic scholar Sanenori, who together with Akihira was chastised for prompting examination candidates as noted earlier. Insha no san is about a carriage and is littered with numerous technical references to spare parts, but as the pseudonym makes clear, the text is to be read as a series of sexually suggestive terms. The textual strategy seems to follow the same pattern, juxtaposing a literary style that presents itself as heavily indebted to ancient Chinese writing with a subject matter that borders on, or crosses over into, the seedy. This is a classic recipe for parody. In these kanbun examples the parody is enforced by the use of established literary genres. Tettsui den follows the given format of the Exemplary Biography model from Chinese dynastic histories, and Insha no san presents itself as just another example of the Hymn genre. Ultimately, Danjo kon’in no fu, too, is a parody, of the poetic exposition, or fu, which in East Asia was a set feature in examination assignments. The use of this form was, by Heian standards, a deliberate degradation of a type of poetry that most Confucian scholars approached with solemnity. This pattern is also evident in the kanshi satire O naki ushi no uta (or Mubigyū ka, Song of the Tailless Ox), in which Minamoto no Shitagō (911 – 983) sings the praise of his disfigured ox.79 The juxtaposition
122 | ivo smits
of a lofty style with mundane subject matter indeed lends itself equally well to satire. As such it was a strategy adopted by Murasaki Shikibu when she described a daughter from a literati household whose talk (full of pompous phrases) and breath (she smells of garlic) suggest the Asian continent (kara).80 While one may wonder about the possible agendas of the respective authors — anything in the range between academic locker-room jokes and the subversion of Confucian values — there is no denying that these texts make use of possibilities apparently not really open to writing in the vernacular. Satire was possible in kana writings of the Heian period, but erotic parody was not.81 One possible exception might be a passage, full of punning, in Tosa nikki (A Tosa Journal, ca. 935) in which women in the entourage of a provincial governor, on his way back to the capital, display their genitals when taking a bath.82 To what extent author Ki no Tsurayuki’s (ca. 872 – 945) ploy to present the narrator as a woman imitating a kanbun diary in a kana text might have muddled possible ideas of decorum and language remains an open question.
Conclusion Surveying the large corpus of kanbun literature of Heian Japan, one must acknowledge its ideological centrality in the contemporary view of the canon of Japanese texts. This centrality gave rise to the formation of a canon of indigenous kanbun, a new textual past inhabited by Japanese authors in which a China was no longer a necessary entity. It also suggested a self-sufficiency of Japan’s literary tradition: Japan’s literature had become a center unto its own. That same ideological centrality made it possible for literati to develop Chinese learning into an expertise that translated itself as cultural capital: in an era when cultural models appeared to follow the new courtly code of miyabi, which resisted meritocratic notions, dedication to kanbun proficiency managed to be the acceptable exception. Literati molded learning into a “way,” and as such heralded a pattern of cultural behavior that would become dominant only in medieval Japan. The nature of Heian kanbun texts is diverse. The majority of texts reflect kanbun’s important role in matters of ceremony and state. However, the Heian perception of texts, especially poetry, as reflections of the situation in which they were composed, provides important clues as to when one may expect literary experiments: that is, in informal settings and a fairly strictly circumscribed number of genres, such as non-verse-topic poems, narrative accounts, or poem prefaces. It is especially within the literary free zone that these genres and situations comprise that one finds kanbun texts that challenge our deep-rooted ideas of Heian literature in general. The traditional emphasis of literary and cultural historians on tale literature and poetry in Japanese obstructs our view of a court literature that, among others things, showed an active interest in a world beyond court society. More generally, it obscures Heian’s linguistic register, for writers in classical Japan could say many things, and they did so in more than one language.
The Way of the Literati | 123
Notes 1. For an insightful analysis of this process, see Kurozumi Makoto, “Kangaku: Writing and Institutional Authority.” Since the late nineteenth century, the terms kanshi (poetry in Chinese) and kanbun in Japan refer to any text in Chinese, usually from China. In keeping with a not entirely logical tradition among Western scholars, I use them exclusively in the sense of poems and prose in Chinese composed by Japanese. Japanese scholars commonly refer to “Japanese kanshi” (Nihon kanshi) or “kanshi from the court period” (ōchō kanshi). Kanbun refers to the language in which a text is written (Chinese, or Sino-Japanese); the corresponding script could in classical Japanese be called mana. However, this term is nonexistent in kanbun texts. The exception are kanbun prefaces to waka anthologies — in other words, within a kana context. Mana was not part of the vocabulary of Heian literati. They referred not to script but to literary categories: bun (rhyming prose), hitsu (rhymeless prose), shi (rhyming poem), fu (poetic exposition), ku (couplet), etc. 2. Haruo Shirane, “Issues in Canon Formation,” 1 – 5; Genji ippon kyō, 37. 3. Meigetsuki, Karoku 1 (1225) 2/16. Teika mentions that his previous, carefully collated edition of The Tale of Genji was “stolen in the Kenkyū era (1190 – 1199).” See also T.J.Harper, “Genji Gossip,” 29. 4. This incomplete but oldest extant catalogue of a private library is Tsūken nyūdō zōsho mokuroku. 5. Honchō mudaishi, 7, no. 463. Honma Yōichi, Honchō mudaishi zenschūshaku, 2:429 – 432. The remark between parentheses is a note Renzen himself added to this line of his poem. 6. Other examples of literati who were chosen to serve the emperor as advisers include Sugawara no Kiyokimi (770 – 842) and Miyoshi no Kiyoyuki (847 – 918). See Francine Hérail, La cour du Japon à l’époque de Heian, 56; and Robert Borgen, Sugawara no Michizane and the Early Heian Court, 48 – 49, 135. 7. Joshua S. Mostow, “Mother Tongue and Father Script,” 135. 8. See, for example, Tomiko Yoda, “Literary History against the National Frame,” 485 – 486. 9. Hérail, La cour du Japon à l’époque de Heian, 47 – 57. 10. Chōya gunsai, 3:60. 11. Only the preface is extant, preserved in Chōya gunsai, 1:18. 12. Only books 7 and 9 are extant. 13. The first part of book 1 is missing. 14. Marian Ury, too, noted that Chinese learning in Heian Japan increasingly became a selfcontained tradition, finding sufficient roots to craft a past of its own (Ury, “Chinese Learning and Intellectual Life,” 341, 389). 15. Sakumon daitai provides detailed analyses of various aspects of kanshi structures and the handling of set verse-topics. Munetada, who built on an earlier manual, gives extensive rules for the use of tone patterns within the poem as well as for parallelism. Tekkinshō contains 460 couplets by 170 Heian poets and two Tang poets. The manual was possibly produced by Fujiwara no Takanori (1158 – 1233). 16. For a brief overview in English of the historical orientations of the prefaces to the three ninth-century collections, see Wiebke Denecke, “Chinese Antiquity and Court Spectacle,” 106 – 110. 17. Chōya gunsai, 18:30. Two minihistories of Heian kanshi are Ōe no Koretoki’s preface to the Nikkanshū and Ōe no Masafusa’s Shikyōki. 18. The Chinese (mana) preface to the Kokinshū mentions one Kaifūsō-Man’yōshū poet by
124 | ivo smits name, Prince Ōtsu (663 – 686). He is singled out as an outstanding kanshi poet. Ōtsu left four kanshi and four waka. 19. One of the better-known exponents of this notion is Chino Kaori. See, for instance, her “Gender in Japanese Art,” 23 – 25. See also David Pollack, The Fracture of Meaning, 55 – 76, for such a use of China by Murasaki Shikibu. 20. Thomas LaMarre, Uncovering Heian Japan, 33. 21. Yoda, “Literary History against the National Frame,” 486 – 487. 22. Eiga monogatari, 6, in NKBT 75:199; translation in William H. McCullough and Helen Craig McCullough, A Tale of Flowering Fortunes, 1:217. See also Eiga monogatari, 8, 19, and 27; NKBT 75:287, 76:104 – 105, 259; translations in McCullough and McCullough, Flowering Fortunes, 1:301, and 2:584, 712. 23. Murasaki Shikibu nikki, 501; Richard Bowring, trans., The Diary of Lady Murasaki, 58. 24. Some six female kanshi poets may be identified for the early ninth century, the only ones for the whole Heian period (Kumagai Naoharu, “Saga chō no joryū sakkatachi,” 71 – 76). 25. The preface to Asatsuna’s Wachū setsuin (939), a now lost kanbun manual, opens as follows: “Now, the way of scholarship puts the composition of literature first. If you recite the Classics, but do not learn to compose poetry, then you may be called a mere bookcase and you will be useless.” Quoted in Sakumon daitai, 352. 26. Shōyūki, Chōwa 1 (1012) 7/17. 27. E.g., Konishi Jin’ichi, “Michi and Medieval Writing.” 28. See Robert N. Huey, “The Medievalization of Poetic Practice,” 651 – 652. 29. Makura no sōshi, 88, 137 – 138; Ivan Morris, The Pillow Book of Sei Shōnagon, 1:91 – 92, amended. The passage echoes another section in The Pillow Book, one that betrays Sei Shōnagon’s familiarity with literati views of the canon, although, interestingly, she mentions only prose genres as important Japanese writings in Chinese: “Texts in Chinese are: Bai Juyi’s collected works, Wenxuan, new-style poetic expositions, Records of the Grand Historian, Annals of the Five Emperors, religious dedications, petitions to the throne, formal requests drawn up by learned professors” (Makura no sōshi, 211, 249). 30. Tametoki’s poem to the Song merchant is Honchō reisō, no. 131. Zhou’s reply is no. 132. Honchō reisō kanchū, 323 – 326. Zhou was shipwrecked on the Japanese coast and was forced to stay in Japan for seven years. 31. Tanahashi Mitsuo, Ōchō no shakai, 120 – 121. One more example would be Akazome Emon, the alleged author of Eiga monogatari, who married Ōe no Masahira. For further details on women authors and their learned fathers, see Mostow, “Mother Tongue and Father Script.” 32. Takemura Shinji, “Chūshaku no genjutsu,” 113 – 114. 33. The Kangakuin was one of several clan-run institutions that started out as combined dormitories and cram schools for students at the State Academy, but gradually evolved into semi-independent colleges. Other private schools also existed. 34. On the State Academy, see Borgen, Sugawara no Michizane, 71 – 88; Ury, “Chinese Learning and Intellectual Life,” 367 – 375; Momo Hiroyuki, Jōdai gakusei no kenkyū; and Hisaki Yukio, Daigakuryō to kodai jukyō. I benefited from Robert Borgen’s personal communications as well. 35. For more detail, see Ury, “Chinese Learning and Intellectual Life,” 346 – 352. 36. For a partial English translation of the Mengqiu, see Burton Watson, Meng ch’iu. 37. Baiyong, which actually contains 120 poems, is an anthology of poems on simple topics, by Li Jiao (644 – 713), a poet who excelled at the genre of poems about things (yongwu shi; J. eibutsu shi).
The Way of the Literati | 125 38. For a translation of Wakan rōeishū, see Rimer and Chaves, Japanese and Chinese Poems to Sing. For general background, see Ivo Smits, “Song as Cultural History”; for use of Wakan rōeishū as illustrative of Heian calligraphic and discursive space, see LaMarre, Uncovering Heian Japan, 116 – 139. 39. See David McMullen, State and Scholars in T’ang China, 207. I am indebted to Dr. Oliver Moore, Leiden University, for a stimulating discussion on Tang examinations. 40. See also Momo, Jōdai gakusei no kenkyū, 87 – 92; and Ury, “Chinese Learning and Intellectual Life,” 370. Such terms as governor seem to have usually been absentee (yōnin) appointments. 41. For Tokimune’s controversial poem and the relevant letters by his examiners, parts of which are quoted in the paragraphs below, see Honchō monzui 7:232 – 240, nos. 176 – 179; and Gonki, Chōtoku 3 (997) 7/19. See also Honchō monzui chūshaku, ed. Kakimura Shigematsu, 960 – 1002; Kawaguchi Hisao, Heian chō Nihon kanbungaku shi no kenkyū, 535 – 536; Ōsone Shūsuke, Nihon kanbungaku ronshū, 2:322 – 323; Kinpara Tadashi, Heian chō kanshibun no kenkyū, 345 – 351, 354 – 390; and Ceugniet, L’office des études supérieures, 30 – 33. 42. In Honchō reisō kanchū, Yanagisawa Ryōichi (392) suggests that Tokimune was born about 985. If true, he would have been about thirteen when he took his exam, which is prohibitively young although there is historical evidence that some boys from scholarly families started learning to compose kanshi at a very early age. 43. Examination poems, and therefore most poetry examples in manuals, consisted of five-character verses (gogonshi), whereas the majority of eleventh-century kanshi are sevencharacter verses (shichigonshi). In other words, in practice Heian poets composed lines of seven characters, while the State Academy program held on to the old ideal of five characters per line. It is a common discrepancy between the classroom and the world outside it. 44. For Masahira, see also Francine Hérail, “Un lettré à la cour de l’empereur Ichijō”; Kawaguchi, Heian chō Nihon kanbungaku shi no kenkyū, 572 – 589; and Ōsone, Nihon kanbungaku ronshū, 2:315 – 342. 45. Nihon kiryaku, Chōtoku 2 (996) 12/26. The Tiantai monk Yuan Qing had arrived in Japan in 995 to request from the chief Tendai abbott (zasu) Senga copies of Buddhist texts no longer preserved in China. The nature of the court’s reply to him is unknown. 46. For Tadana, see Kawaguchi, Heian chō Nihon kanbungaku shi no kenkyū, 533 – 542; and Kinpara, Heian chō kanshibun no kenkyū, 324 – 352. In 1000 Michinaga was given a copy of Fusōshū by Tadana’s widow. The general assumption is that Michinaga later presented this anthology to the throne (Midō kanpaku ki, Chōhō 2 [1000] 2/21). 47. The expression comes from Han Feizi (third century B.C.) and was popular with scholars throughout East Asia. 48. “To enter the Dragon Gate” referred to success at the examinations. 49. Ōsone, Nihon kanbungaku ronshū, 2:323. 50. Honchō monzui 7:232 – 240, nos. 176 – 179; Chōya gunsai, 13:276 – 278. 51. Masafusa’s interviewer, Fujiwara no Sanekane, added, “It is not just Tadana’s poetry, he also embellishes his prose with bits from old collections.” Masafusa talks about Tadana’s bun and ku, meaning prose and verse with rhyme (in), whereas Sanekane’s comment refers to Tadana’s hitsu or rhymeless prose. The distinction between rhyming and rhymeless was perhaps more important than such categories as “poetry” and “prose” (Gōdanshō 5:204, 535, no. 62). 52. It would seem that ke (informal) did not function as a category equal to hare (formal) in Heian Japan. One could speak about formal poems (hare no uta) but not about informal poems (ke no uta). For the late Heian period, this argument is made persuasively by Nishiki Hitoshi in “Insei ki utaawase no kōzō to hōhō,” 27 – 29.
126 | ivo smits 53. Chōya gunsai, 13:270 – 273. For more detail, see Horikawa Takashi, “Shi no katachi, shi no kokoro,” 121 – 126. 54. For an excellent overview of sekiten in Nara and Heian Japan and the Chinese poems composed at banquets following the ritual, see I.J.McMullen, “The Worship of Confucius in Ancient Japan.” 55. With “study groups” I do not mean the more or less formalized gatherings to encourage learning (kangakue). I am thinking of more informal and loosely organized gatherings, often at someone’s home and at times including no more than three or four people. These meetings generally belong to the dokusho (reading) category. 56. The Kitano Shrine, immediately northwest of the capital, was founded in 947 as a place where Tenjin, the deified Michizane, could be worshiped. The Kichijōin was built in 881 by Michizane in memory of his father Koreyoshi (812 – 880). In 1066 (some sources have 934) a new Tenjin Hall was added to the Kichijōin (var. Kisshōin), where one could worship Michizane. This hall was commonly referred to as Tenjin Gyodō or Seibyō. See Borgen, Sugawara no Michizane, 308 – 326. 57. The kangakue, first organized by Yoshishige Yasutane (931? – 1002) between 964 and 986 but resuscitated and organized until 1122, were Dharma gatherings (hōe), at which literati discussed matters of faith with monks and read Buddhist scriptures; afterward, poems on religious topics were composed. 58. This was the same examination, the difference being that when it was held at the Suzakuin, the emperor was present to hand out the topic for the poem and students were rowed to an island in the middle of the pond for poetry composition. 59. Sometimes an even finer distinction was followed. Topics consisting of four characters were called hikudai, reserving mudai for topics of six or seven characters. See the definition in the Ōtaku fukatsushō (1276)‚ a kanbun treatise by the Shingon monk Ryōki. 60. Sakumon daitai, 365. Several anthologies of poem prefaces, both for kanshi and waka, were compiled from the twelfth century onward. To name but three that are extant: Fusō kobunshū (Collection of Japanese Old Prose, mid-twelfth century), Honchō shōjoshū (Small Prefaces of Our Court, date unknown), and Shijoshū (Collection of Poem Prefaces, shortly after 1132, comp. Fujiwara no Atsumitsu, 1063 – 1144). 61. Tanahashi, Ōchō no shakai, 121 – 127. 62. Sakeiki, Chōgen 7 (1034) 11/25, 12/9. See also Naumann, “Enzyklopädische und emanzipatorische Züge,” 453. In general on Akihira, see Ōsone, Ōchō kanbungaku ronkō, 18 – 55. 63. A German translation is available in Hagen Blau, Sarugaku und Shushi, 333 – 358. Sarugaku (lit. “monkey music”) was a form of popular entertainment that seems to have combined dancing, acrobatics, singing, and skits. Akihira’s “Account,” however, is not about sarugaku itself but about the people who come to watch it. They are a blend of low-ranking courtiers and commoners. The name Uemon no jō (lit. “Captain of the Right Gate Guards,” a post open for those of sixth or fifth court rank) and the positions of some of his kin suggest the man has a low-rank courtier background. 64. Honchō monzui 1, no. 18, and 9, 132 – 272, nos. 237 and 238. For a similar observation, see Jacqueline Pigeot, Femmes galantes, femmes artistes, 58. 65. Kanke bunsō, 200 – 209. NKBT 72:259 – 265. For a translation of this sequence, see Burton Watson, Japanese Literature in Chinese, 1:93 – 94; and Borgen, Sugiwara no Michijane, 187 – 188. 66. Also known as Meikō ōrai (Akihira’s Model Correspondence). The two letters in question are nos. 117 and 118. See Unshū shōsoku, 173 – 175. 67. See, for example, Midō kanpaku ki, Kenkō 1 (1004) 9/7: the famous calligrapher Fujiwara no Yukinari (972 – 1027) brings Michinaga “the first book of the ballads” that he has copied
The Way of the Literati | 127 out. Eight days later he brings Michinaga the second book, so that now Michinaga has the complete set of Bai’s new ballads. 68. Regent Fujiwara no Tadamichi (1097 – 1164) organized such a study session, which resulted in a small series of Bai Juyi imitations in shi form. Hosshōji dono gyoshū, 254 – 255, nos. 85 – 89; Ivo Smits, “Reading the New Ballads,” 175 – 182. In his Tsurezuregusa (Essays in Idleness), Kenkō (1283 – 1352) mentions one of these xinyuefu study sessions hosted by Emperor Go-Toba (1180 – 1239; r. 1183 – 1198). Tsurezuregusa, 226: 271 – 272. 69. Chūyūki, Kajō 2 (1107) 3/30; Marian Ury, “The Ōe Conversations,” 359. Ury quotes this passage in reference to what was to become Gōdanshō, but I would suggest that it has broader implications, extending to Masafusa’s records of unusual practices and people. 70. E.g., Roppyakuban utaawase (ca. 1194). The topics for nos. 1141 – 1202 are “Love for courtesans,” “Love for kugutsu,” “Love for divers,” “Love for woodcutters,” and “Love for merchants.” 71. For a more detailed case study of how late Heian kanshi poets used the intratextual poetics of the xinyuefu and explored the social margin as subject matter, see my “Reading the New Ballads.” 72. Honchō monzui, 12, no. 377; SNKBT 27:337 – 339, or NKBT 69:429 – 436 (annotated yomikudashi edition). The only studies that make any use of Tettsui den at all are Naumann, “Enzyklopädische und emanzipatorische Züge,” 455 – 460; Ōsone, Ōchō kanbungaku ronkō, 137 – 138; and Tanahashi Mitsuo, Kodai to chūsei no hazama de, 39 – 57, 64. Burton Watson characterized this work thus: “Interesting as it may be in conception, it turns out to be one of the most tedious works of pornography in all literature.” Watson, Japanese Literature in Chinese, 1:53; also quoted in Donald Keene, Seeds in the Heart, 348. Kakimura Shigematsu (1879 – 1931) was apparently so scandalized by Tettsui den, as well as Danjo kon’in no fu (see text below), that he omitted them altogether from his 1922 edition of Honchō monzui. 73. Heian scholars were aware of the existence of erotic literature in China. They preserved the oldest known example from Tang China, You xianku (A Dalliance in the Immortals’ Den, J. Yū senkutsu), a text from the 690s that “engages in a sort of literary and sexual boasting in front of a male audience,” to quote Paul Rouzer. For a discussion and translation, see Paul Rouzer, Articulated Ladies, 204 – 216, 313 – 354. However, while You xianku’s vocabulary may have influenced certain Heian poems, it does not seem to have affected the more explicit parodies treated here. 74. Honchō monzui, NKBT 69:430. 75. Dōjō hōshi den (Biography of the Priest Dōjō) by Miyako no Yoshika (834 – 879) retells the story of the miraculous Dōjō, a monk of Gangoji in Nara, who was famous for his incredible strength. The story also appears in Nihon ryōiki (1, no. 3) and seems to be part of a cycle of legends from Owari Province. 76. By the looks of it, Akihira’s inventive streak did not extend to waka. Two conventional waka by him were posthumously included in the fourth imperial anthology, Goshūishū (Later Collection of Gleanings, 1086; nos. 166 and 423). 77. Honchō monzui, 1:15; SNKBT 27:130 – 131 (see NKBT 69:340 – 344, for an annotated yomikudashi version); Watson, Japanese Literature in Chinese, 1:53 – 56. 78. Insui kōi kōkō (Honchō zoku monzui, 11:726). “Kōkō” might also be read “Takakari,” and some hold that it actually refers to someone’s name. However, it is more likely a question of word play. On the one hand, the word kōkō (Ch. gao hong) suggests a Chinese name, such as Chen Hong (fl. ca. 813), the author of Changhen ge zhuan, a prose version of Bai Juyi’s “Song of Endless Sorrow.” After all, kōi (captain) was an office in the ancient Chinese bureaucracy. Another association might be with the term kōju, which refers to a gifted Confucian scholar. On a different linguistic level, kō (or kari) is the equivalent of gan (or kari, “goose”), which
128 | ivo smits was an abbreviation for gantō (or karikubi in Edo texts), “goose head,” meaning the glans of a penis. E.g., Tettsui den, NKBT 69:435, 436; Shin sarugakuki, 146, 305 (“his glans is raised high,” echoing the pseudonym Takakari/Kōkō). 79. Honchō monzui, 1, no. 41, SNKBT 27:135 (see NKBT 69:359 – 363, for an annotated yomikudashi version); Watson, Japanese Literature in Chinese, 1:65 – 67. 80. Genji monogatari, vol. 1; NKBT 14:83; Royall Tyler, trans., The Tale of Genji, 34. 81. Donald Keene stresses a dichotomy between waka, in which longing is expressed “so indirect that its erotic aspects might pass unnoticed,” and poetry in Chinese as the only language available for explicit erotica. Keene, Seeds in the Heart, 346 – 347, 348. 82. Tosa nikki, 38; for a somewhat restrained translation, see Helen Craig McCullough, Classical Japanese Prose, 83 – 84.
6
d Edward Kamens
Terrains of Text in Mid-Heian Court Culture
I
n Murasaki Shikibu’s semiprivate living quarters (ca. 1008 – 1010), there were two cupboards. One was “crammed to bursting point” with “old poems and tales” (furu uta, monogatari); the other was full of miscellaneous Chinese books (fumi domo) left to her by her late husband. It was to the latter, she reports in her diary, that she was drawn in those times when idleness weighed upon her (tsurezure semete amarinuru toki): “Whenever my loneliness threatens to overwhelm me, I take out one or two of them to look at; but my women gather together behind by back. ‘It’s because you go on like this that you are so miserable. What kind of lady is it who reads Chinese books?’ (nadeu onna ka mannabumi wa yomu).”1 The lady in question had in fact, according to her own account, been reading Chinese texts since childhood, but, in anticipation of such censure, had long “avoided writing the simplest character.”2 And at yet other times when she sought distraction from her personal woes (mi no usa), she says, she had done so in the reading and rereading of monogatari — perhaps including her own, The Tale of Genji.3 From just such frequently cited passages as these, in a variety of literary texts of the Heian period, scholars have constructed an image of cultural practices configured by conspicuous borders that barred or at least strongly deterred women from encounters with “Chinese” texts, while men remained free to move at will across and among a variety of textual domains. Such passages, juxtaposed and contextualized in various ways, give support to the notion that Heian women were discouraged, and even discouraged one another, from reading Chinese texts. Other passages, almost always including the deliberately playful gender masquerade with which Ki no Tsurayuki’s Tosa nikki (ca. 930) begins, have provided the basis for the perception that Heian women were likewise barred from, or barred themselves from, the writing of Chinese texts as well. And yet there are many well-known exceptions that complicate this picture. One good example is the Great Kamo Priestess (Daisaiin) Senshi’s preface to her collection of Buddhist poems in Japanese, Hosshin wakashū (dated 1012), in which, writing in Chinese, she nevertheless cites her sense of alienation from “Chinese letters” (kanji) — as well as Sanskrit writing (bongo) — as the reason for her choice of “the thirty-one-syllable Japanese poem” (sanjūichiji no uta) to serve as her votive medium.4 To my knowledge, no one has ever seriously suggested that Senshi is not the author of this preface, or that its juxtaposition of languages is aberrant or accidental; rather, along with the many quotations from Buddhist scripture — in Chinese, of course — that stand beside her poems in the Hosshin wakashū text,
130 | edward kamens
these multiple languages coexist in the space of this one work, performing their respective, complementary roles and interacting to form multiple tiers of signification.5 This juxtaposition may provide us with another telling model for the mid-Heian period’s textual terrain. Relatively recent historical conditions, some of which shall be discussed below and which are also discussed in the essay by Ivo Smits, have placed Japanese writings (wabun, kana bungaku), especially those by women, at the center of most conventional perceptions of Heian literary culture, with Chinese writings (kanbun) at the periphery — a construction that in many ways runs counter to historical reality, insofar as we are able to recover it through the reading of texts themselves. The purpose of this essay is to complicate this model further, to show its limitations, and to offer some alternative ways of perceiving how the literary terrain of mid-Heian Japan was encountered, accessed, and in some cases altered by at least some of the men and women who traversed it. The art historian Chino Kaori often spoke and wrote about what she called the “dual binary structure” of the culture of the Heian period. In her published work and many lectures in Japan and abroad, Chino developed this conception as a way of understanding how Chinese writing, Chinese painting, and other cultural forms of continental origin found their places within the complex amalgam that we encounter as Heian court culture. Chino’s dual binary structure was, for her, a way of analyzing some of the interactive relationships between kara-e and yamato-e (“Chinese” or “Japanese” pictorial images and styles), kanji and kana writings (that is, “Chinese” or “vernacular” orthographies), otoke-de and onna-de (“male” or “female” calligraphy), hare/ke, omote/ura (“public” vs. “private” or “outward” vs. “inward”) configurations, and much more.6 She also spoke and wrote often of the existence of a “China within Japan,” a conception that in and of itself complicates conventional notions of native and foreign, self and other, center and periphery.7 While Chino is no longer with us to share what promised to be a significant further development of her vision of Heian culture, particularly in terms of the meaning of this dual binary structure in relation to issues of gender, one particular statement she made in a workshop presentation at Harvard University in September 2001 resonated powerfully at the time and provided me with the point of departure for this discussion. As a way of illustrating the interactive role of Chinese writings within the broader scope of mid-Heian literary culture, Chino noted that (and I paraphrase here), “Murasaki Shikibu and Sei Shōnagon acted as if they did not know Chinese (shiranai furi o shite) [because of the identification of Chinese writings with masculinity and officialdom], but in fact their writings reveal the active presence of Chinese texts in their literary experience.” Chino was of course not the first to remark upon the ways that the authors of Genji monogatari, Murasaki Shikibu nikki, and Makura no sōshi (The Tale of Genji, The Diary of Lady Murasaki, and The Pillow Book of Sei Shōnagon) demonstrate that they did have extensive access to and a flexible command of an array of canonical Chinese texts; it was simply her way of describing their sometimes coy denial of that knowledge (shiranai furi o shite) alongside their ample
Terrains of Text in Mid-Heian Court Culture | 131
demonstrations of a fairly sophisticated familiarity with Chinese writings that struck such a vibrant chord with me. I recall Chino’s formulation, for example, when reading the section of Makura no sōshi that begins, “On the tenth day of each month the empress [Teishi] ordered that dedications of sūtras and of images be made on behalf of the late chancellor” [her father, Fujiwara no Michitaka, 953 – 995] (kotono no ontame ni, tsukigoto no towoka, kyau hotoke nado kuyau sesasetamahishi wo).8 This diarylike episode relates that, after the memorial observance in the ninth month (four months after Michitaka’s death in 995), some of the male participants remained together to drink wine and recite Chinese poems (sake nomi, shi zuji nado suru ni). When Sei Shōnagon and the empress heard one of them, Fujiwara no Tada nobu (967 – 1035), chant a brief line from a memorial Chinese verse composed by Sugawara no Fumitoki (899 – 981), a Japanese poet of an earlier generation, they showered Tadanobu with lavish praise for his strikingly adept and appropriate quotation: “It was splendid, and I wondered how he could have thought of such an appropriate passage” (hata imijiu medetashi, ikade sa ha omohiidetamahikemu), exclaims Shōnagon. The empress concurred: “Wasn’t he magnificent,” she said. “Those lines were just right for the occasion” (medetashi na. imijiu kehu no reu ni ihitarikeru koto ni koso are). Tadanobu could not have won such praise had his citation of the verse in question (“Where is he now/when moon and autumn have returned at the appointed time?” tsuki, aki to ki shite mi izuku ka) not been readily recognized, not only by the gentlemen in attendance but also by the women listening to them. Thus, it is clear that this verse — and perhaps the entirety of Fumitoki’s poem, composed on the first anniversary of the death of Fujiwara no Koretada (or Koremasa, 924 – 972) — was locatable in a terrain of text shared by both the men and women present on this occasion. In the Makura no sōshi telling of this episode, there is no obstacle delaying or deterring Teishi’s or Shōnagon’s access to the significance of Tadanobu’s utterance. One man’s words of lamentation are instantaneously recognized, when repeated by another, by the person to whom they are likely to mean the most — the mourning Teishi — and by her literary companion Shōnagon as well. Nothing impedes the communication intended here; no one need pretend that they do not understand. The resonance of Tadanobu’s utterance was no doubt further enhanced by the fact that the esteemed Koretada was Michitaka’s (and Tadanobu’s) paternal uncle; and Shōnagon’s inclination to praise Tadanobu was no doubt at least partly grounded in her amatory relationship with him — the somewhat troubled course of which becomes the topic of the remainder of this section. So Tadanobu’s timely and adroit quotation impressed not only as a sign of his literary acumen but also as a sign of his political astuteness and because of the personal regard of at least one of his listeners — here, the one who is telling us this anecdote. Tadanobu serves extraordinarily well as an example of an historical figure active in the intercultural and multilingual textual field or terrain of these times. As we can sense from this anecdote, this field found form and expression not
132 | edward kamens
only in the medium of written and read text but in orally uttered and heard text as well. Tadanobu’s literary activities, engaged in not as excursions or diversions from his court duties but as inherent extensions of them and as formalizations of his political and social relationships, range widely across this field in a manner that I think should be seen as normative. His dexterity in moving across this panorama of forms, languages, and contexts should not be seen as anomalous but rather as illustrative of the open, interactive, cross-referencing field of activity engaged with and accessed in various ways by both men and (at least some) women of the time. Like some other figures of his time, such as Fujiwara no Kintō (966 – 1041) — who also appears from time to time in literary interactions conducted in multiple languages with women such as Murasaki Shikibu and Sei Shōnagon in their own writings — Tadanobu emerges as, among other things, a mediator across and among literary practices and gendered spheres of cultural action. For reasons that have to do with the ways that the canon of classical Japanese literature has been constructed over time (for the most part, in relatively recent time), as shall be discussed below, it is Murasaki Shikibu and Sei Shōnagon who are almost always treated as the exemplary producers and shapers of the literary culture of the mid-Heian period; but men such as Tadanobu, who interacted with them in that same literary terrain and freely traversed what we often think of as its demarcated sectors, need to be returned to this picture. An examination of Tadanobu’s writings in a variety of forms, languages, and contexts — not only his kanshi (poems in Chinese) and kanbun (Chinese prose) prefaces (in Honchō monzui, a compendium of Chinese writings by Japanese authors assembled no later than 1066, and Honchō reisō, a kanshi anthology compiled ca. 1007 – 1008) but also some examples of his Japanese poems composed in a variety of hare (public) settings, such as for poetry contests (utaawase), for inscription on painted screens (byōbu uta), to commemorate imperial progresses (miyuki), and to contribute to Buddhist devotional offerings (shakkyōka) — serves a number of useful purposes. Foremost, we see here an array of sectors of the intercultural and multilingual literary field, and we can question whether any of these sectors should properly be characterized as central or peripheral. An assessment of the range in the oeuvre of this one figure can also help shift some of our perceptions of the norms of literary culture in the Heian world: we can see how frequently literary production took place in relatively impersonal, programmed settings, rather than in private, reflective circumstances. Finally, the coexistence of multiple literary languages in this field, seen and heard by men and women alike, suggests that a relatively open, permeable, and multifaceted collective literary discourse was characteristic of these times. This picture differs considerably from the one that emerges from our conventional regard for kana writings (for the most part, women’s writings) as the central paradigms of Heian literary culture. I will examine some of the reasons for the durability of this bias toward the female kana canon, but through this study, I hope to suggest some new ways that we might picture to ourselves a much more variegated horizon of literary textuality in Heian court culture circa 1000.
Terrains of Text in Mid-Heian Court Culture | 133
Reconfiguring the Terrain There is nothing all that new in arguing along the lines that I am suggesting; rather, it is a question of where and how we place our focus and emphasis. In her chapter “Chinese Intellectual Life” for the Heian Japan volume of The Cambridge History of Japan, Marian Ury wrote, “A view that is fortunately losing currency among Western students of Japan holds that to the Japanese the Chinese language remained permanently alien.”9 She went on to document in ample detail how Chinese learning was both assimilated and appropriated into the indigenous culture. As a separate cultural tradition or gathering of traditions, however, Chinese learning became itself a Japanese tradition. Reified, it was possessed, transmitted, hoarded, honored, displayed, boasted of, cultivated, and lamented when it failed to flourish. It came to have its own patriarchs, a history, and curiosities. The most appropriate metaphor might be that of a fund of intellectual capital, consisting of the segment of Heian Japan’s intellectual inheritance that depended directly on the use of the Chinese language and education in Chinese books. Originally brought in from abroad, it still received supplements from abroad, though in reduced number, but during the Heian era it can also be seen replenishing itself from its own resources.10 The managers of this “fund of intellectual capital” were of course male, and Ury discusses many of them and their works, but she does not concern herself with the issue of women’s access to or alienation from this resource. Yet the identification of specific textual territories as the domains of almost exclusively male or female readers or writers has been a long-standing tenet of writing about Heian literary culture — and it is one that has come under increasing scrutiny in recent years, as gender-based criticism has been brought to bear upon established assumptions, models, and modes of reading of the Heian textual corpus. A good example of scholarship in this contemporary vein can be found in the work of Tomiko Yoda.11 On the subject of the linkage of specific orthographies to the respective genders and to national origins, she observes, The assumption of the neat division between kana-feminine-Japanese versus mana-masculine-Chinese in Heian court society has stimulated debates that pit one against the other — for instance, asking which of the two ultimately represents the essence of Heian aristocratic culture. Such an inquiry, however, precludes the possibility that multiple cultural values and logic may have operated in Heian court society without necessarily constituting a sharp dichotomy. The multivalent approach to Heian culture is suggested here not so much to recuperate it as an idealized field of free-floating plurality but as a starting point for exploring its historical specificity in a more nuanced and methodologically self-conscious manner.12 The evidence available to us is indeed unlikely to yield a representation of the Heian literary terrain as a “field of free-floating plurality,” but neither does it sup-
134 | edward kamens
port an uninflected account of insurmountable categories or sectors that could never be broached, from either side. Yoda concurs with Thomas LaMarre, who, she says, “approaches Heian discursive space in general as being comprised of multiple styles and variations in which yamato and kara modes overlap, double, and hybridize with each other.”13 In that spirit, she argues that whatever the exact nature of mana taboo [by which term she refers to documented examples of women’s avoidance of use of the “Chinese” mana orthography], it did not operate as an absolute prohibition; neither could women’s relation to Chinese letters be characterized as that of clear segregation. What women did experience, it seems, was the pressure to feign a certain distance and unfamiliarity with that part of their own world.14 This feigning is of course the shiranai furi of Chino’s more informal discussion of the same matter. Yoda also asks, “When women were told to stay away from mana, were they prohibited from Chinese language or from Chinese script?”15 In other words, if such a taboo existed, did it operate in visual fields only, in practices of reading and writing, or did it extend into oral and aural practices? The evidence that I will discuss below, from some of Tadanobu’s interactions with literate women courtiers, suggests that rōei, the voiced performance of verse, was one especially significant oral and aural medium for shared encounters with and appreciation of Chinese texts — both those composed in China and by Japanese writers using Chinese — by both men and women. But in almost every known instance of such sharing, males were the ones doing the singing, and women were listening, recognizing, critiquing, and enjoying. This does not necessarily mean that men were at the center of this practice and women on its periphery; it may have been the opposite, but a more helpful characterization might see rōei performance as a space occupied by participants with multiple roles — males vocally activating texts, women (when present) joining with other men in the appreciation of the musical effects as well as the textual significations, admiring of the taste and skill of the performers, but not even feigning distance from the text or its performance. This is just one way in which the model of center and periphery seems helpful as a starting point for reconsideration of text and gender interrelationships in Heian culture, but having begun our inquiries in this way, we soon see the need to move beyond this spatial conceptualization toward more-subtle characterizations of the historical phenomena that are our objects of study. Such spatial conceptualizations of the relative significance, dominance, prestige, and influence of specific textual languages and their associated genres, and the gender of their practitioners, lie at the core of most accounts of Heian literary history and often guide readings of the texts themselves; but for this very reason these received conceptualizations deserve to be questioned anew. Very different and, I think, more historically accurate conceptualizations take shape when we read the texts in question with some skepticism about these presuppositions and biases, which of course have their own historicity. That historicity is explored in another revisionist study, Kurozumi Makoto’s
Terrains of Text in Mid-Heian Court Culture | 135
“Kangaku: Writing and Institutional Authority,” available in both Japanese and English.16 From the outset, Kurozumi announces, “I intend to show that while kanbun and kana are often thought of as antithetical writing systems, with their respective disciplines and canons in competition with each other, they were in fact closely interrelated, as were kangaku and kana-based disciplines.”17 He explicitly contests the placement of vernacular kana texts at the literary “center”: [H]iragana literature (waka collections from the Kokinshū on, women’s diaries and monogatari) began appearing in the mid-Heian period in the tenth and eleventh centuries. While this development deserves attention, its importance has been exaggerated, creating the mistaken impression that the Japanese language of the Heian period was centered on hiragana-based documents and that this constituted the main stream of linguistic culture.18 Kurozumi challenges the tendency to focus attention exclusively on vernacular works of the Heian age (which by and large we associate with female writers and readers), reading them in isolation from or as contestations of the Chinese/ male literary domains. In this configuration, which is largely a product of early- modern and modern ideologically driven canonization processes, women’s vernacular works and those written later in the genres that they putatively established are placed at the center, and Chinese writings are of course dismissed to the margins.19 For the Heian period, at least, Kurozumi argues instead for recognition of “a general trend toward linguistic and cultural syncretism, a Japanese/Chinese (wakan) fusion.”20 This syncretism is not a bland abstraction: it was realized in the oral and aural registers activated in the complex dynamics of Heian reading and writing practices and in a “linguistic consciousness which valued both Chinese and Japanese and promoted contact and fusion with Chinese from within Japanese.”21 I see this notion of “contact and fusion with Chinese from within Japanese” as a good model for rethinking such things as the dynamics of Heian women’s contact with Chinese texts, which at least some of them encountered at times as written texts but perhaps even more often as articulated, verbalized texts, made audible and active in a variety of settings in the complex amalgam of Heian court culture.22 Furthermore, Kurozumi asserts that the tendency to treat the tenth-century vernacular literary monuments (the Kokin wakashū and The Tale of Genji) as the exemplary literary productions of the time and to take their depictions of court society as the only valid images of the past “emerged later, as a contrast to the fierce image of the warriors who had seized power. This perception began to circulate during the Muromachi period after the study of Heian waka and monogatari had become the object of highly retrospective classical study.”23 And in an essay in the same volume, Tomi Suzuki goes on to chronicle the latter stages of the canonization process that ultimately placed kana writings at the center: Heian kana literature, long associated with femininity, was designated [from the 1890s onward] the basis of “national literature” as a result of the phonocentric notion of “national language” (kokugo) that emerged in close relationship
136 | edward kamens
with the genbun-itchi (union of spoken and written languages) movement and that increasingly stressed direct and unmediated expression. In this newly constructed body of “national literature,” from which all the texts written in kanbun, or classical Chinese, were eliminated, Heian waka, especially the Kokinshū — a canonical text for a thousand years — were devalued because they were not considered to be direct, unmediated expression.24 One element of Suzuki’s argument that is particularly important here is the notion that waka could be and has been critiqued in terms of the extent of its perceived “direct, unmediated” qualities or lack thereof. For better and for worse, this tenet has indeed often guided readings of waka, which are all too often assumed ipso facto to be lyrical expressions that pour forth in unmediated surges from the poet’s heart (kokoro). Scholarly considerations of aspects of poetry that do not meet this criterion or fit this model are all too often pushed to or beyond the margins of consideration. One of my purposes here, in examining the oeuvre of Tadanobu, is to emphasize the extent to which this privileging of an ahistorical ideology of poetic composition has skewed many representations of waka compositional practice and caused it to be treated in isolation from other text-making and text-activating practices such as quotation, citation, and chanting. When judged from his extant works and descriptions of his activities, Tadanobu turns out to be a good example of a literary practitioner whose text-making and textciting activities (in both Chinese and Japanese) more often than not transpired in specific social settings governed by programmatic protocols — which means that his texts are not so much expressions as performances of expression.25
Kanshi and Waka no jo Tadanobu was the second son of Fujiwara no Tamemitsu (942 – 992), who was chancellor (daijō daijin) from 970 to his death in 992. His paternal uncles Koretada, Kanemichi (925 – 977), and Kaneie (929 – 990) all served as sesshō (regent for child emperor) or kanpaku (regent for adult emperor) or both. Kishi (969 – 985), his younger sister by the same mother, became Emperor Kazan’s junior consort (nyōgo) but died shortly thereafter from complications of pregnancy. These and other bloodlines and marriage ties placed him at what we might think of as the center of court society, although neither he nor his father achieved the highest levels of office or wielded real political power. Tadanobu rose to the office of dainagon (grand councillor) in 1010, after serving for a decade as master of the household (chūgū daifu) to Ichijō’s consort Fujiwara no Shōshi (later known as Jōtōmon’in, 988 – 1074). His career and the scope of his social and literary activities followed much the same trajectory as that of three other leading noblemen of his day — Fujiwara no Kintō, Minamoto no Toshikata (960 – 1027), and Fujiwara no Yukinari (972 – 1027). At his death at age sixty-nine, Tadanobu held the senior second rank and was minbukyō (minister of popular affairs) in addition to dainagon. While both Tadanobu and his elder brother Sanenobu (964 – 1001) may have
Terrains of Text in Mid-Heian Court Culture | 137
received personal literary instruction from Minamoto no Tamenori, who is best known as the author of Sanbōe and of Kuchizusami, the latter a learning guide for Sanenobu,26 no details of Tadanobu’s formal education are known.27 We can only assume that he was well prepared to take part in the literary activities that were a feature of many social gatherings in the residences of the likes of Fujiwara no Michinaga (966 – 1027) and his son and successor Yorimichi (990 – 1074) or in the royal palace or other royal residences. He is known to have participated in at least twenty-six gatherings at which the composition of Chinese verse was a major feature, and almost all of these took place in the palace, at one or another of Michinaga’s residences, or at Yorimichi’s Uji residence.28 Culled from these occasions, a total of sixteen of his kanshi and one preface (jo) in Chinese, introducing a collection of commemorative waka (see below), survive in various collections.29 The following poem, composed at a gathering at Michinaga’s residence on the twenty-fourth day of the third month of 1006, is one of three produced on this occasion included in Honchō reisō (the others are by Kintō and Ōe no Michinao [n.d., active ca. 1012 – 1017]).30 The five-character topic was “Flowers and birds are spring’s treasures” (kachō wa shun no shicho tari): kachō wa izure no hi yori ka kokin ni kikoyu, When were the praises of flowers and birds first sung? Who first said, haru no shicho tarite sengin ni masaru to. “They are spring’s treasures, worth more than a thousand pieces of gold”? tsumu mo masa ni yoku fūzen no iro wo chirasubeku, Yet spring does not hoard its flowers — it scatters their colors before the winds; musaboru mo rotei no ne o aitsutaen to hossu. It does not simply revel in its birds — it sends their songs echoing under covers of dew. taikei no yoryō no gotoshi sangatsu no go, So third-month songs at dusk are a late spring bonus; kōjun no seikei no gotoshi ichien no shin. a colorful garden at month’s end is extra salary. koko yori omoietari hinso ni arazarukoto wo, And so, I tell myself, I am in no sense poor, dogi ni au goto ni suigin wo hoshiimama ni sura wa31 and whenever I find myself in blessed surroundings such as these, I give myself up to drink and song. We can treat this poem as a typical example of those that capable kanshi poets could produce to meet the expectations of social gatherings presided over by the politically powerful and at which both semiprofessional poet-scholars (such as Michinao and Masahira) and amateur but nevertheless accomplished practition ers (such as Tadanobu and Kintō) were present. Tadanobu expatiates upon the topic (dai) in a manner that serves not so much to express a personal perspective as to shape the poem-making act itself as a compliment to the host. He makes a gracious gesture of gratitude for being present and returns the favor with the
138 | edward kamens
production of a text that modestly celebrates its own capacity to reactivate the cadences of its own highly conventional form. We see this poem as a fragment of a specific moment of social interaction, but it also quietly insinuates itself into a much larger corpus of poems that looked and sounded much like it, and thereby fulfills its primary task. A task of quite another sort is addressed in another example of Tadanobu’s surviving kanshi. The Shingon abbot Ninkai (951 – 1046) was called upon in the summer of 1033, as on several other occasions, to conduct rites for rain during a particularly fierce drought. When he was apparently successful, Tadanobu was one of several courtiers commissioned to write poems in praise of his powers. den’en kansō shite imada sōsei sezu, The fields were sere, no living thing could grow; takaku shinkin wo ugokashite koku wo inoru sei ari. This caused His Majesty to be gravely concerned and inspired his supplications for good crops. Anokudatsu no nami haruka ni sawagu nochi, After the waves on Lake Anavatapata had beat loudly on those distant shores, Jōrin’en no tsuki hajimete kumoru hodo. And once the moon had begun to pass through clouds over Shanglinyuan,32 yūun tareshikite ten munashiku...nari Heavy clouds lowered, draping the heavens... kantaku amaneku hodokoshite chi...tairaka nari. And sweet rains nursed the earth in all directions. hitori hazuraku wa chikagoro uruoi wo fukumu uchi Yet I alone am embarrassed, for during this time of gathering moisture shika karetsukite in no osoku naru koto wo.33 My poems have dried up, my rhymes have been slow to come. Flattering though this must have been to Ninkai — whose deeds had been deemed efficacious, while Tadanobu chides himself for having had nothing to offer toward the cause — the abbot was nevertheless peeved that more tangible forms of reward for this and the previous year’s rainmaking were slow to come from the authorities.34 Once again, it was Tadanobu, presumably in his official capacity as dainagon, who composed the conciliatory response — in the form of yet another kanshi and a waka as well.35 Sketchy as the details of this episode may be, it serves to remind us that exchanges of poems sometimes had such practical uses as these. If this is indeed an example of the administrative use of poetry, in this instance in communications between the court and its commissioned subcontractor (a Shingon prelate in his eighties with an impressive record of effective service as a thaumaturge)36 it invites us to look for other examples that may in turn help us understand more about the status of poetic language and poetic text with respect to other formal
Terrains of Text in Mid-Heian Court Culture | 139
languages used in official and semiofficial discourse. In other words, poetry will appear to be a discourse that was at least at times available for or subject to integration with other languages, not relegated to a sphere defined as or restricted as literary. It is in any case apparent that the composition of poetry, in Chinese or Japanese and on some occasions both, was often an activity integrated with a wide variety of court observances and rituals. Again, we might ask if the composition, presentation, sharing, and subsequent preservation and circulation of poems had more than a decorative function. That is, to what extent, and by what means, did the inscription or utterance of poems in such settings become a virtually necessary part of, rather than an ancillary coda to, these ritual activities? Such questions are also pertinent in the following discussion of Tadanobu’s waka, given what we know about the circumstances of composition of most of his extant Japanese poems. Here, we might also make note of a Chinese preface he composed for preservation together with the waka written by himself and other courtiers at a banquet held on the twenty-first day of the eleventh month of 1030, on the day after the onhakamagi (first dressing in trousers) for Princess Shōshi, the eldest daughter of Emperor Go-Ichijō and Michinaga’s daughter Ishi. The ceremony took place when Shōshi was five years old, having been postponed during the period of mourning following her grandfather Michinaga’s death in the twelfth month of 1027.37 The rhetoric of such prefaces is invariably formal and ingratiating, but this does not necessarily mean that there is no significance in the fact that Tadanobu associates the royal princess who was the focus of these celebrations with the seventh-century empresses Kōgyoku and Jitō. He points out that these royal women were surrounded by worthy advisers who came to their aid when destiny decreed that they serve as female sovereigns (ōjō or kōjō).38 The implications as far as the child Shōshi is concerned seem far-fetched given the political conventions of the period, but such aggrandizing rhetoric may at least serve to mark the importance of what would otherwise be a more routine life event for a royal daughter with maternal bloodlines connecting her to the most powerful branch of the most powerful family in the land. It may also be significant that this occasion — like so many others — was deliberately marked by the production of texts in both Chinese (Tadanobu’s preface) and Japanese (his waka and those by several others.)
Waka Tadanobu’s waka composed at this post-hakamagi banquet is one of four reproduced in Eiga monogatari’s account of the event, and like the others — by Shōshi’s adult uncles Yoshinobu (995 – 1065), Yorimichi, and Norimichi (997 – 1076) — it activates time-hallowed figures of longevity on the royal child’s behalf: watatsuumi no kame no senaka ni iru May the years of your life
140 | edward kamens
chiri no yama to narubeki kimi ga miyo ka na be as vast in number as that mountain of dust that builds up on the backs of turtles in the sea!39 This is waka rhetoric deployed to make the requisite ingratiating courtly gesture: Tadanobu senses the social and political expectation to respond in this manner and avails himself of the conventions of poetic usage that have been used many times before in similar ways (and that would be used in similar ways many times again) to reaffirm his relationships to the parties concerned and demonstrate anew his continuing capacity to perform in precisely this way. According to the Eiga monogatari account, the poem making at the banquet got under way when Go-Ichijō himself said, “Waka nado arubeshi” (There ought to be some [Japanese] poems [at a gathering like this]).”40 This may mean not simply that the making (and uttering) of poems was a fitting ancillary activity or an elegant sideshow at the banquet, but that the occasion itself needed these quasi-liturgical recitations of new settings of old poetic figures to make it complete and to affirm the participants’ sense of being present at an historically significant repetition of time-honored practices. The recitation of poems with such familiar tropes would serve to locate the occasion in a temporal continuum, and their subsequent recording would help preserve collective memory of this particular occasion’s place in time. One might make similar conjectures about the significance of many other instances of public (hare) poetic productions. Within Tadanobu’s oeuvre, at least, these are numerous, but in this respect he is typical of court poem makers of his time, who frequently found themselves called upon or took the opportunity to contribute to and commemorate an array of events with verses that were integrated with the program for those events; their poems would subsequently serve as part of the written record of the events that occasioned their production. (As readers of Makura no sōshi know, this was a kind of poem making for which Sei Shōnagon found herself temperamentally ill-suited.41) The program for some such events might call for the production of poems to be integrated with the production of paintings: byōbu (screen) and uta (poem) presentations accompanied many important life-cycle events for royal women and noblewomen in particular. This practice seems to have been entering a new stage of development during Tadanobu’s time. When Michinaga presented his daughter Shōshi at the palace as Emperor Ichijō’s consort in the tenth month of 999, he made a stir by commissioning some of the highest-ranking men of the court, including Retired Emperor Kazan, to compose the poems for her dowry screens, rather than consigning the task to more ordinary court poets.42 Tadanobu’s contribution was a verse coordinated with a section of the screen that depicted “a house surrounded by pine trees where people were playing music on flutes” (Jōtōmon’in judai no toki, onbyōbu ni, matsu aru ie ni fuefukiasobi shitaru hito aru tokoro wo yomihaberikeru): fuetake no yofukaki koe zo kikoyu naru The sound of the bamboo flute is heard deep in the night;
Terrains of Text in Mid-Heian Court Culture | 141
mine no matsukaze fuki ya souran and is the wind from the pines on the peak joining in?43 The poem fulfills programmatic expectations: it activates the trope of conjoined man-made and natural sounds to represent a semidivine music that cannot be heard but that, through the medium of word and picture, becomes part of the orchestration of festive airs that celebrate Shōshi’s marriage. Poetry competitions were important opportunities for courtiers such as Tadanobu to practice and prepare for such performances. At such gatherings, participants honed their skills in addressing set topics by deploying tried-andtrue tropes in new configurations, at the same time also satisfying the host’s desire to demonstrate his or her capacity to orchestrate and facilitate the collective display of these skills and to provide the setting in which these elegant performances could be staged. Tadanobu made his first recorded appearance at an utaawase at the age of twenty, in a contest held in the royal palace in the sixth month of 986. His contribution for the team on the left tied with that of the representative of the right, Fujiwara no Koreshige, in a round with the assigned spring topic Warbler (uguisu): [Shōshō Tadanobu (the left):] kōritoku kaze no oto ni ya Hearing the sound of the wind that melts the river’s ice sugomoreru tani no uguisu haru wo shiruramu the warbler that huddled in its nest in the valley must now have learned that spring has come. [Koreshige (the right):] uguisu no naku ne nodoka ni kikoyu nari The warbler’s song falls gently upon the ear: hana no negura mo ugokazarubeshi its nest of blossoms must not have been disturbed.44 Tadanobu’s strategy was to activate a time-tested anthropomorphic trope — a bird awakens and sings when it becomes aware of seasonal change — while Koreshige deployed the conceit of the nest embedded in or made of flowers (hana no negura). In the opinion of the unnamed judge, these were respectable performances, neither outdoing the other. Expectations were fulfilled by both, if not with particular distinction by either. Tadanobu’s only other known utaawase appearance occurred many years later, in the tenth month of 1032, when Shōshi, as Jōtōmon’in, the dowager empress, held a chrysanthemum contest (kikuawase) at Kayanoin, her brother Yorimichi’s residence. This event took as its models several kikuawase of the early and midtenth century, but it was the first revival of those model practices after several decades of hiatus. The participating team poets (kataudo) were for the most part ladies-in-waiting in Jōtōmon’in’s intimate service; noblemen, including Tada nobu, are listed in the contest records as well but their role appears to have been
142 | edward kamens
limited to that of advisers (omoibito or nennin) to the respective left and right teams.45 With chrysanthemums brought to the Kayanoin from Hōjōji for display before the participants, to serve as the visible and tangible referent for their verses, this was an occasion for festive renewal of past practices that celebrated both the potent floral symbol of longevity and the Fujiwara family itself. Tadanobu was well qualified for his role as advisory omoibito by this time: he could call upon many decades of experience in performative poem production. For example, he had been among those invited to compose poems on the topic Colored leaves afloat upon the water (kōyō mizu ni ukabu) when Koichijōin (Atsuakira, a former crown prince) made an autumn excursion to Ōigawa. Tadanobu contributed the following to the verse record of the event: aki fukaku nariyuki toki wa When autumn reaches its peak Ōigawa nami no hana sae momijishinikeri even the flowers in the waves of Ōi River change color as do the maple trees!46 Here again, he had exercised a trusted trope, the mitate of foam as “wave flowers,” to fulfill his task. A performance of this kind served through recapitulation of familiar figures to extend and perpetuate the traditions of Ōigawa poesy and their special relationship to the history of royal visitations to that site.47 Tadanobu had also been a key performer at other gatherings where poems were produced to celebrate the longevity and accomplishments of particular Fujiwara family members, such as the Senzai awase (garden match) at the Tsuchimikado mansion in the eighth month of 1023, which honored Dowager Empress Shōshi with poems on the topics The autumn moonlight is bright (aki no tsuki, hikari sayakanari) and The waters of the lake will long remain clear (ike no mizu nagaku sumu).48 Later that year, he was at the banquet celebrating the sixtieth birthday of Michinaga’s wife, Rinshi, and joined with Michinaga, Kintō, Sanesuke, Yorimichi, and several others to compose verses for her.49 Once again, in the ninth month of 1024, he was a member of the party at the Kayanoin when the family received Go-Ichijō for a viewing of the annual royal horse races, and he joined with the emperor and others to compose poems on the seasonal topic The radiant bloom of the chrysanthemums on the bank will long endure (kishi no kiku hisashiku niou).50 Given this experience as well as his intimacy with Michinaga and his immediate and extended family, it is not surprising that Tadanobu should have been among those whom Michinaga called upon when he organized collective programs for the composition of devotional Buddhist verses. The date and the votive beneficiary of Michinaga’s nijūhappon no uta (verses composed on topic texts extracted from and representative of each of the twenty-eight chapters of the Lotus Sūtra) are not known, but some scholars believe the project to have been carried out in memory of one of his daughters after her death. Tadanobu’s poem addresses as its topic “the essential meaning of the ‘Fortitude’ chapter” (Kanjibon no kokoro wo) of the sūtra:
Terrains of Text in Mid-Heian Court Culture | 143
kazu naranu inochi wa nani ka oshikaran In this insignificant life, for what should I feel regret? nori toku hodo wo shinobu bakari zo I tolerate it only to be here to share these Lotus teachings!51 Here, the compositional task is to rework the Chinese text of the scripture into the cadences of the Japanese poem; the voice that implicitly utters Tadanobu’s poem occupies the same position and attitude as do the bodhisattvas in the sūtra passage who similarly announce their determination to withstand all hardships in order to partake of the Buddha’s instruction. This is another kind of performative exercise, motivated at least in part by devotional objectives, but rendered with much the same set of skills that Tadanobu and poets like him brought to other types of daiei production; and, once again, Tadanobu is the mediator in a transformation from a Chinese “original” to a Japanese re-presentation. But some poems in Tadanobu’s oeuvre appear to be less impersonally performative than these, and we need to examine them alongside the rest of his work to appreciate its range across the compositional terrain. In the previously discussed instances, we have seen Tadanobu supplying verses to fulfill expectations in various social situations, but in the following examples the form of the poem itself is supplying Tadanobu with his communicative medium, as he produces poems that more directly respond to specific events and conditions. In these poems, therefore, we can legitimately identify the expression of personal sentiments, and here Tadanobu’s voice is more his own rather than that of a collective consciousness recapitulating its time-tested tropes. At some time before 1031, and not long after the death of Michinaga, when Tadanobu learned that Yorimichi was taking a group of courtiers on a flowerviewing expedition but not including him in the party (Uji no saki no daijō daijin, hanami ni namu to kikite tsukawashikeru), he used the vehicle of verse to convey his disappointment: inishie no hanami shi hito wa tazuneshi wo You, with whom I have gone flower viewing in the past, once again seek them out; oi wa haru ni mo shirarezarikeru but spring, it seems, is not alone in forgetting those who now are old. Yorimichi’s reply refers directly to the absence of his late father: tazunen to omou kokoro mo The heart that seeks out the flowers inishie no haru ni wa aranu kokochi koso sure cannot feel as once it did in those bygone springs!52 Tadanobu had implied that Michinaga would not have excluded him, but Yorimichi neutralizes this complaint simply by citing his own lingering grief. Such give-and-take may seem largely rhetorical, but the exercise of such rhetoric may
144 | edward kamens
in fact have been significant in delineating the state of relations between these two men. Senzai wakashū, the imperial waka anthology (chokusenshū) compiled late in the twelfth century, also presents (as poem no. 912) what would appear to be a poem Tadanobu wrote in the course of a love affair with an unnamed correspondent: “When a woman told him she wanted to withdraw to a place deep in the mountains, he sent...” (onna no, fukaki yama ni mo iramahoshiki yoshi ihite haberikereba tsukawashikeru...): yama yori mo fukaki tokoro wo tazunemiba Were you to search for depth greater than that of the mountain to which you say you wish to go, waga kokoro ni zo hito wa irubeki you would do well to look into my heart.53 This seems to be the only koi no uta (love poem) in Tadanobu’s oeuvre. Its prose preface may tell of genuine circumstances, or it may not; in any case, the poem’s existence suggests that Tadanobu’s range apparently extended into this topical and thematic sector as well. Chapter 30 of Eiga monogatari, “Tsuru no hayashi,” describes the double shock felt through the court when Michinaga, who had long been ill, and Yukinari, who had not, died on the same day in the twelfth month of 1027. The account includes an exchange between two men, Kintō (by this time retired to religious reclusion at Nagatani) and Tadanobu, who had been particularly close to the recently deceased courtiers. I see this exchange (which appears in no other text) as a document of surviving friendship; here, the exchange of poems does seem to serve as a means of offering solace, one to the other. [Kintō:] mishi nito no nakunariyuku wo kiku mama ni I hear of their deaths one after another — those men I once knew — itodo fukayama zo sabishikarikeru and life deep in these mountains seems lonelier still. [Tadanobu:] kienokoru kashira no yuki wo haraitsutsu Brushing back hair white as lingering snow, sabishiki yama wo omoiyaru ka na I let my thoughts wander to the mountains where you dwell in loneliness.54 If Kintō and Tadanobu really did share these poems with one another, as Eiga monogatari reports, we may indeed treat them as activations of tested tropes as powerful salves: two aging courtier-poets, shocked by loss and anxious about change, turn to these familiar figures and cadences as outlets and containers for their shared grief, and by doing so they find and offer one another comfort and reassurance.
Terrains of Text in Mid-Heian Court Culture | 145
Rōei Tadanobu’s literary productions span several sectors of the textual terrain and are representative of Heian compositional practices in the array of languages, audiences, occasions, and rhetorics that shaped them. Rōei, the performative reproduction of texts, was yet another medium in which, according to his contemporaries, he particularly excelled. When he so adroitly and movingly recited Fumitoki’s memorial verse to honor the memory of Michitaka, as reported in Makura no sōshi, a mixed audience of women and men shared the space of communication and appreciation.55 We can see this space as an important intersection, where a performative art practiced almost exclusively by males offered itself on this occasion (and at least some others) to women who were ready to receive it with unimpaired comprehension of its uttered Chinese text and able to relate it to its written Chinese (or Chinese-within-Japanese) point of origin.56 The thirteenth-century setsuwa collection Kokon chomonjū relates how Tada nobu made a similarly strong impression at an apparently all-male gathering: Once, the Higashi Sanjō’in Regent and Former Chancellor Kaneie attended the moonlight nenbutsu service on the night of the thirteenth day of the ninth month at the Tōhokuin. When it had grown quite late, and all was very quiet, he summoned Tadanobu, the minister of popular affairs, and said, “How can we let this night go by without something special? How about some rōei? ” Tadanobu respectfully acknowledged his request. For some time he looked as if he were struggling with indecision, and all those present were waiting and listening closely to see just what verse he might choose to chant. Then he sang: “Our meditations on the Honored One of the Land of Utmost Bliss last through the night” (gokuraku no son wo nenzuru koto ichiya). It was quite wonderful, beyond compare (tagui naku medetekarikeri). [Ki no] Tadana, who had written this verse, was in fact of the company on this occasion. What pleasure he must have had in hearing his own verse chanted as a rōei by such a one as Tadanobu!57 Here again, Tadanobu’s success in fulfilling the host’s expectations was achieved by transferring a remembered text from its original written context to a new setting. Tadana’s lines came from a preface to poems he and other confrères composed at a gathering of the Kangakue — the mixed lay-aristocratic and clerical Society for the Advancement of Learning, which focused particularly on Pure Land texts and devotions — probably in the 970s or 980s. In the process of its reactivation by Tadanobu, the chanted quotation is transformed into a verbalized maxim that effectively connects one earlier devotional occasion to another through this explicit textual link. At the same time, Tadanobu creates and reinforces a community of shared appreciation and understanding: honoring the older Tadana (957 – 999), he also draws his auditors into the collective moment of recognition and admiration. That which was so especially wonderful (medetashi) in this performance was much the same as that which was medetashi to Teishi and Shōnagon in the episode recounted in Makura no sōshi, section 29.
146 | edward kamens
In section 155 (“Kotono no onbuku no koro...” [while we were in mourning for His Excellency]) of Makura no sōshi, Sei Shōnagon again has reason to be impressed with Tadanobu’s tasteful citations. She immediately recognizes the lines of a poem by Bai Juyi when Tadanobu says he plans to recite them at the following night’s Tanabata Festival observances, but she notes that the identification of the source and the appropriateness of the selection are likely to have been lost on the other women listening, as she is, from behind the screen. She also doubts that Minamoto no Nobukata, a gentleman also present, has the acumen necessary to appreciate the moment as she does.58 This exchange is explained in the following flashback to the previous summer, a night at the beginning of the fourth month, when, Shōnagon says, she took special delight in listening to Tadanobu and Nobukata singing an impromptu rōei duet on lines from a Tanabata poem in Chinese by Sugawara no Michizane. She further praises Tadanobu for cleverly explaining why he chose to sing verses on an autumn topos (the festival of the ninth day of the ninth month) on a summer night.59 Not long afterward, she somewhat jokingly advises the emperor himself to delay a promotion for Tadanobu because, she says, his new duties will reduce the opportunities for her and others to hear him sing. She certainly does not hesitate in presenting herself as an astute aficionado of rōei performance: “Tadanobu is wonderful at reciting Chinese poetry” (shi wo ito okashiu zujihaberu mono wo), I said. “Now that he has been promoted, who is going to be left to give us lines like ‘Hsiao of K’uai-chi, having visited the ancient tomb...’? Your Majesty had better make sure that he continues coming here, even if it means that he has to wait a little longer for his new post. It would be too sad to lose him.”60 As modern commentators have noted, Shōnagon is making her own adroit allusion here to get her message across. The rōei lines she refers to are from a rhyming preface by Ōe no Asatsuna (886 – 957) on the sorrows of parting from old friends — a fitting quotation, since Shōnagon wants to prevent any greater distancing from Tadanobu. But she is also demonstrating the extent to which she is an active and equal member of the community of connoisseurs — along with Tadanobu and the emperor himself — who collectively circulate and savor such texts. In the much-discussed section 78, “The Captain First Secretary, Tadanobu, having heard certain false rumours...” (Tō no Chūjō no suzuronaru soragoto wo kikite...), Shōnagon at first reveals some anxiety over the receipt of a letter from this sometime paramour, but she finds that it is troubling not for what it says about their relationship at this stage but for the kind of response that it demands: It was elegantly written on heavy blue paper, and there was nothing about it to worry me. I opened it and read: “With you it is flower time as you sit in the Council Hall
Terrains of Text in Mid-Heian Court Culture | 147
’neath a curtain of brocade.” (ranshō no hana no toki/kinchō no shita) And below this he had added, “How does the stanza end?” (sue wa ika ni, sue wa ika ni).61 Shōnagon has no trouble recognizing the quotation from Bai Juyi’s collected works (Hakushi monjū),62 but is unsure of the form that her reply should take; she wishes she could confer with the empress, to draw upon her knowledge of such texts and her intuition regarding the proper way to join in such literary repartee: I was at a complete loss. If Her Majesty had been there, I should have asked her to look at the letter and give her opinion; but unfortunately she was asleep. I had to prove that I knew the next line of the poem, but were I to write it in my somewhat faltering Chinese characters it would make a bad impression. I had no time to ponder since the messenger was pressing for a reply. Taking a piece of burnt-out charcoal from the brazier, I simply added the following words at the end of Tadanobu’s letter: Who would come to visit This grass-thatched hut of mine? (kusa no iori wo/tare ka tazunen) Then I told the messenger to take it back to Tadanobu. I waited for a reply, but none came.63 Shōnagon’s solution, arrived at with minimal delay, was to render part of the next couplet of the Chinese original as if it were a Japanese poem while retaining the original’s representation of the distance and difference between the environs of the court and the dwelling of someone who chooses to live apart from it. It is impossible to say whether Shōnagon knew the original verse as a written or as a voiced text, or as both. What her correspondence with Tadanobu produces, however, is a unique admixture. Her concern for the way in which her answer will look places emphasis on its orthographic form (she seems to share this anxiety with Murasaki Shikibu, who also tells us that she “avoided” writing kanji 64), but what she creates is a text that is no longer simply Japanese or Chinese, nor specifically male or female, but one that crosses or diverts these boundary lines to engage Tadanobu on a newly configured textual plane. Sei Shōnagon was not the only woman in Ichijō’s court who could meet Tadanobu halfway to create new discourse in such textual terrain. In a section of Murasaki Shikibu nikki that begins “Her Majesty went over to the Dedication Hall just before dawn on the eleventh” (jūichinichi no akatsuki, midō e watarasetamau) and proceeds to describe religious rites as well as entertainments provided for Imperial Dowager Shōshi and her mixed male and female retinue on a visit to the Tsuchimikado mansion, the diarist evocatively describes the events of one particular evening of festivities.65 A section of the Hinohara version of the Murasaki Shikibu nikki ekotoba (an illustrated handscroll version of selected passages and scenes in the diary, produced in the early thirteenth century) has been identified
148 | edward kamens
as an illustration of this scene (figure 6.1).66 If the corresponding text passage was ever part of this version, it is now lost. But at the center of the painting sits the imposing figure of Tadanobu, just as described in the memoir: When it was all over, the senior courtiers took the boat and rowed out onto the lake one after the other. On the eastern veranda of the hall in front of the open side door sat Master of Her Majesty’s Household Tadanobu. He was leaning against the railings of the steps that ran down to the water’s edge. While His Excellency happened to be inside with Her Majesty, Tadanobu took the opportunity to exchange a few words with Lady Saishō; what with her inside the screens trying not to appear too intimate in front of Her Majesty, and him outside, it was quite a performance.67 The relative positions of the participants, as described here, are spatially significant. Lady Saishō and other women, including the observing, reporting Shikibu, are in the interior (one woman is barely visible in the painting); outside, Tadanobu reclines on the steps. Between them, a space for looking and listening together opens up. A hazy moon emerged. The sound of His Excellency’s [Michinaga’s] sons all in the one boat singing songs in the modern style was refreshing and quite delightful, but the sight of Minister of the Treasury Masamitsu [Fujiwara no Masamitsu, a gentleman in his fifties], who had got in with them in all seriousness but who was now sitting there meekly with his back to us, not unnaturally loath to take part, was rather amusing. The women behind the screens laughed among themselves. “And in the boat he seems to feel his age” (fune no uchi ni ya oi wo ba kakotsuramu), I said. The master of the household must have heard me. “Hsu Fu and Wen Ch’eng were empty braggarts” (Jōfuku Bunsei kyōtan ohoshi), he murmured. I was most impressed.68 In this case, it is Shikibu who seizes the moment to open a dialogue with a rendering of a verse from Bai Juyi’s collected works (topically appropriate for its reference to sailing across open waters) into the cadences of what would seem to be part of a Japanese poem. Tadanobu responds and caps her verse with a chanted rendering of the lines that follow in the Chinese original. She finds both his voice and his whole physical manner (koe mo, sama mo) utterly elegant and fitting for the moment (koyonau imamekashi), but of course her role in this creative or re-creative performance is equally impressive. Together, Shikibu and Tadanobu invent another anomalous, hybrid text, which is nevertheless analogous in its mixing of languages to the conjoinings of passages from Chinese verses (by both Chinese and Japanese poets) with Japanese poems in Wakan rōeishū — which may ultimately be our best model for any envisioning of a literary praxis of mixed scripts and mixed activated texts, accessible to both men and women and representative of the variegated corpus of texts they could at least potentially and jointly see, hear, know, and use. Likewise, in this scene at the Tsuchimikado mansion, neither the male nor the female participant in joint text making skips a beat: drawn from a corpus they both share, the words come readily to their lips. Shiki-
Terrains of Text in Mid-Heian Court Culture | 149
Figure 6.1. Detail from the Hinohara version of Murasaki Shikibu nikki ekotoba. Ta danobu (center) observes boaters at the Tsuchimikado Mansion, while Sei Shōnagon and others comment from within. Private collection; reproduction courtesy of Chūō Kōron Shinsha.
bu’s voice, and then Tadanobu’s, write an audible joint script on a space they also share. With light hearts and knowing grins, they claim their respective places in this textual terrain, marking new tracings on its capacious surface — and, in so doing, they articulate and inscribe what may serve us well as a paradigmatic cross-section of their literary terrain.
Notes 1. Murasaki Shikibu nikki, 92; Richard Bowring, trans., The Diary of Lady Murasaki, 55 (slightly revised). 2. Murasaki Shikibu nikki, 97; Bowring, Diary, 57 – 58. 3. Murasaki Shikibu nikki, 57; Bowring, Diary, 34. 4. Hosshin wakashū no kenkyū, ed. Ishihara Kiyoshi, 37, 71 – 73; Edward Kamens, The Buddhist Poetry of the Great Kamo Priestess, 66. 5. In Tamakatsuma, Motoori Norinaga took Senshi to task for her open avowals of Buddhist faith while serving as Kamo priestess, but he did not criticize her for mixing Chinese and Japanese texts in Hosshin wakashū. See Kamens, Buddhist Poetry, 20 – 21. 6. See, for example, Chino Kaori, “Gender in Japanese Art.” 7. Ibid., 22 – 25. 8. Sect. (dan) 129 in Makura no sōshi, 242 – 244; sect. 88 in Ivan Morris, trans., The Pillow Book of Sei Shōnagon, 152 – 153. 9. Marian Ury, “Chinese Learning and Intellectual Life,” 341. 10. Ibid., 346. 11. Tomiko Yoda, “Literary History against the National Frame, or Gender and the Emergence of Heian Kana Writing.” See also idem, Gender and National Literature, 81 – 110. For another thought-provoking rereading of much of the material discussed in this same connection, see Joshua S. Mostow, “Mother Tongue and Father Script.” 12. Yoda, “Literary History,” 486.
150 | edward kamens 13. Ibid., 480 – 481. Yoda refers to two of LaMarre’s published works: his article “Writing Doubled Over, Broken,” and his book Uncovering Heian Japan. 14. Yoda, “Literary History,” 489. 15. Ibid., 479. 16. Kurozumi Makoto, “Kangaku: sono shoki, seisei, ken’i” and “Kangaku: Writing and Institutional Authority.” 17. Kurozumi, “Kangaku: Writing and Institutional Authority,” 204. 18. Ibid., 205. 19. This aspect of the canonization process is also one of the central concerns of Yoda in Gender and National Literature; see esp. 25 – 80. 20. Ibid., 211. 21. Ibid., 206. 22. There are also figures in earlier and later Heian court culture whose literary activities traverse multiple textual languages and forms in exemplary ways. The exemplary figure of this type in the latter part of the ninth century is Sugawara no Michizane; for a thorough study, see Robert Borgen, Sugawara no Michizane and the Early Heian Court (e.g., 268). For a study of Minamoto no Tsunenobu and Minamoto no Toshiyori, two mid-eleventh-century figures who likewise operated in multiple literary registers, see Ivo Smits, The Pursuit of Loneliness. 23. Yoda, “Literary History,” 210. 24. Tomi Suzuki, “Gender and Genre: Modern Literary Histories and Women’s Diary Literature,” in Inventing the Classics, ed. Haruo Shirane and Tomi Suzuki, 72. 25. For further development of this characterization of waka composition as performance rather than as expression, see Edward Kamens, “Waking the Dead.” 26. Ury, “Chinese Learning and Intellectual Life,” 348 – 355, provides a lively summary of the contents of this work. 27. Fukui Michiko, “Fujiwara no Tadanobu kō,” 17. 28. Ibid., 19. 29. Ibid., 21. Fukui’s article includes a useful table indicating the known topics, locations of compositional events, dates, and sources in which the poems and preface are now to be found. 30. Kawaguchi Hisao and Honchō reisō o yomu kai, eds., Honchō reisō kanchū, 19 – 25. Ōe no Masahira’s poem on the same topic, presumably from the same occasion, is included in his collection Gō rihō shū; a fragment of Fujiwara no Yukinari’s poem on the same topic appears in Kōzei shikō. The gathering in question is also mentioned in Yukinari’s journal Gonki, but there is no entry for this date in Midō kanpaku ki. 31. For annotated texts and detailed analyses, see Kawaguchi, Honchō reisō kanchū, 19 – 21; and Honchō reisō zen chūshaku, ed. Imahama Michitaka, 1:132 – 165. 32. Tadanobu implies that rain-bearing clouds have moved eastward toward Japan from locations on the Asian continent — a mythical lake in the Himalayas (Anokudatsu) and a historical garden of the Han capital at Chang-an (Jōrin’en). 33. Fukui, “Fujiwara no Tadanobu kō,” 25. See also “Ono no Sōjō shō’u gyōbō ga’u no shi,” 413, in Zoku Gunsho ruijū, vol. 12a, no. 318. These two sources offer disparate versions of the first line of the third couplet; the transcription and translation offered here are tentative. I wish to thank Professor Satō Michio of Keiō University for his assistance in the reading and interpretation of this and the preceding poem. 34. Ninkai conducted the rite on the fourteenth day of the sixth month, and rain fell on the eighteenth. He had nine similar successes in the period 1028 – 1044, earning for himself the appellation Ame sōjō. See Allan G. Grappard, “Religious Practices,” 536. 35. “Ono no Sōjō shō’u gyōbō ga’u no shi,” 414.
Terrains of Text in Mid-Heian Court Culture | 151 36. Grappard, “Religious Practices,” 536. 37. There is an account of this event in Eiga monogatari, bk. 31, “Tenjō no hanami.” See Eiga monogatari, 3:189 – 190. 38. The preface is preserved in Honchō monzui, bk. 11. See Honchō monzui chūshaku, ed. Kakimura Shigematsu; 2:600 – 603; and Honchō monzui in SNKBT 27:322. 39. Eiga monogatari, 3:190 – 191. 40. Ibid. 41. See, for example, the well-known hototogisu episode in which she threatens to give up writing occasional poetry altogether: Makura no sōshi, sect. 95, esp. 191 – 194; sect. 65 in Morris, Pillow Book, esp. 123 – 125. 42. Fujiwara no Sanesuke (957 – 1046) was one of those who objected to this ostentatious demonstration of Michinaga’s capacity to trump prior practices. For a detailed discussion of the relevant passage in Sanesuke’s Shōyūki (Chōgen 1/10/28), see Kawamura Hiroko, “Michinaga, Yorimichi jidai no byōbu uta,” 109 – 125. 43. Senzai wakashū, no. 960. See Senzai wakashū, ed. Katano Tatsurō and Matsuno Yōichi, 287. 44. “Kanna ninen rokugatsu tōka dairi utaawase,” nos. 3 – 4, in Heian chō utaawase taisei, ed. Hagitani Boku, 1:638. Hagitani discusses the problems of the exact dating of this contest but notes that it took place just days before Kazan’s forced abdication and precipitous raku shoku shukke (he took vows and had his head shaved at Hanayama Gangyōji on Kanna 2 [986] 6/23). The chief plotter Kaneie’s sons Michinaga and Michitsuna were both among the contest participants (ibid., 650). 45. “Chōgen gonen jūgatsu jūhachinichi Jōtōmon’in Shōshi kikuawase,” in Heian chō utaawase taisei, ed. Hagitani Boku, 2:818 – 819, 825, 829. 46. Mandai wakashū, no. 1251. There is uncertainty about the date of the excursion; it may have taken place in 1013, before Atsuakira’s retirement, or in 1018, about a year after his removal from the succession (Mandai wakashū, jō, ed. Yasuda Noriko, 205; Fukui, “Fujiwara no Tadanobu kō,” 26). 47. See Edward Kamens, Utamakura, Allusion, and Intertextuality in Traditional Japanese Poetry, 189 – 192, 198 – 207. 48. Eiga monogatari, bk. 19, “Onmogi,” in Eiga monogatari, ed. Yamanaka Yutaka, 2:353; William H. McCullough and Helen Craig McCullough, trans., A Tale of Flowering Fortunes, 2:596. The second of Tadanobu’s daiei poems explicitly calls for Shōshi to “live on for a thousand generations” (chiyo no manimani/yorozuyo no/nagarete sumeru). 49. Eiga monogatari, 2:371; see also McCullough and McCullough, Flowering Fortunes, 2:608. 50. Eiga monogatari, 2:413 – 425; see also McCullough and McCullough, Flowering Fortunes, 2:636. 51. Shin kokin wakashū,no. 1928; Shin kokin wakashū, 562. 52. Goshūi wakashū includes only Tadanobu’s poem (as no. 113; Goshūi wakashū, 43), but Eiga monogatari, bk. 31, “Tenjō no hanami,” presents the full exchange (Eiga monogatari, 3:202.) 53. Senzai wakashū, 273. 54. Eiga monogatari, 3:174; McCullough and McCullough, Flowering Fortunes, 2:771. 55. See note 5, above. 56. Jonathan Chaves suggests that a passage in the “Ongaku” chapter of Eiga monogatari describes women performing rōei (“Chinese Poets in the Wakan rōei shū,” 20). Nothing in the McCullough and McCullough translation of the passage (2:563) or in the original text (Eiga monogatari, 2:294) supports this interpretation: those singing rōei on this occasion, as usual, are male.
152 | edward kamens 57. Kokon chomonjū, 1:192 – 193. The dating of the passage is problematic because of inconsistent details: Kaneie was dead long before his granddaughter Jōtōmon’in commissioned the construction of the Tōhokuin at Hōjōji. 58. Makura no sōshi, 285; Morris, trans., Pillow Book, 175 – 176. 59. Makura no sōshi, 286; Morris, trans., Pillow Book, 176. 60. Makura no sōshi, 288; Morris, trans., Pillow Book, 178. 61. Makura no sōshi, ed. Matsuo and Nagai, 136; Morris, trans., Pillow Book, 89 – 90. 62. The couplet also appears in the mountain retreats (sanka) topical section of Wakan rōeishū. See Wakan rōeishū, 292; see also J.Thomas Rimer and Jonathan Chaves, eds. and trans., Japanese and Chinese Poems to Sing, 167. 63. Ibid. 64. See note 2, above. 65. Murasaki Shikibu nikki, 100 – 102; Bowring, Diary, 60. The dating of this passage is in dispute. 66. Murasaki Shikibu nikki ekotoba, 66 – 67. 67. Murasaki Shikibu nikki, 100 – 102; Bowring, Diary of Lady Murasaki, 60. 68. Ibid.
7
d Samuel C. Morse
The Buddhist Transformation of Japan in the Ninth Century The Case of Eleven-Headed Kannon
T
he significance of the Shingon and Tendai traditions in the history of Japanese Buddhist art during the early Heian period is indisputable. Yet it is important to acknowledge that those teachings were available only to a culturally privileged, literate male minority with close connections to the court. Temple histories and inventories as well as texts from the period describing popular Buddhist beliefs and practices, such as the Nihon koku genpō zen’aku ryōiki (Miraculous Stories from the Japanese Buddhist Tradition) and the Tōdaiji fuju monkō (Text of Buddhist Recitations from Tōdaiji), attest to the vitality of a Buddhism far different from that described in the official histories and in the biographies of Kūkai (774 – 835) and Saichō (767 – 822). Of particular importance for an understanding of the development of Buddhist culture throughout Japan in the early Heian period are works of art — the statues that served as objects of worship at the large number of temples that were established throughout Japan during the ninth century. These images both reveal a religious experience fundamentally different from the one articulated by texts and document the spiritual lives of the majority of the Japanese of the time, not just the ideas of a few great thinkers and their aristocratic patrons. These works of sculpture describe a Buddhism that flourished away from the centers of political and religious authority in the Heian capital. An assessment of this artistic evidence reveals that Buddhist sculpture in Japan underwent a number of momentous transformations at the start of the Heian period. One was primarily technical; wood replaced bronze, clay, and lacquer as the primary medium of the sculptor’s craft. A second transformation was iconographic; images of deities only occasionally worshiped in the previous century were produced in unprecedented numbers. A third transformation was geographic; statues of the period can be found in temples across Japan, frequently in sanctuaries in remote mountain locations or at temples associated with indigenous cults. These transformations must be understood in the context of changes in religious practice as Buddhist beliefs became more widespread as evidenced especially in the cult of the Eleven-Headed Kannon (Jūichimen Kannon). Numerous statues of this deity were enshrined in regional temples during the first two centuries of the Heian period, rapidly eclipsing images of other
154 | samuel c. morse
Figure 7.1. (Above) Eleven-Headed Kannon. Early ninth century. Wood. H. 100 cm. Hokkeji, Nara. Figure 7.2. (Right) Eleven-Headed Kannon. Early ninth century. Wood, single woodblock construction. H. 83.9 cm. Ryōsenji, Nara.
esoteric manifestations of Kannon such as Thousand-Armed Kannon (Senjū Kannon) and Fukūkenjaku Kannon, which were more widely worshiped in the eighth century. The majority of these statues were carved of unpainted wood, directly conveying to the devotee the nature of the material from which they were fashioned. Included in this group are some of the most distinctive works of early Heian sculpture, such as those at Hokkeji (figure 7.1) and Ryōsenji (figure 7.2) in Nara, Chōenji (figure 7.3) in Osaka, and Jikō Enpukuji (figure 7.4) in Wakayama, yet most of these images have been infrequently studied because there is little or no firm documentary evidence associated with them. The cult of Eleven-Headed Kannon in the ninth and tenth centuries also seems to have had a particularly strong local character. Nowhere was it more popular than at the northern end of Lake Biwa and in the provinces of Wakasa and Echizen along the coast of the Japan Sea immediately to the north and west. Works from this region include some that are relatively well-known such as the
The Buddhist Transformation of Japan in the Ninth Century | 155
Figure 7.3. (Left) Eleven-Headed Kannon. Early ninth century. Wood, single woodblock construction. H. 45 cm. Chōenji, Osaka Prefecture. Figure 7.4. (Above) Eleven-Headed Kannon. Ninth century. Wood with polychrome. H. 149.3 cm. Jikō Enpukuji, Wakayama Prefecture.
statues at Tadadera (figure 7.5) and Hagaji in Obama, but most, such as the statues at Keisokuji in Takatsuki-chō, Shiga Prefecture, and at the Futagami Kannondō (figure 7.6) in Fukui City have escaped the attention of most members of the art historical community. What, then, accounts for this rise in popularity in early Heian Japan of Eleven-Headed Kannon, a relatively obscure member of the Buddhist pantheon? Why are the majority of the extant images of the deity carefully fashioned from unpainted wood, and why are so many images of the deity concentrated in one particular region? To answer these questions it is first necessary to understand the place of the deity in the religious life of the preceding Nara period. Such an overview is essential because the discourse on the history of Buddhist
Figure 7.5. (Left) Eleven-Headed Kannon. Late eighth century. Wood. H. 154.0 cm. Tadadera, Fukui Prefecture. Figure 7.6. (Right) Eleven-Headed Kannon. Late ninth century. Wood. H. 185.3 cm. Futagami Kannondō, Fukui City.
The Buddhist Transformation of Japan in the Ninth Century | 157
practice in the early Heian period has been written from the perspective of the Shingon and Tendai religious traditions that eventually emerged as ascendant. The histories of the temples on Mount Kōtakami at the northern end of Shiga Prefecture exemplify this phenomenon. According to the Kōtakamiyama engi, the legendary history of the mountain compiled in 1441, the sanctuaries on the mountain were first founded by the mendicant monk Gyōki (668 – 749) and later revived by Taichō (682 – 767), who came from neighboring Echizen. They were then revitalized by Saichō, although today most Kōtakami temples are affiliated with the Shingon sect. Yet the early Heian images at the site — a statue of the Healing Buddha, two statues from a set of Twelve Divine Generals, and two statues of Eleven-Headed Kannon — are not connected with either the Tendai or the Shingon phases of Mount Kōtakami’s history, but instead are technically and stylistically related to imagery produced in Nara. Despite the later esoteric transformations of these sanctuaries, they were under the administrative control of Kōfukuji during the fifteenth century, a fact that is indicative of the continuation of a strong Nara influence on the region into the Muromachi period. Despite such evidence, both traditional histories and more-contemporary scholarship describe the Buddhism of the early Heian period as a time when the schools of Nara declined and the esoteric sects rapidly gained power and authority, and ignore the strong continuities that existed with the previous century.1 Much is made of Kūkai’s role in the esotericization of the great temples of Nara initiated by the founding of the Shingon’in at Tōdaiji in 810, and the institutionalization of Shingon rituals at court begun with the construction of the Shingon’in at the Imperial Palace in 834. However, less attention has been paid to the continued influence of Nara monks in the religious life of the Heian capital throughout the ninth century. For example, the Buddhas Names Ceremony (Butsumyōe), a rite of repentance held to seek protection for the nation, to ensure a good harvest, and to seek protection from pestilence, was held for the first time at the Imperial Palace in 838 and was presided over by monks who came from the great Nara temples.2 Indeed, during the ninth century it was monks from Nara, rather than adherents of the newly established Shingon and Tendai sects, who were primarily responsible for spreading Buddhism throughout Japan. For example, Tokuitsu (active during the early ninth century), a Hossō monk from Nara and one of Saichō’s chief critics, founded temples near modern Aizu-Wakamatsu, and Ken’na (active ca. 835 – 870), a monk from Gangōji, was active in Harima and established sanctuaries along the southern shores of Lake Biwa. Whereas the influence of Shingon and Tendai doctrine on the visual arts is readily apparent in works from later in the Heian period — the many lavishly ornamented versions of the Lotus Sūtra or the Mandalas of the Two Worlds — their influence on the art of the early Heian period is far less certain. Indeed, while the extant objects dating to the ninth century that were produced in response to the teachings of Saichō and Kūkai number fewer than thirty and are concentrated at a small number of temples, statues of Eleven-Headed Kannon are found at temples across Japan. This artistic evidence suggests that worship of the deity
158 | samuel c. morse
flourished throughout the ninth and tenth centuries and that the cult was more widespread and far more popular than any promoted by the esoteric communities. The goal of this chapter is to bring Buddhist sculpture of the early Heian period out of the shadow of Saichō and Kūkai and, through an examination of the circumstances behind the creation of these statues, to learn something of the spiritual lives of the majority of the Japanese, particularly those living at some distance from the Heian capital, at the start of the Heian period.
Worship of Eleven-Headed Kannon in Early Heian Japan The patterns of devotion to Eleven-Headed Kannon in the ninth century were first established in the former capital of Nara. While a few images date to the beginning of the eighth century, sūtras describing the deity and the benefits from its worship were not known in Japan until the mid-Nara period. These texts were introduced by monks such as Dōji (? – 744) and Genbō (? – 746) upon their return from lengthy periods of study in China. For example, after twenty-eight years in Chang’an, Genbō returned in 735 with more than five thousand texts and commentaries.3 Furthermore, documents in the Shōsōin indicate that a Jūichimen kyō (Eleven-Headed Kannon sūtra) was first copied in one of the scriptoria of Nara in 733, and historian Ishida Mōsaku believes this text was the Avalokiteśvaradaśamukhhadhāran. ī (Jūichimen shinjū kyō) translated into Chinese by Xuanzang (602 – 664) in 654.4 Four years later, a text clearly identified as Xuanzang’s translation as well as the Jūichimen kanzeon shinjū kyō, an older version of the sūtra translated by Yasogupta (active in the late sixth century) in the 570s, are also recorded as having been transcribed.5 The records from the scriptoria document requests made by monks for particular texts, and the frequency with which the Eleven-Headed Kannon sūtras appear indicate that by the middle decades of the eighth century, the Japanese were certainly aware of the benefits that resulted from worship of the deity. A different version of the Jūichimen shinjū kyō included as the fourth chapter of the esoteric compilation known as the Dāranī jikkyō and translated into Chinese by Atikūta (mid-seventh century) was available in Japan by 742.6 The most important commentary on the text, the Shiyi mian shenzhou xin jing ishu (Jūichimen shinjū kyō gishō), attributed to Xuanzang’s disciple Huizhao (651 – 714), is recorded as having been present in Nara as well.7 A fourth text, a translation combining elements of Xuanzang’s version with the ritual instructions included in Atikuta’s version, was undertaken by Amoghavajra (705 – 774) sometime in the 750s. This sūtra was brought to Japan in the ninth century, first by Kūkai and then later by Ennin (794 – 864) and Enchin (814 – 891).8 The four translations of the sūtra share many similarities. In each the vow taken by Eleven-Headed Kannon emphasizes overcoming both illness and suffering. All include incantations known as dhāran.ī that were believed to contain the essential powers of the deity. In the version translated by Yasogupta, the text reads,
The Buddhist Transformation of Japan in the Ninth Century | 159
Kanzeon spoke to the Buddha, saying, Lord, I possess a secret dhāran.ī called Eleven Heads. This secret vow has been spoken by eleven myriads of buddhas, and now I will relate it. It is for the benefit of all sentient beings. It is because I wish to cause all sentient beings to concentrate on good dharma. It is because I wish all sentient beings to have no troubles. It is because I wish to eradicate sickness from all sentient beings. It is because I wish to remove and eradicate all hindrances, calamities, and evil dreams. It is because I wish to eradicate all untimely deaths and deaths from illness. It is because I wish to eradicate those with evil intentions and to cause them to be under control. It is because I wish to banish the trouble caused by various demons and evil spirits and to never allow it to occur again.9 Each text also describes ten benefits that will accrue to the devotees who, each morning, purify their bodies and recite the dhāran.ī of the deity 108 times. Seven benefits relate to physical protection — against illness; evil enemies and threats; harm from the poison of insects or spells of demons; harm by swords or staffs; drowning; fire; and untimely death. Another benefit centers on the acquisition of riches — receiving wealth, jewels, clothing, and food in unlimited supplies. Two vows are more spiritual — receiving salvation from the myriad buddhas and words of praise from various exalted beings. The texts further provide instructions for worship and pharmaceutical prescriptions for curing specific diseases.10 The translation by Xuanzang, for example, states that drinking ritually purified cow’s milk while reciting mantras prevents all diseases. Earaches and migraines abate when incense made of greenwood sap is mixed with birch bark and applied in the ear. A broader range of illnesses is healed by tying a cord to the head of an image and around the neck of the afflicted while reciting mantras. Other ritual remedies described in the sūtra include bathing a statue of the deity with the urine of a bull first presented to it during a mantra recitation to remove evil dreams, hindrances, and illnesses and washing one’s body with water first used to clean incense rubbed over the surface of an image to discard impediments. In addition to providing individual remedies, Eleven-Headed Kannon’s power to subdue epidemic diseases was of particular importance to its subsequent popularity. The translation by Xuanzang states, Moreover, if human or bovine pox occurs in the land one should kindle a fire of ninba wood in front of this image [of the deity]. Then taking up another piece of wood of this tree one should cut it into 1,008 pieces. Taking up each piece in succession one should rub them with mustard-seed oil making a recitation and then throw them one at a time into the fire until they no more are left. Then, one should take up a red cord and make knots for seven recitations. If one makes a knot for each of seven recitations and then ties the cord to the top of the head of the Buddha atop the head [of the statue], epidemic disease should recede and completely disappear. Once the epidemic has completely dissipated, one should take off the ritual cord.11
160 | samuel c. morse
For those with sincere faith in the deity’s therapeutic and thaumaturgic powers, Eleven-Headed Kannon also responded directly to a wide range of physical and emotional needs; thus it appealed to the entire populace and not only to members of the imperial court. All four versions of the sūtra stress that efficacious worship of Eleven-Headed Kannon requires that an image of the deity be carved of white sandalwood. The text translated by Yasogupta describes the process in the following terms: These good men and these good women should make an image of Kanzeon using white sandalwood. The wood must be fine grained, well aged, and of a rectangular shape. The length of the body should be one shaku [foot] and three sun [inches], and the statue should be made with eleven heads. The three heads in the front should be made with the faces of a bodhisattva. The three heads on the left should be made with angry faces. The three heads on the right should be made with the faces of a bodhisattva, but with tusks rising upward. The head in the rear should be made with a face laughing wildly. The head on the top should be made with the face of a buddha. Each of the faces should be facing forward, and around the body there should be a mandorla. Each of the eleven heads should wear a crown, and in the crowns there should be images of Amida. Kanzeon should hold a water vessel in its left hand, and from the mouth of the vessel should rise a lotus blossom. The right arm should be extended, and it should pass through jewelry. The hand should be in the gesture of the absence of fear. The body of the image should be carved with jeweled ornaments.12 The three early translations further state that the image should have two arms; however, the later translation by Amoghavajra dictates that the image should be fashioned with four. The translations by Atikūta and Amoghavajra also include detailed instructions for performing fully developed esoteric rituals dedicated to the deity. Yet chapter 4 of the Darani jikkyō was rarely copied as an independent text, and no four-armed images of the Eleven-Headed Kannon are known from the ninth and tenth centuries. Thus, it is possible to conclude that neither Atikūta’s nor Amoghavajra’s texts had any influence on the sculpture of early Heian Japan. As Ryūichi Abé and Yoshida Yasuo have pointed out, certain esoteric manifestations of Kannon were worshiped during the eighth century well before a more complete understanding of esoteric doctrine was accessible.13 In addition to Eleven-Headed Kannon, Thousand-Armed Kannon and Fukūkenjaku Kannon were frequently invoked because they were believed to respond to the immediate needs of the populace, and their dhāran. ī were thought to possess therapeutic powers. Analysis of the records of the presentation of candidates for ordination (upasoku kōshinge) dating from the second quarter of the eighth century, for example, reveals that memorization of the sūtras and dhāran. ī related to all three deities was a part of the curriculum for novitiates.14 Statues of all three manifestations of Kannon were also well-known to the residents of Nara: Fukūkenjaku Kannon at the Sangatsudō at Tōdaiji, Eleven-Headed Kannon at the Jūichimendō
The Buddhist Transformation of Japan in the Ninth Century | 161
on the grounds of Saidaiji, and Thousand-Armed Kannon in the Main Image Hall at Tōshōdaiji. Yet during the first two centuries of the Heian period, only a few images of Thousand-Armed Kannon and Fukūkenjaku Kannon were sculpted. In contrast, numerous images of Eleven-Headed Kannon were produced, indicative that by the ninth century it was considered to possess more efficacious powers than its two counterparts. Rites of repentance dedicated to Eleven-Headed Kannon were the greatest cause for the deity’s rise in popularity. These rituals were held in the Nara capital perhaps as early as 737 and became institutionalized at the imperial court by the early 750s. Rites of repentance were one of the most distinctive aspects of Buddhist practice in early Japan, dedicated not only to Eleven-Headed Kannon, but also to a variety of deities including the Healing Buddha (Yakushi), Amida, and Kichijōten and were some of the few rituals to be sponsored by both the ruling elites in the capital and residents of the provinces. In this context, Nakamura Hajime makes a number of important observations, frequently cited by other scholars, about the differences between rites of repentance as they were practiced in Japan and their original form in India.15 He points out that whereas the Indian rite of desanā was held within the monastic community to absolve individual monks and nuns of their personal indiscretions, in China and in particular, in Japan, these rites were performed to seek more-immediate benefits through the collective purification of evil karma. For those in the capital, the goals of the practice were most often national in character — the guarantee of the well-being of the emperor or the protection of the nation. However, for those in the provinces, the aims were of a practical and immanent nature — the alleviation of drought, the curing of illness, or the eradication of disease. The inclusion of indigenous deities as participants in these rites was also distinctive to the practice in Japan.16 Satō Michiko observes that the practice of expunging the defiled karma of the previous year, an essential aspect of rites of repentance, strongly resembles Japanese ancient religious practices that required purifying acts to be held at the palace and sacred locations throughout the land before the start of the new year. In this sense, she suggests that the rites of repentance simply recast existing customs in Buddhist terms. One result of this convergence was that the ritual came to be regularly held in the second month as part of a ceremony to mark the start of the new agricultural year and was given the designation shunie (ceremony of the second month).17 The best known of the rites of repentance, instituted in the eighth century and still held at the Nigatsudō on the grounds of Tōdaiji, is popularly known as the omizutori. Celebrated in the second month of the lunar calendar and dedicated to Eleven-Headed Kannon, this shunie ritual seeks the assistance of the myriad kami and the aid of various deities in the Buddhist pantheon to ensure the success of a new agricultural cycle and to gain protection for the nation. This rite was first held in 752 not at Tōdaiji, but at a chapel in the Queen Consort’s Household Agency (Shibichūdai), and, as Horiike Shunpō points out, although the ritual originally was private in nature, it eventually took on a public function when it was moved to Tōdaiji in 760.18
162 | samuel c. morse
The popularity of the rite of repentance is documented in various sources from the eighth century. When the health of Emperor Shōmu (701 – 756; r. 724 – 749) deteriorated in the late 740s and early 750s, his consort, Kōmyō (701 – 760), frequently sought the assistance of the Buddhist clergy to effect a cure. These measures were undertaken to ensure that Shōmu would live to witness the dedication of the Great Buddha at Tōdaiji. Thus, in 751, Kōmyō ordered forty-nine learned monks to gather at Shin Yakushiji (dedicated to the Healing Buddha and originally founded to cure a previous illness of the emperor) to perform a ritual to prolong life (zoku myōhō).19 These measures seem to have been ineffective, so in the first month of the next year, one thousand monks and nuns were ordained as a pious gesture and as a means of gaining karmic benefits, but again Kōmyō’s efforts on behalf of the ailing emperor were unrewarded.20 Having appealed to the Healing Buddha with little effect, a month later, she ordered Jitchū (? – 824), abbot (bettō) at Tōdaiji, to establish a chapel for rituals of repentance dedicated to Eleven-Headed Kannon as part of a new attempt to prove a cure for the emperor. The existence of the chapel is confirmed by a requisition for a copy of the Darani jikkyō to be used at the Queen Consort’s Household Agency in the following year.21 Kōmyō’s appeal to Eleven-Headed Kannon was apparently successful, for Shōmu lived for four more years. Other rites of repentance dedicated to Eleven-Headed Kannon are described in the Nihon ryōiki, which provides evidence that the deity was worshiped on a popular level distant from the capital.22 During the reign of Emperor Shōmu, Dharma Master Daie of Yakushiji was summoned by the nuns of Sayadera in Kii Province to perform a rite of repentance dedicated to Eleven-Headed Kannon. Kyōkai, the author of the text, tells of a wicked husband who, learning that his wife had gone to the temple to join the congregation participating in the event, castigates Daie and accuses him of having seduced his wife. This evil man then drags her home and rapes her; as a consequence of his actions an ant bites his penis, and he dies in acute pain. While Kyōkai’s primary concern is a moral one and focuses on the retribution that resulted from opposing the Buddha’s teaching, the tale is also significant because it documents rites of repentance that were open to commoners, unlike the exclusive rituals held at the Queen Consort’s Household Agency. Another rite of repentance dedicated to Eleven-Headed Kannon was held at Kojimadera, situated in the mountains at the southern edge of the Yamato plain.23 As recorded in the Enryaku sōroku, a biographical text compiled during the reign of Emperor Kanmu (737 – 806; r. 781 – 806), the principal imperial consort, Fujiwara no Otomuro (760 – 790), constructed a hall on the temple’s grounds during the last decades of the eighth century so that rituals could be held in it every fall and spring. The temple had been founded more than twenty years earlier by a certain Hōon (? – 795), who was known as a faith healer. According to the Genkō shakusho, an early-fourteenth-century Buddhist hagiography, Hōon took the tonsure at fifteen and became a monk at Kōfukuji, where he studied under the tutelage of Genbō, who was also noted for his abilities as a mendicant.24 After Genbō was banished from Nara in 745 for becoming too involved in affairs of state,
The Buddhist Transformation of Japan in the Ninth Century | 163
Hōon became disillusioned with the religious life in the capital and retreated, at the age of thirty, to the mountains of Yoshino, where he studied the dhāran. ī of the various manifestations of Kannon. But in 752 Hōon was summoned to the palace to attempt to cure Emperor Shōmu’s recurring illness. He must have been successful, for he was offered an official ordination. The monk declined this honor and instead returned to the mountains where he founded Kojimadera and enshrined in its small hall a sandalwood (danzō) image of Eleven-Headed Kannon sixteen feet in height.25 Forty years later Hōon’s skills at healing were again put to good use when he restored Emperor Kanmu to health after the physicians in attendance at the palace were unable to relieve his suffering. Both tales clearly include considerable fictional elaboration, but they provide evidence of the popularity of Eleven-Headed Kannon among both commoners and the court nobles. Both stories indicate that the deity was particularly valued by monks associated with Hossō teachings, and, as the biography of Hōon makes clear, Eleven-Headed Kannon was often invoked because of its therapeutic powers. These points have particular bearing on the rite of repentance dedicated to the deity said to have been held in 737 by Taichō, the monk associated with Mount Kōtakami. Taichō was born in the village of Asōzu in the province of Echizen in 682.26 Precocious in his attachment to the Buddhist faith, he is said to have fashioned Buddhist images out of mud as a child; when he was fourteen, Eleven-Headed Kannon is reported to have appeared to him in a dream. In 716, when Taichō was thirty-four, the kami of Hakusan, the sacred mountain situated where modern Fukui, Gifu, and Ishikawa prefectures join, appeared to him in another dream; and when he met the deity some days later, she was in the guise of Eleven-Headed Kannon. Taichō first visited the capital in 722, when he cured an illness afflicting Empress Genshō (680 – 748; r. 715 – 724) and as a result received the title of dharma master (zenji). In 736 the monk again journeyed from his native Echizen to study in Nara, where he received from Genbō a copy of an Eleven-Headed Kannon sūtra. When the great smallpox epidemic threatened the capital in the next year, Taichō is said to have performed the first rite of repentance dedicated to Eleven-Headed Kannon held in Japan, for which he was again rewarded with high monastic rank. Taichō subsequently returned to his home province, where he eventually died in 767. Like the account of Hōon, much of the story of Taichō’s career is certainly fiction, yet enough details of his life correspond to information in other, morereliable sources to conclude that certain aspects are in all likelihood true: a mendicant monk from Echizen had a vision that prompted him to travel to the capital to seek out religious texts, and while he was there he participated in the rituals held in the capital to attempt to quell the smallpox epidemic that ravaged Japan in 737 and that claimed the lives of the heads of the four branches of the Fujiwara clan. Supporting this conjecture is the fact that two versions of the Eleven-Headed Kannon Sūtra were first copied in that very same year and that Eleven-Headed Kannon remains the Buddhist manifestation of the kami of Mount Hakusan. The Taichō legend is just one of a number of connections
164 | samuel c. morse
between Nara and Wakasa, Echizen, and the northern end of Lake Biwa during the eight and ninth centuries, and it helps to provide a context for the popularity of the Eleven-Headed Kannon cult outside the capital at the time. Many estates under the control of Tōdaiji were established in the Hokuriku region, including the Kusooki estate just to the south of the modern city of Fukui. Within its original borders is the Futagami Kannondō, which houses the earliest extant statue of Eleven-Headed Kannon in Echizen (see figure 7.6). Carved from a single solid block of hinoki (cryptomeria) probably stained red, the statue can be dated to the end of the ninth century by the distinctive treatment of the drapery, which juxtaposes broad, smooth areas with a section covered with alternating sharp and rounded folds.27 Although the specific circumstances surrounding the sculpting of the image are unknown, connections between Echizen and the Nara capital as well as the emphasis in the Taichō legend on the Eleven-Headed Kannon cult provide a relevant context. Although it is impossible to correlate the sculpting of the image with a particular event, epidemics were endemic to Japan throughout the ninth century, and it is not hard to imagine the work’s having been fashioned in response to a catastrophic event.28 The Taichō legend provides a context for a number of statues of Eleven-Headed Kannon housed in temples clustered at the northern end of Lake Biwa. The most famous of these sanctuaries, Dōganji, is said to have been founded on Mount Kōtakami in 736 by Taichō at the behest of Emperor Shōmu to combat the great smallpox epidemic. Among five other temples near the peak of the mountain was Kannonji, also said to have been founded by Taichō, which counted Keisokuji and Shakudōji among its associated ritual centers. Like Dōganji, both Keisokuji and Shakudōji were later moved off the mountain, and both house statues of ElevenHeaded Kannon.29 The Eleven-Headed Kannon at Keisokuji, carved from a single block of hinoki, shares many formal characteristics with the statue at the Futagami Kannondō and can also be dated to the end of the ninth century. As for the statue at Shakudōji, which was sculpted from a single block of keyaki (zelkova) and then polychromed, the folds of the drapery are more two dimensional, and the patterns they establish are less assertive, indicative of a date in the middle of the tenth century. Other temples traditionally associated with Taichō are found around the shores of Lake Biwa and in the southeast corner of Kyoto Prefecture along one of the early routes linking Ōmi with Nara.30 The existence of a group of sanctuaries that share a common history and often share a common object of worship attests to the credibility of the Taichō legend and the vitality of the Eleven-Headed Kannon cult in the region during the ninth and tenth centuries. The influence of Nara on the religious culture of the region is further confirmed by the important role played by Wakasa in the omizutori ritual. According to the Tōdaiji yōroku, the late twelfth-century history of Tōdaiji, on the first day of the fourteen-day rite of repentance dedicated to Eleven-Headed Kannon, Jitchū summoned the kami throughout Japan to be in attendance. All appeared with the exception of Wakasa Hiko (called Onyū Myōjin in the text), who was away fishing. It was not until the ritual was almost finished that Wakasa Hiko realized his mistake. To apologize he and his consort took the form of black and
The Buddhist Transformation of Japan in the Ninth Century | 165
white cormorants and miraculously flew underground from Wakasa to Nara. They thereby connected the Negori River to a two-shafted well, known as the Wakasa well and situated on the hill behind the Great Buddha Hall, thereby providing a permanent source of freshwater for use in the ceremony. The ritual “sending” of water to Tōdaiji has been held at Jingūji in Obama at least since the last quarter of the seventeenth century and includes a rite of repentance dedicated to the Healing Buddha and a procession of pine torches along the Negori River. Like other tales describing the early history of Buddhism in Japan, much of this story is in the realm of legend. Yet on a map of Tōdaiji dating to 756, the Wakasa well is clearly labeled, an indication that the tradition must be accorded some authenticity. Moreover, the credence given to the tradition in medieval times is attested to as well by the inclusion of the two cormorants next to the well in the late Kamakura-period Nigatsudō Mandala.31 The statue of Eleven-Headed Kannon at Tadadera in Obama (see figure 7.5) provides additional proof of the influence of the capital on the region and evidence that the Eleven-Headed Kannon cult had penetrated there by the second half of the eighth century. According to temple legends, Tadadera was founded during the first reign of Empress Shōtoku (718 – 770; r. 749 – 758, 764 – 770) when a triad of the Healing Buddha was installed as the main object of worship at the temple. The three statues that make up that triad today are all stylistically different, suggesting that the present grouping is a later reconfiguration: whereas the Healing Buddha and Kannon statues date to the first part of the ninth century, the formal characteristics of the Eleven-Headed Kannon indicate that it was sculpted close to the time of the temple’s legendary founding.32 Carved from a single solid block of unpainted hinoki, this statue shares certain formal characteristics with one of the earliest images of Eleven-Headed Kannon known in Japan: the painting on the interior wall of the Image Hall at Hōryūji. On both works the heads are small in proportion to the bodies, and the scarves delineate two loops immediately below the waist and are tightly gathered on the forearms. However, the stiff frontal pose of the statue, the ungainly proportions of the body, and the awkward treatment of the edge of the drapery at the top of the skirt betray the provincial origins of the work. The statue of the Eleven-Headed Kannon at Tadadera is just one of a number of early images of Eleven-Headed Kannon found in the Wakasa region. Also in Obama is Hagadera, which according to its temple legend was founded by Gyōki in 717 at the behest of Empress Genshō. Dating to the middle decades of the ninth century, the statue of Eleven-Headed Kannon that serves as the main image of the temple today also perpetuates both formal and technical characteristics of sculpture produced in Nara during the second half of the eighth century.33 While the image is carved from a single block of hinoki, many of the surfaces, particularly on the skirtlike garment and the scarves in direct contact with the body, have been modeled with lacquer paste in a manner reminiscent of wood-core dry-lacquer sculpture.34 Moreover, the folds of the drapery are deeply carved and set up repetitive patterns over the legs. These two images provide further evidence of the active worship of Eleven-
166 | samuel c. morse
Headed Kannon in the Wakasa region in the late eighth and ninth centuries. The distinctive technical and formal qualities of the works indicate that the statues were not brought to Wakasa from the capital, but that local sculptors produced them to meet the religious needs of the local population. It is possible to imagine a situation similar to one described in tale eleven in the second book of the Nihon ryōiki. Kanki was a monk from the Nagusa district in Kii Province, who was both “gifted in carving...[and] such a learned monk that he fulfilled the role of a speaker in a ceremony and was influential among the people.”35 During the reign of Emperor Shōmu, he carved statues of Shakyamuni and his attendants for his local temple, Nōōdera, in his home village, and in 779 he began work on a ten-foot image of Eleven-Headed Kannon. Although Kanki died before he could complete the project, a talented carver from his village subsequently finished the image on his behalf. Both statues in Obama were probably produced under similar circumstances, for stylistically (Tadadera) and technically (Hagadera) they deviate from established metropolitan norms. The statues and the texts describe a pattern that would frequently be repeated throughout the early Heian period. As metropolitan practices penetrated into the more remote parts of Japan, their forms became altered in response to local conditions and needs. In contrast to the large amount of textual evidence documenting the ElevenHeaded Kannon cult in the eighth century, records from the ninth century are much scarcer. Many extant statues, however, attest to a rise in Eleven-Headed Kannon worship throughout Japan at the time, and the textual evidence that does remain clearly indicates that the deity continued to be invoked for both thaumaturgic and therapeutic ends both in the capital and in the provinces. For example, the increased institutionalization of the worship of the deity can be ascertained from an edict promulgated by Emperor Ninmyō (810 – 850; r. 833 – 850) in 837: There is nothing more efficacious in bringing peace to the people and to the realm than the mystical power of Eleven-Headed Kannon. I request that at each of the kokubunji in the Five Home Provinces and those along the Seven Roads, seven monks who have undergone suitable training perform the ElevenHeaded Kannon ritual for seven days and seven nights.36 Unfortunately, the text does not tell specifically what prompted the emperor to make such a declaration. The fact that it was issued in the second month, however, permits speculation that he might have been seeking to establish rites of repentance as part of the shunie to mark the beginning of a new agricultural year. This record also demonstrates that the kokubunji, the national system of monasteries established by Shōmu during the mid-eighth century and that once functioned exclusively as mechanisms for state control as well as outposts of metropolitan culture, were being used for rituals that responded more directly to the immediate needs of the populace during the ninth century. Rites seeking the direct intervention of Kannon and Eleven-Headed Kannon, and of the Healing Buddha in particular, were held with considerable frequency at the kokubunji throughout the early Heian period. Eventually these practices transformed the
The Buddhist Transformation of Japan in the Ninth Century | 167
temples to such a degree that statues of the Healing Buddha replaced images of the Historical Buddha at the majority of the kokubunji across Japan.37 The benefits of Kannon’s power continued to be sought in times of crisis as well. In the second month of 849 the Bureau of Divination alerted the emperor that there was reason to believe that epidemics would be severe and that floods would occur in the fourth and fifth months.38 In his edict responding to this portent, the emperor noted that recently many young people had died from disease, and he instructed that the powers of the kami and the buddhas be enlisted to prevent such events from recurring. Offerings were made to shrines; and at all the kokubunji, kokubuniji (national nunneries), and jōgakuji (officially sanctioned temples), the Lotus Sūtra and other texts for the protection of the nation were recited during the day, and Kannon was worshiped at night for a period of seven days. Fourteen years later, in 863, Ken’ei, a lecturer (kōshi) from the province of Hōki, reported to the emperor the difficulties experienced by peasants in his province due to poor harvests and outbreaks of pestilence. To alleviate their suffering, Ken’ei stated that he had used some of the income he received to make paintings of Kannon and of the Thirteen Thousand Buddhas, which he had placed in the kokubunji of the province, but that he needed more funds to perform rituals. In a world where disease was a constant source of worry, the protection provided by Eleven-Headed Kannon would have been particularly welcome; thus, it is not surprising that the emperor granted Ken’ei’s request.39 Assistance from Kannon and the Healing Buddha was sought again in the twelfth month of 875, when the Bureau of Divination determined that the upcoming year would be particularly inauspicious.40 To prevent the occurrence of drought, pestilence, disasters, and the outbreak of war, offerings were made to the shrines in each province, and for three days each kokubunji, kokubun niji, and jōgakuji was directed to have seven monks recite the Kongō Hannya kyō during the day and chant the names of the Healing Buddha and Kannon during the night. Although no specific mention is made of which manifestation of Kannon was used in these rituals, many of the statues of Eleven-Headed Kannon can be dated to the middle decades of the ninth century, a time when outbreaks of disease were endemic. The existence of these images suggests that while the court was attempting to deal with these calamities on a national level, individuals or religious confraternities commissioned images of Eleven-Headed Kannon to seek protection on a regional one. The cult, which had first been established in the centers of religious and political power in Nara, had been adopted by the devout of all social classes across Japan.
Sandalwood Imagery and Eleven-Headed Kannon The use of unpainted wood was another distinctive aspect of the Eleven-Headed Kannon cult throughout the early Heian period. Unpainted sculpture in aromatic woods, frequently referred to in texts as danzō (sandalwood images), has a long history in the Buddhist world. It can be traced to India, where tradition claimed
168 | samuel c. morse
that the legendary “first image” of the Buddha, commissioned by King Udayana of Kauambi, was carved of “ox-head” sandalwood (sendan) and was five shaku in height. This tale, recorded in the Zōitsu Agon kyō, accorded high value to unpainted statues fashioned from finely grained aromatic wood.41 The Chinese pilgrim Faxian (early fifth century) claimed to have seen the original statue on his travels to India in 400, and a century later, Emperor Wu of the Southern Liang dynasty (502 – 556) sent two deputies to India to claim the image but was forced to accept a copy in purple sandalwood (shitan) instead.42 Xuanzang reported that he had come across a statue said to be Udayana’s first image in Khotan in 644 and that when he reached Kauambi, he discovered the actual figure of the Buddha carved of sandalwood that had been commissioned by King Udayana. Xuanzang played an important role in popularizing sandalwood imagery when he brought “four carved sandalwood images of the Buddha” back to Chang’an from India in 645.43 Thus, the legend of the first image marked the beginning throughout East Asia of a practice of conferring high sanctity to unpainted statues carved from aromatic woods. Another text that describes the merits of image making, the Tathāgata Pratibimba Pratisthanuśamsa Sūtra (Daijō zōzō kudoku kyō), records that the first image of the Buddha was made of purple sandalwood (shitan) and measured seven shaku in height.44 Translated into Chinese in the late seventh century, this sūtra first appeared in Japan in 736, one year before the appearance of the Zōitsu Agon-kyō and three years before the record of Xuanzang’s journey.45 It was through these texts and the various versions of the Eleven-Headed Kannon sūtras that the Japanese of the mid-eighth century learned of the continental tradition of the sandalwood imagery. In fact, the religious authority accorded to sandalwood imagery led the authors of some later sources to associate danzō with the earliest history of Buddhism in Japan. According to the Fusō ryakki, compiled by Kōen (? – 1169) in the eleventh century, an eight-shaku piece of sandalwood (danboku) floated ashore on Awaji Island in the fourth month of 595.46 The local residents, not recognizing its rarity, wanted to use the log for firewood; however, Prince Shōtoku, understanding that it was “the fragrant wood sendan and that it came from India,” ordered a Korean sculptor from Paekche to carve it into an image of Kannon several shaku in height. Shōtoku then enshrined the finished statue at Hisoji in the Yoshino district, where it was said to give off a miraculous light intermittently. The exotic origins of sandalwood are also alluded to in the Tōnomine ryakki, the early history of the Danzan shrine-temple complex. According to the text a three-shaku image of Nyoirin Kannon made of white sandalwood (byakudan) and housed in one of its halls, the Nyohōdō, was found floating in the ocean by the priest Jōe (? – 711) when he returned to Japan from Tang China in 665.47 While this statue is no longer extant, a sandalwood image of Eleven-Headed Kannon formerly in the shrine’s possession is generally believed to have been brought back by the monk.48 The artist seems to have attempted to follow specifically the prescriptions in the Eleven-Headed Kannon sūtras in crafting the small and precisely carved image: finely detailed jewelry sheathes much of the body, and
The Buddhist Transformation of Japan in the Ninth Century | 169
the left hand holds a water bottle. The only variation is in the treatment of the right hand, which clutches a rosary rather than being held in the gesture of allaying fear. Two other danzō were also brought to Japan from the continent during the eighth century. One, which most scholars believe to be a rare iconographic variant of Eleven-Headed Kannon, is the Nine-Headed Kannon at Hōryūji, recorded in the 747 inventory as having been brought from Tang China in 719.49 The provenance of the second, an Eleven-Headed Kannon in the collection of Jinpukuji in Yamaguchi Prefecture, is less clear.50 In their formal qualities, both follow the prescriptions of the sūtras; however, while the Hōryūji work is sculpted of sandalwood, the Jinpukuji statue is made of cherry. These three Chinese works provide evidence that the eighth-century Japanese learned about Eleven-Headed Kannon imagery not only from texts, but also from actual works of art. Danzō appear in Japan with increased frequency during the second half of the eighth century. The Amida kekaryō shizaichō, an inventory of materials used in rites of repentance dedicated to Amida compiled in 767, records two danzō statues of Kannon in the Amidadō, founded in 741 and later incorporated into Tōdaiji.51 In 752, Hōon enshrined his monumental danzō of Eleven-Headed Kannon at Kojimadera, and in the following year the Chinese monk Jianzhen (Ganjin; 688 – 763) arrived in Japan with a white sandalwood (byakusendan) image of Thousand-Armed Kannon.52 A story in the Nihon ryōiki dated to 758 likens a statue of the Healing Buddha that had lost both ears and been buried in the sand to the “danzō statue made by King Udayana (uten danzō).”53 Other records of danzō in textual sources include portable shrines of Yakushi and Fukūkenjaku Kannon, a set of the Four Divine Kings and an eight-shaku statue of ThousandArmed Kannon.54 None of these images remains today, but clearly not all of them were made of actual sandalwood, for it does not grow in Japan; and in South Asia, where it is indigenous, it never grows to great size. With no sandalwood available to them except for a small number of blocks brought from the continent, the Japanese sculptors of the ninth century were obliged to turn to native materials for their images of Eleven-Headed Kannon. In fact the use of woods other than sandalwood for danzō is not without scriptural justification. Huizhao’s commentary on Xuanzang’s translation of the Eleven-Headed Kannon Sūtra authorizes the use of baimu (Japanese hakuboku) when white sandalwood cannot be obtained: It was asked, “If there is no white sandalwood, out of what kind of wood should the people of this country make their images?” It was answered, “If by any means sandalwood can be obtained, they should make them from it. If they cannot get it reasonably, then they should make the images out of baimu.”55 In both China and Japan, baimu was used to refer to different varieties of cypress (hinoki and sawara) and to Japanese nutmeg (kaya), all of which are extremely light in color and highly aromatic.56 Apparently, the Chinese sculptor of the Jinpukuji statue also found cherry appropriate for an image of Eleven-Headed Kannon. Huizhao’s commentary, the early tales, and the statues themselves thus offer
170 | samuel c. morse
a picture of how the Chinese and Japanese interpreted the term danzō during the eighth and ninth centuries. When one of the varieties of sandalwood was available, they sculpted images out of this exotic aromatic wood; but because it was so rare in both countries, sculptors looked for native materials that shared some of sandalwood’s properties: light color, fine grain, and high aroma. In Japan in particular, this concept of sacred wood seems to have engendered a sympathetic response among a people who had long regarded great trees as divine and had developed unusual sensitivity to the inherent properties — aesthetic as well as religious — of the cypress, nutmeg, and cryptomeria that grew majestically throughout the land. The artistic evidence from the ninth century clearly confirms this interpretation. Only one Japanese image of Eleven-Headed Kannon dating to the ninth century was carved from sandalwood, a statue for which the provenance is completely unknown.57 Both in scale and in the detail of the carving, this work, now owned by the Nara National Museum, closely follows the prescriptions of the Eleven-Headed Kannon sūtras. All other early Heian works use substitute materials, especially hinoki and kaya, and although some are small in scale, others are much larger. Representative of the group of small-scale images is the statue housed today at Chōenji in Osaka (see figure 7.3). Carved from a solid block of hinoki, the image measures forty-five centimeters in height. The drapery and scarves that cover the volumetric body are characterized by deeply carved folds, which set up a variety of assertive patterns, and the robe is further embellished by whorl-shaped details that are meant to represent overlapping layers of cloth. Like many early Heianperiod works of sculpture, nothing is known of the history of the statue; but on the basis of style it can be dated to the middle decades of the ninth century. Other small-scale images of Eleven-Headed Kannon dating to the ninth century include the statue at Ryōsenji (see figure 7.2), carved from a solid block of kaya, remarkable for its exaggerated proportions and glowering expression, and the statue at Kaijūsenji, also made of kaya and distinguished by its precisely carved drapery and animated pose.58 The Eleven-Headed Kannon at Hokkeji (see figure 7.1) is representative of works on a larger scale. Standing just over one meter in height, the image is noteworthy for its exaggerated contrapposto; full, fleshy body; and wide-open eyes, the pupils of which are fashioned from copper. The block of kaya from which the statue was carved was left unpainted; however, a yellow pigment seems to have been rubbed into the surface to make the wood appear more like sandalwood. This treatment was used on other ninth-century works including the statue of Eleven-Headed Kannon housed today at Jikō Enpukuji in Wakayama Prefecture (see figure 7.4). With no large blocks of sandalwood available, the sculptors of the ninth century were obliged to invent a variety of techniques to make their images conform as closely as possible to the descriptions in the sacred texts. Based on the evidence cited above, it becomes possible to propose an explanation for the adoption of wood as the primary material of the Japanese sculptor’s craft at the start of the ninth century. The fashioning of Buddhist images out of
The Buddhist Transformation of Japan in the Ninth Century | 171
wood received an important boost from the establishment of an independent sculpture atelier at Tōshōdaiji in the 760s; it produced numerous single woodblock statues during the latter decades of the eighth century. This workshop became a source not only for technical expertise, but also for stylistic innovation. The volumetric bodies and repetitive drapery patterns that characterize much of the wood sculpture of the early Heian period have their source in the Buddhist imagery in China, the styles with which Jianzhen and the monks and artisans that accompanied him would have been most familiar and which they adopted for statues at the temple. But most important was the rapid rise of the ElevenHeaded Kannon cult in response to raging epidemics in the eighth and ninth centuries. Efficacious worship of the deity required images in unpainted wood, and wood was the only material available to patrons and artists away from the state-supported sculpture ateliers in the Nara and Heian capitals.
The Legacy of the Eleven-Headed Kannon Cult The strength of the Eleven-Headed Kannon cult during the latter part of the Heian period, especially in Nara and Ōmi, is documented by both images and texts. These sources reveal that the deity continued to be invoked in ways consonant with those of the previous two centuries, primarily to seek immanent benefits. One example is the statue commissioned by the Shingon monk Shōbō (832 – 909), who is best known as the founder of Daigoji. Although Shōbō first took the tonsure in 847 in Heian under the supervision of Shinga (801 – 879), a disciple of Kūkai, he soon thereafter traveled to Nara, where he resided at Tōdaiji for over twenty years and studied Sanron, Hossō, and Kegon teachings. One of the positions that Shōbō held after he left Nara and returned to the Heian capital was abbot of Gūfukuji (better known as Kawaharadera), a post that had been awarded to high-raking members of the Shingon sect since Kūkai first stayed there on his journeys to Mount Kōya.59 Throughout his lifetime Shōbō was particularly active as a sculptor and was responsible for the production of more than fifty works, one of which was a sixteen-foot (jōroku) danzō image of Eleven-Headed Kannon (not extant today) enshrined at Gūfukuji during the first decade of the tenth century. Although the specific circumstances surrounding the commission of the image are not described in detail in Shōbō’s biography, it is possible to conclude that even for one of the most influential Shingon monks of the day, artistic forms reflecting patterns of religious practices established in Nara during the eighth century still held currency. Two events in the late tenth century also attest to the continued strength of the Eleven-Headed Kannon cult even as Pure Land Buddhism had begun to capture the attention of the residents of the Heian capital. In 951, in response to the outbreak of an epidemic in the capital, the evangelical monk Kūya (903 – 972), known for his single-minded commitment to Amida, commissioned a statue of Eleven-Headed Kannon and a single set of the Daihannya kyō in six hundred fascicles for Saikōji, a temple at the eastern edge of the Heian capital.60 This statue, now housed at Rokuharamitsuji, is not a danzō, but Kūya’s choice of
172 | samuel c. morse
Figure 7.7. Eleven-Headed Kannon. Twelfth century. Wood with cut gold leaf. H. 39.2 cm. Kōgenji, Shiga Prefecture.
Eleven-Headed Kannon reveals that belief in the efficacy of the deity in combating pestilence had not waned. At the end of the century, Heisū (928 – ?), who studied both Kegon and Shingon teachings at Tōdaiji, founded a temple in Ujidawara to function as a private retreat.61 Known as Zenjōji, the sanctuary was halfway between Ishiyama and Kizu on one of the ancient routes linking Ōmi and Nara. While administrative duties kept Heisū at Tōdaiji, work on the main hall and the main object of worship, an eight-shaku image of Eleven-Headed Kannon, began in 991 and was completed in 995.62 Again, although no contemporary documents describe the rituals in which the image was used, a letter dating to 1151 detailing the founding of the temple states that Heisū performed rites of repentance there; this revelation is not surprising given the continued vitality of the Eleven-Headed Kannon cult at Tōdaiji.
The Buddhist Transformation of Japan in the Ninth Century | 173
The image at Zenjōji, standing in a stiff frontal pose and fashioned from a block of cherry for the front and hinoki for the back, preserves a number of formal characteristics of the sculpture of the previous century. These included the volumetric body, the alternating sharp and rounded drapery folds on the legs, and the whorl-like patterns of drapery between the legs and along the edges of the end of the skirt, which hangs down from the waist. In fact, the Zenjōji ElevenHeaded Kannon is just one of a number of archaicizing works produced at the behest of monks from Nara at the end of the tenth century and the beginning of the eleventh.63 The statue provides crucial evidence that not only were Nara monks perpetuating forms of worship established in the Heijō capital, but they were perpetuating early stylistic norms as well. As Pure Land beliefs gained hold over the religious lives of the Japanese during the eleventh and twelfth centuries, images of Eleven-Headed Kannon were sculpted with less frequency. Yet in areas where the cult dedicated to the deity had been long established, such as Ōmi and Echizen, images of the deity were still regularly produced. One work (figure 7.7) in the collection of Kōgenji, situated at the foot of Mount Kōtakami, is especially noteworthy. While the garments are decorated with kirikane (cut gold leaf), reflecting the lavish tastes of the period, the statue is carved from a single block of kaya (in the manner of the early Heian period), which has been stained to resemble sandalwood and measures just under forty centimeters in height, the size prescribed by the Eleven-Headed Kannon sūtras. Like the works from the tenth century mentioned above, this image reveals the tenacious hold the ElevenHeaded Kannon cult had across Japan throughout much of the Heian period. Few of the statues discussed here are well-known. Many have been ignored both by art historians and by historians of religion because they are not recorded in texts and are at temples isolated from the centers of power and religious authority. What textual evidence that does remain, such as the biography of Taichō, is frequently legendary. Yet as has been shown above, many of these sources provide a degree of access to the images whose early histories have been often obscured by the Shingon and Tendai traditions. A careful examination of ElevenHeaded Kannon worship shows that the spread of Buddhism to the provinces in the ninth and tenth centuries was directed not by monks from the newly established esoteric sects, but by monks with strong connections with Nara. Hossō monks in particular, who are better known for the study of philosophically complex doctrine, promoted modes of worship in both the capital and the provinces that responded directly to the immanent needs of many Japanese of the ninth century. The statues and the cult of Eleven-Headed Kannon force us to revise our understanding of the relationship between center and periphery in early Heian Japan. While the temples of the Shingon and Tendai sects and the sculptures and paintings housed in them will continue to be better known, it is the works of art in lesser known peripheral sanctuaries that more accurately reveal the nature of the religious lives of the Japanese of the early Heian period. Theirs are lives documented not by texts, but by images.
174 | samuel c. morse
Notes 1. For discussions of Buddhism in the ninth century, see Sonoda Kōyū, “Heian bukkyō no seiritsu”; and Hayami Tasuku, “Heian bukkyō ron.” Ryūichi Abé is one of the first scholars to acknowledge continuities between mikkyō practices in the eighth century and those in the early Heian period (Abé, The Weaving of Mantra, 151 – 184). 2. For a discussion of the position of Nara monks in the early Heian period, see Ueshima Toru, “Heian shoki bukkyō no saikenchō”; and Sone Masato, “Heian shoki Nanto bukkyō to gokoku taisei.” 3. Shoku Nihongi, Tenpyō 18 (746) 6/18. 4. Shiyi mian shenzhou xin jing, in Taishō shinshu daizōkyō (hereafter T) 1071, 152a – 154c; Ishida Mosaku, Shakyō yori mitaru Nara chō bukkyō no kenkyū, 84. 5. Shiyi mian guanyin shenzhou jing, T 1070:149a – 152a; Sawa Ryūken in “Jūichimen kannon no hyōgen ni tsuite,” 61; Ishida Mosaku, Shakyō yori mitaru Nara chō bukkyō no kenkyū, 84, 89. 6. Darani jijing, T 901, chap. 4, 812b – 825c. Nakamura believes the name should be Adhigupta (Nakamura Hajime, “Keka no seiritsu,” 31). 7. T 1802. Ishida Mosaku, Shakyō yori mitaru Nara chō bukkyō no kenkyū, 115. 8. Shiyi mian guanzizai pusa xin miyan yogui jing, T 1069, 139a – 146a. For information about the introduction of the sūtra in Japan, see Mikkyō daijiten, 850. 9. Shiyi mian guanyin shenzhou jing, T 1070:149a. 10. The list is from T 1071, 153c – 154c. 11. T 1071, 154c. 12. Shiyi mian guanyin shenzhou jing, T 1070, 150c – 151a. 13. Yoshida Yasuo, Nihon kodai bosatsu to minshū, 155 – 168; Abé, The Weaving of Mantra, 159 – 176. 14. The record of presentation for Hata no Kimi Toyotari dated Tenpyō 4 (732) includes the Jūichimen jō; those of Kamo Agatanushi Kurobito and Isonokamibe Oshiyama datable to Tenpyō 6 (734) include the Jūichimen kyō darani. Subsequent records indicate that study of dhāran. ī of Eleven-Headed Kannon continued to be an important part of the course of study. In total, the dhāran. ī of Eleven-Headed Kannon was studied on thirteen occasions. That of Thousand-Armed Kannon was studied on twenty-five occasions, and that of Fukūkenjaku Kannon on eleven occasions (Nara ibun, 2:508 – 510). 15. Nakamura, “Keka no seiritsu,” 25 – 34. 16. Yokomichi Mario, “Shunie no aramashi,” as quoted in Nakano Genzō, Rokudō e no kenkyū, 9. 17. Satō Michiko as quoted in Nakano, Rokudō e no kenkyū, 9 – 10. 18. Horiike Shunpō, “Nigatsudō shunie to Kannon shinkō,” 176 – 177. The date for the founding of the rite of repentance is from the “Tōdaiji gon bettō Jitchū nijū kyōkujō” detailing Jitchū’s career. In it he indicates that he has performed the rite of repentance for twenty-seven days beginning on the first day of the second month consecutively for seventy years between Tenpyō Shōhō 4 (752) and Daidō 4 (809). See Tōdaiji yōroku, 269. 19. Shoku Nihongi, Tenpyō Shōhō 3 (751) 10/20. 20. Shoku Nihongi, Tenpyō Shōhō 4 (752) 1/15. 21. The document is included in Horiike, “Nigatsu-dō shunie to Kannon shinkō,” 175. 22. Kyoko Nakamura, Miraculous Stories from the Japanese Buddhist Tradition, 175 – 176. 23. For the biography of Hōon and the history of Kojimadera, see Tsuji Hidenori, “Hōon hōshi gyōjō kō.” 24. Genkō shakusho, chap. 9, “Shaku Hōon,” 136 – 137.
The Buddhist Transformation of Japan in the Ninth Century | 175 25. “Kojima sanji konryū engi taishi den,” 103. The Daigoji bon was compiled in Jōgen 1 (1207). 26. The oldest extant version of Taichō’s biography, the Taichō kashō denki, is in the collection of Kanazawa bunkō and was transcribed in Shōchū 2 (1325) from a text first written down in Tentoku 1 (957). See Hakusan shiryō shū, ed. Anada Sanjrō et al., 1:219 – 225. Other versions of the biography are in the Ozoe Mitsutani family collection and at Heisenji and Ochi Jinja. For a discussion of the biography of Taichō, see Hongo Masatsuna, “Kodai Hokuriku no shūkyō bunka to kōryū,” 372 – 377. 27. Nishikawa Shinji believes that the statue can be stylistically related to works such as the seated Miroku at the Jison’in in Wakayama datable to Kanpyō 4 (892) and Eleven-Headed Kannon at Ryūgeji in Hiroshima. Thus, he dates the statue to the late ninth or early tenth century. See Nishikawa Shinji and Nagasaki Ichirō, “Echizen ni okeru kodai zōzō to sono haikei,” 923. Inoue Tadashi proposes a late eighth-century date (Kobutsu-chōkoku no ikonoroji, 189 – 194). 28. William Wayne Farris, Population, Disease, and Land in Early Japan, 645 – 900, 160 – 161. 29. For illustrations of the statue at Keisokuji, see Ishimoto Yasuhiro, Kōkoku no Jūichimen Kannon, 116 – 119; and 120 – 123 for the statue at Shakudōji. 30. In Shiga Prefecture these temples include Chūsenji, Sōshūji, and Ōsakidera in Makinochō; Ōtanidera, Yataka Gokokuji, and Jōheiji in Ibuki chō; Iwamadera and Hōzōji in Ōtsu city. In Kyoto Prefecture they include Kontaiji in Wasoku chō, Jindōji in Yamashiro chō, and Daidōji in Ujidawara chō. See Hongō Masatsugu, “Taichō to Hakusan shinkō,” 893 – 895. 31. For an illustration of this painting, see Nara kokuritsu hakubutsukan, ed., Tōdaiji Nigatsudō to omizutori, 23. 32. Gogota Hisanori, “Tadadera mokuzō Jūichimen Kannon bosatsu ryūzō ni tsuite,” 1 – 22. 33. For an illustration of this statue, see Mizuno Keizaburō, ed., Mikkyō jiin to butsuzō, pl. 65. 34. Asai Kazuharu, “Hagadera Jūichimen Kannon zō ni tsuite,” 31 – 33. 35. Nakamura, Miraculous Stories, 263. 36. Shoku Nihon kōki, Jōwa 4 (837) 2/2. 37. By Genkyō 5 (879) the main image of the kokubunji for the province of Sagami had been replaced with a statue of Yakushi. Other kokubunji with images of the Healing Buddha as objects of worship during the ninth century include those in Sado, Tanba, Tango, Awa, and Iyo. See Sandai jitsuroku, Genkyō 5 (879) 10/3; Nakano Genzō, Keka no geijutsu, 112 – 128; Nishio Masahito, Yakushi shinkō, 50 – 51. 38. Shoku Nihon kōki, Kashō 2 (849) 2/25. 39. Sandai jitsuroku, Jōgan 5 (863) 4 – 3. 40. Sandai jitsuroku, Jōgan 15 (873) 12 – 13. 41. Ekottarāgama Sūtra, chap. 28, T 125, 706a. See also Alexander C. Soper, “The Best Known Indian Images.” 42. James Legge, trans., A Record of Buddhist Kingdoms, 57. Although the earliest known attempt by the Chinese to obtain the first image of the Buddha for themselves did not take place until 502, sandalwood images were made in south China as early as the late 460s. Sambao kandong lu records that the sculptor Ho Jingshu of the Song dynasty (420 – 478) sculpted an image made of sendan during the years 465 – 471 (Mōri Hisashi, “Heian jidai no danzō ni tsuite,” 134). 43. Daiciensi sanzang fashi zhuan, T 2053, 252b – 253a; Si-yu-ki, Buddhist Records of the Western World, trans. Samuel Beal, 235.
176 | samuel c. morse 44. T 694. The text was translated in 691. See entry for “Daijō zōzō kudoku kyō” in Mochizuki Shinkō, ed., Bukkyō daijiten, 4:3277b. 45. Ishida Mosaku, Shakyō yori mitaru, 24, 34, 145. 46. Fusō ryakki, 39. 47. Tōnomine ryakki, 2:501. 48. For illustrations of the image, see Nara kokuritsu hakubutsukan, ed., Danzō, entry 29. 49. The passage in Hōryūji garan engi narabi rūki shizaichō reads, “One danzō brought over from Tang in Yōrō 3 (719)” (Nara ibun, 1:345). Some scholars believe that the statue was actually brought to Japan in the previous year by Dōji. See Kameda Tsutomu, “Kichijōten to jōdai no Konkōmyō kyō no bijutsu,” 48. For a discussion of the various theories about the relationship between the entry in the Shizaichō and the statue, see Uehara Shōichi, “Kannon bosatsu (Kumen Kannon),” 58 – 60. 50. For illustrations of this work, see Nara kokuritsu hakubutsukan, ed., Danzō, entry 30. 51. Amida kekaryō shizaichō, 671, 679; Nakano, Rokudō e, 14; Takei Akio, “Narachō no Amida keka,” 13. 52. Tōdaiwajō tōseiden, included in Tōshōdaiji, 1:91. 53. Nakamura, Miraculous Stories, 212. 54. The portable shrines are recorded as having been in the Ushirodō at Kōfukuji. See Kōfukuji rūki, 19. The Four Divine Kings were sculpted by the monk Kaimyō for the eastern precinct at Yakushiji (Shoji engishū [Daigoji bon], 116). The Thousand-Armed Kannon was at Katsuoji before 780 (Shaji engi shū [Gokokuji bon], 286). 55. Inoue Kazutoshi, “Danzō kō,” 16 – 17, quoting the Shiyi mian shenzhou xin jing ishu. 56. Inoue Tadashi, Danzō, 26 – 27. 57. Although it first gained some attention in 1909 (“Hara ke zō no Jūichimen Kannon zō,” Kokka, no. 224 [January 1909]) and has often been included in exhibitions since the early 1970s, this statue has until recently received little scholarly attention. Kurata Bunsaku, who first recognized the importance of the work, believed it to date from the beginning of the ninth century; however, he harbored some doubts about both the date and the statue’s provenance. For discussions of the statue, see Tokyo kokuritsu hakubutsukan, ed., Heian jidai no chōkoku, entry no. 12; Nara kokuritsu hakubutsukan, ed., Kannon bosatsu, 44 – 45; Nara kokuritsu hakubutsukan, ed., Danzō, entry 31; Nara kokuritsu hakubutsukan, ed., Tenpyo, entry 78. 58. For illustrations of this image see Nara kokuritsu hakubutsukan, ed., Danzō, entry 36. 59. Saeki Arikiyo, Shōbō, 118 – 126. Shōbō was appointed to the post in 879 and again in 883, resigning the position in 891. The date for the statue comes from Inoue, Danzō, 38. 60. Details of the founding of the temple and the sculpting of the image can be found in the Kōya rui, a biography of the monk compiled in 973, and the Rokuharamitsuji engi of 1122. See Itō Shirō, “Rokuharamitsuji Jūichimen Kannon zō,” 91 – 95; for an illustration of the statue see color plate 7. 61. Information about Heisū and Zenjōji can be found in Mizuno Keizaburō, “Zenjōji no chōkoku to sono shūhen,” 168 – 174. 62. For an illustration of this statue see ibid., pl. 8. 63. See Mizuno Keizaburō, “Hosei ikō ni tsuite,” and idem, “Zenjōji,” 180 – 184.
part III
Establishing New Religious Spheres
8
d Ryūichi Abé
Scholasticism, Exegesis, and Ritual Practice On Renovation in the History of Buddhist Writing in the Early Heian Period
T
here was an epistemic shift in the production of Japanese Buddhist texts in the early Heian period, a shift that enabled Buddhists to incorporate the elements of meditation, ritual, and religious practice in general within the science of scriptural exegesis. Until the early ninth century, the exegetic texts written by Japanese Buddhist scholars were concerned entirely with doctrinal issues. By contrast, by the mid-tenth century, the great majority of Buddhist commentarial texts had their focus on ritual practices, especially on the rituals of esoteric Buddhism, the ritual practices that became integral within the management of the Heian court and the courtiers’ lives. To appraise the significance of this change, this essay will examine Buddhist writing against the broader background of early Heian textual production and then compare as case studies two commentarial texts on the Prajñā-pāramitā Heart Sūtra, which is popularly known in Japan as Hannya shingyō, or simply, Shingyō. One of the commentarial texts was written by Sanron master Chikō (709 – 781), the other by Shingon master Kūkai (774 – 835). By the late Nara period Chikō’s text established itself as a classic, setting the standard for the early Heian Buddhist scholarpriests to exercise their exegeses. It represented the exemplary mode of textual production under a regime whose ruling ideology was dominantly Confucian. In contrast, Kūkai’s text can be understood as a challenge to the established method of interpretation, a challenge aimed at creating a distinctly Buddhist method for interpreting Buddhist texts. By comparing these two texts, I hope to illustrate a seminal change in the history of Buddhist textual production in which ritual language, the element that had been peripheral in Buddhist intellectual discourse until the early Heian period, began to assume the central role not only in shaping Buddhist thinkers’ ideas but in engendering texture in their writings. It was such a change that in turn enabled the Buddhists of the mid- and late Heian periods to make their religious practices integral with and even pivotal for the management of the court and the state.
180 | ryŪichi abé
Authority of Kangaku: The Beginning of the End of the Ritsuryō Age The early Heian period was marked by a great surge of scholastic activities grounded in kangaku, the study of Chinese texts (kanseki). Confucian studies, especially the study of its ideology of kingly rule, was avidly promoted by Emperor Kanmu and the succession of the early Heian emperors.1 Kanmu’s adoption in 784 of the Gongyang and Guliang Commentaries on the Spring and Autumn Annals as official texts for statecraft signaled the emperor’s effort to centralize the state, foreshadowing the move of court and capital from Nara to Kyoto in 794.2 In his 806 decree, Emperor Heizei, who succeeded Kanmu, made it mandatory that all imperial princes and sons of the aristocrats of the fifth rank or above enter the State Academy (Daigakuryō) and study Confucian classics there.3 In addition to the discipline of Confucianism (meikyōdō), its central curriculum, the State Academy was important in promoting the study of law, history, and writing. Sons of prominent aristocratic families entered the State Academy to learn the basics of Confucianism, then moved to kidendō, the curriculum on history and literature, in which they learned matters immediately relevant to administrative procedures of the court.4 Others moved on to the discipline of law (meihōdō). Legal experts of the court worked hard to improve the implementation of the ritsuryō rules, the body of laws reflecting Confucian political and social ideals, adapted originally from the Tang penal and administrative codes. The compilation of comprehensive collections of kyaku — amendments to the ritsuryō rules — and shiki — its bylaws — began as a state project under Heizei’s rule and was completed during the reign of Emperor Saga (r. 809 – 823). The resulting Kōnin kyaku and Kōnin shiki provide coverage of all the amendments and bylaws issued between 701 and 819.5 In 833, a team of legal scholars led by Minister of the Right (udaijin) Kiyohara no Natsuno (782 – 837) concluded the compilation of the Ryō no gige, the official exegesis of the Ritsuryō.6 Near the end of the ninth century, Professor of Law (meihō hakase) Koremune Naomoto (fl. 889 – 898, ? – 907) compiled Ryō no shūge in fifty volumes, another exhaustive commentary on the Ritsuryō, in which he collected a range of works produced privately by legal experts and scholarly houses.7 Composing historical texts was another important area of academic production for early Heian aristocratic scholars. Five out of the six official volumes of imperial history, the Rikkoku shi, were composed during this period. Using the style of Chinese dynastic histories as their model, the Shoku Nihongi and the Nihon kōki, completed in 797 and 840, respectively, were intended as sequels to the Nihon shoki, the first imperial history compiled in 720. In contrast, the 869 Shoku Nihon kōki, the 878 Montoku tennō jitsuroku, and the 901 Nihon sandai jitsuroku follow the convention of qijuzhu (kikyochū) — collected sayings of the emperor compiled, edited, and annotated by his personal attendants — typified in the Latter Han History in their manner of recording the deeds of emperors and their daily court lives. Despite their difference, all six histories share a common desire to portray rulers as exemplars of Confucian moral virtue.8
Scholasticism, Exegesis, and Ritual Practice | 181
According to the Nihon sandai jitsuroku, the last among the six books of the imperial history, a fire broke out in the Daigokuden, the central audience hall in the Imperial Palace compound in the evening of the tenth day of the fourth month of Jōgan 18 (876). The fire swiftly reduced the Daigokuden and its adjacent structures to ashes. On the next day Emperor Seiwa (r. 858 – 876) asked his courtiers whether he should conduct the affairs of state and whether the emperor’s ministers should attend to the business of the court as usual. A group of scholars led by Yoshibuchi no Nagasada (n.d.), an expert on Confucian classics, basing their opinions on the Book of Rites and Zuo’s Commentary on the Spring and Autumn Annals, recommended that both the emperor and his court lament the loss of the hall for three days and abstain from their official duties during that period. But Kose no Fumio (fl. 878), Miyako no Yoshika (834 – 879), and other scholars of historical learning cited precedents in the Guliang Commentary, the Han History, the Latter Han History, the Record of Wei, the Book of Liao, and other Chinese sources to suggest that the emperor and ministers should observe rites lamenting the loss for three days but should not otherwise interrupt the conduct of court business. The emperor adopted the latter recommendation and immediately resumed his regular duties.9 As this example illustrates, it was customary for the early Heian court to rely heavily on authoritative Chinese texts on rituals, laws, and history in order to make important political decisions. The Zoku kojidan, a twelfth-century compendium of episodes surrounding the Heian imperial court and its aristocrats, provides the following information regarding the learning of the Heian courtiers: To be chosen to record the proceedings of the imperial cabinet meeting and compose its edicts was a matter of great honor and significance. Only those holding the dual appointment of special minister (sangi) and secretary-general (daiben) in the Council of State (Daijōkan) were entitled to perform such a task. The cabinet was the battleground for elite aristocrats to boast their great learning. They vied with one another by displaying their ability to refer to diverse books of Chinese classics. Quoting these texts freely from memory, they advanced their arguments and recommendations for the cabinet’s adoption. Without asking each cabinet member about their sources, without having any text in his reach, the secretary was expected to write down all the arguments and recommendations as immediately as he heard them.10 This episode portrays the particular affinity that developed between literary studies and statecraft in the early Heian court and its intelligentsia. In order to receive appointments to high offices and serve effectively as statesmen at court, aristocrats needed to be well versed in Confucianism, law, history, and a gamut of exegetic literature written on the principal texts in these fields. Furthermore, they were required to have sophisticated skills in the art of writing. This requirement explains why the early Heian period marked the apex in the development of kanshi, poetry in Chinese.11 Emperor Saga appointed Special Minister Ono no Minemori (778 – 830) to compile the first imperial anthology of Chinese poems, which was completed in 815 as the Ryōunshū. In its introduction,
182 | ryŪichi abé
Minemori states, “Writing is the great work of managing the state, the ever-thriving enterprise.”12 Two additional imperial kanshi anthologies followed: the Bunka shūreishū and the 827 Keikokushū. One of the chief editors of the Keikokushū, Councillor-general (dainagon) Yoshimine no Yasuyo (785 – 830) writes, “The task of writing is to unveil the meaning of all that transpires between heaven and earth, to distinguish the rank among people, and to understand the nature and principles underlying all things in the world.”13 These anthologies contain poems written by imperial princes, ministers, and aristocratic court officials, many of whom were scholars responsible for the production of Confucian exegeses, legal commentaries, and historical records. Their poems were typically understood as ōseishi (C. yingzhishi), poems responding to the ruler’s invitation in support of the emperor’s virtuous reign. Thus, studying poetry comprised an essential part of refining literary skills — especially in the disciplines of rhetoric and poetics — which in turn empowered nobles to serve the court most effectively.14 For example, Secretary-general Miyako no Yoshika, one of the principal editors of the Montoku tennō jitsuroku, collected his own verses and poems (titled Toshi monjū) into six fascicles intended as texts for younger generations of scholars to learn grammar, syntax, and literary devices necessary to write well in classical Chinese. Minister of Treasury (ōkurakyō) Shigeno no Sadanushi (785 – 852), one of the chief compilers of the imperial poetic collection Bunka shūreishū, composed the Hifuryaku, a massive Chinese dictionary, in one thousand fascicles, in which he painstakingly identified the usage of each word or character in more than fifteen hundred primary texts of diverse genres.15 The intertwining of the arts of writing and statecraft, as suggested in these examples, resulted largely from the Confucian utilitarian bent in the early Heian intelligentsia’s attitude toward poetry, textual studies, and language. One of Confucius’ sayings in chapter 17 of the Analects epitomizes their philosophy: My young friends, why do you not study the Book of Poetry? Poems stimulate your emotions, broaden your observation, expand your fellowship, and express your sorrows. They help you in your immediate service to your parents and in your more remote service to the rulers. They widen your acquaintance with names of birds, animals, and plants.16 It is this Confucian pragmatic approach to language that allowed the production of Chinese texts in law, philosophy, history, and poetry to flourish. However, this constellation revolving around the authority of the Chinese texts and Confucian studies began to change, rather drastically, from the mid-Heian period. The superabundant production of kyaku saturated the legal procedure with conflicting principles, rules, and guidelines and accentuated ironically the inapplicability of the ancient system of ritsuryō to the historical reality of Heian society. The formulation of additional sets of shiki continued as well. However, by the mid-Heian period, emphasis shifted to detailed descriptions of the swiftly growing bodies of rites, ceremonies, customs, and rules of conduct in the Heian court and its aristocratic society.17 The compilation of imperial histories ceased with the 901 Nihon sandai jitsu
Scholasticism, Exegesis, and Ritual Practice | 183
roku. The courtiers’ interest moved away from depicting emperors as idealistic virtuous rulers to the composition of mundane, realistic records of administrative procedures — as exemplified by the Seiji yōryaku, an outline of administrative procedures in 130 volumes composed by Koremune Kotosuke (n.d.), who served in the court of Emperor Ichijō (r. 986 – 1011).18 With the disappearance of imperial collections of kanshi poems, the aristocrats’ fervor for composing such poems also declined.19 The compilation of Kokin wakashū, which was ordered in 905 and completed in 913 or 914, marked a new tradition of waka anthologies, poems in Japanese, compiled under the imperial aegis. In contrast to the imperial kanshi collections — whose poems articulated and constituted praise of the ruler — the depiction of natural beauty, seasonal change, love, joy, and sorrow comprised the primary subjects for the poems in the Kokin wakashū. At the same time, there was a significant growth of monogatari (fictional narrative literature written in Japanese) such as Murasaki Shikibu’s The Tale of Genji. Precisely because of its fictionality, which was held to be devoid of utility for affairs of state, this genre of writing was conspicuously absent from the offerings of the early Heian court authors.20 The birth of these new categories of writing suggests that a break between the art of writing and the art of statecraft developed in the mid-Heian period. In contradistinction to the words of Confucius quoted earlier, Yoshida Kenkō (1283 – 1352?) would eventually observe, “To boast talent in diverse arts brings shame to the gentleman. True, both rulers and ministers cherish the subtle ways of attaining skills in poetry, songs, and musical instruments. However, it is utterly foolish in our age for a ruler to think of governing his realms through the mastery of these arts.”21 Thus, intellectuals of the thirteenth century had thoroughly embraced this separation, developed first in the tenth century. While the Heian courts continued its support for Confucian studies, Confucianism lost its authority and centrality for the management of the state.22 With the development of multiple court rituals integrating strongly esoteric Buddhist elements, Buddhism rose as the dominant ideology of the state in the mid- and late Heian periods. Together with other genres of writings in Chinese, the production of Buddhist texts grew abundant during the early Heian period. As Inoue Mitsusada pointed out in his classic study, the academic activities of the Six Nara schools (Nanto rokushū) reached their summit in the late eighth and early ninth centuries. According to Inoue, few exegetic and doctrinal texts were produced by Nara scholar-priests during the early and mid-Nara periods, while its production multiplied exponentially at the end of the Nara period and in the early Heian period.23 To provide an institutional foundation to their academic activities, the celebrated annual lecture-conferences at the Nara temples — such as the Yuima’e at Kōfukuji (802), the Saishōō’e at Yakushiji (830), and the Hokke’e at Daianji (832)24 — were established under the aegis of the early Heian emperors. It was these major temples of Nara — rather than the new temples of the Shingon and Tendai schools in Kyoto — that continued to receive heavy patronage from the court and performed the rites and ceremonies for the state. Thus, contrary to
184 | ryŪichi abé
prevailing views, Nara temples and their six schools formed the nucleus of the early Heian Buddhist establishment whose institutional might and sphere of influence dwarfed the incipient Shingon and Tendai schools. The Buddhist clergy functioned in short as a secondary bureaucracy complementing the aristocratic officials of the imperial court.25 A case in point is the Sōgō (Office of Monastic Affairs), a governmental office in the Genbaryō (Agency of Foreign Affairs), which in turn belonged to the Jibushō (Ministry of the Aristocracy). The Sōgō officials are unique because they were chosen from among the most eminent priests at the Nara temples and, at the same time, composed a formal part of the imperial court bureaucracy. Their primary duty was to maintain order in the clergy and make the Buddhist community prosper in accord with the principles set in the Ritsuryō, especially the Rules for Priests and Nuns (Sōniryō) in volume 7. Article 14 describes the qualifications necessary for priests seeking appointment to the Sōgō: “Article 14. Those who are to be appointed to the office of Sōgō must be models of virtuous conduct, exhibit strong leadership, and be worthy of reverence of both ordained and laypeople. Above all, they must be skilled in supervising matters relating to the Dharma.”26 In his Ryō no shūge commentary, Koremune Naomoto provides the following elucidations: “Virtuous” (toku) and “conduct” (gyō) refer, respectively, to inner and outer qualities. “Virtuous conduct” is applicable to those who, having filled their minds with virtue, generously give others in their religious practice. According to Kong Anguo’s Annotated Book of History, virtue is grounded in respectful deeds; thus we find the expression worthy of reverence. Based on Anguo’s text, the Old Commentary (Koki) says, “Virtuous conduct means that both speech and deeds of sagacious priests establish the norm for others.” The Administrative Commentary (Ryōshaku) says, “For interpreting virtuous conduct see the Rules of Appointment and Promotion (Senjoryō), in volume 12 of the Ritsuryō.” Anata’s Commentary (Anaki) says, “Virtuous conduct means the power to win the respect and reverence of others, both clergy and lay.” According to Ato’s Commentary (Atoki), “Supervising the matters related to the Dharma means the work of spreading a net of control over the clergy to protect the Buddhist teaching.” The Red Ink Commentary (Shuki) says, “Virtuous conduct here must be understood as exactly the same thing for the lay. It applies to both clergy and laity. Although a priest of virtuous conduct resides only in one temple, he gains the respect and reverence of priests in all other temples.27 Naomoto’s legal exegesis reflects the importance of kunko or kunkogaku (C. xun guxue) — literally, the “discipline for interpreting ancient terms” — the method of interpretation stereotypical for Chinese studies and the essential mode of learning for the early Heian aristocracy. In China, kunkogaku developed along with the study of Confucian texts, as Confucianism served as the orthodoxy for the state in the Han, Sui, and Tang dynasties.28 According to Ikeda Genta, kunkogaku, as
Scholasticism, Exegesis, and Ritual Practice | 185
adopted by Nara and early Heian intelligentsia, demonstrates the following three traits.29 First, it divides the passages in the original text into progressively smaller semantic units — phrases, words, letters, and even strokes within a single character. This operation is necessary to determine syntactical breaks in the original passages, which are devoid of punctuation signs indicating the end of a phrase, sentence, or paragraph. Second, the proper usages of divided phrases, words, or characters are identified in texts of established authority — typically, exegetic texts from ancient or earlier periods. Third, an effort is made to quote from as many of these authoritative texts as possible, so that readers are able to acquire encyclopedic knowledge of primary texts in which there appears a particular character, word, or expression to be elucidated; readers also acquire a clearer understanding of that character, word, or expression itself. In the above example, the term in question is tokugyō, “virtuous conduct,” which is a Buddhist term indicating the mastery of the three studies (sangaku) — the monastic law, the science of meditation, and the cultivation of wisdom — and the six bodhisattva practices (rokudo) grounded in the acts of giving, upholding precepts, patience, courageous effort, practicing meditation, and advancing wisdom. Neither the Ryō no shūge nor any of the exegetic texts it cites makes any mention of these Buddhist connotations, however. Naomoto’s exegesis strives to interpret tokugyō in the context of Confucian studies. For this goal, it first cites the Book of History, one of the five Confucian classics, in the version annotated by Kong Anguo, the celebrated early Han Confucian scholar and the eleventhgeneration descendant of Confucius. Kong Anguo’s edition is one of the official texts adopted at the State Academy and is frequently referred to in the main text of the Ritsuryō. Setting the tone with the Book of History, Naomoto liberally quotes from the earlier commentaries on the Ritsuryō, all of which reinforce his Confucian redefinition of the Buddhist term.30 At the heart of kunkogaku is the Confucian principle of the rectification of names (J. seimei; C. zhengming). According to this idea, the correct meaning and appropriate usage of every word were established by ancient sages and encapsulated in the classics. That is, a society is able to sustain or restore its order by “rectifying” the use of words and by eliminating disruptions from the just correspondence between words and things.31 In short, with the application of kunkokagu, the ritsuryō legal system provided the political and institutional framework for early Heian intellectuals, including Buddhist clergy, to carry out their academic activities in accordance with the Confucian theory of language and society. Under such a political, social, and cultural climate, how did Buddhists actually study their scriptures and engage in their own production of exegetic texts? In the following sections I strive, first, to study the typical mode of production of early Heian Buddhist exegetic texts, then to analyze the significance of the new methods of interpretation introduced through esoteric Buddhist texts, and finally, to assess the repercussions of such a change in the general history of writing in Heian society.
186 | ryŪichi abé
Chikō and the Marginality of Buddhist Texts Throughout the Nara and early Heian periods, the Prajñā-pāramitā Heart Sūtra was one of the most popular Buddhist scriptures for both the ordained and lay, for their merit-producing acts of chanting and copying. In 774, for example, when the nation was threatened by an epidemic, Emperor Kōnin (r. 770 – 781) commanded the nation to recite the sūtra: It is said that the Prajñā-pāramitā is the mother of all the Buddhas. When I, the Son of Heaven, recite it, the nation is safe from invasions and rebellions; when my people invoke it, their households are protected from the demons of illness. Let us rely on its compassionate power to save us from our present misfortune. I therefore encourage all those in every province under heaven, both men and women, both young and old, constantly to recite the sūtra. Those of you who serve my court, in both civilian and military ranks, recite the sūtra on your way to work and at any interval between your duties.32 Among exegeses on the Heart Sūtra written by Japanese scholar-priests, the Maka hannya haramita shingyō jutsugi by Sanron master Chikō (709 – 781) of Gangōji was regarded as the most authoritative commentary in the early Heian period.33 Chikō was a celebrated Buddhist scholar of the mid-Nara period, renowned as well for his practice of Pure Land Buddhism, especially for his commissioning of the Pure Land mandala at Gangyōji.34 Keikai of Yakushiji, in his tale the Nihon ryōiki, composed about 787, describes Chikō as follows: “He [Chikō] was gifted with heavenly intelligence and was unexcelled in his fame as a great pundit. He composed commentaries on the Ullambana, the Greater Prajñāpāramitā, the Prajñā-pāramitā Heart, and many other sūtras and gave lectures on Buddhist doctrine that attracted multitudes of students.”35 Modern scholars have identified fourteen titles in fifty fascicles as Chikō’s authentic writings, a number that places him as one of the most prolific authors among the Nara and early Heian Buddhist scholars.36 Only two of his compositions survive as complete works: the Shingyō jutsugi, which is discussed here in detail, and commentary on the Vimalakīrti Sūtra (Yuimakyō). But a number of relatively large sections and fragments from his other works were preserved as quotations in the writings of early Heian scholars. For example, quoting extensively from Chikō’s compositions, Sanron master Anchō (763 – 814) of Daianji composed the Chūron shoki,37 a monumental work illustrating Nāgārjuna’s nondualist philosophy as articulated in his Mādhyamikakārikā. In 830, in response to Emperor Junna’s request, Master Gen’ei (? – 840) of Saidaiji composed a treatise capturing the gist of the Sanron school’s philosophy,38 relying heavily on Chikō’s scholarship.39 Instead of authoring their own commentaries on the Heart Sūtra, Anchō, Gen’ei, and other Sanron scholars of the early Heian period used Chikō’s work as a guide to interpret it. One of Chikō’s aims was to deliver a critique on the Hossō (Yogācāra) school’s interpretation of the Heart Sūtra, which since the appearance of the celebrated Chinese Yogācāra master Xuanzang’s translation had become
Scholasticism, Exegesis, and Ritual Practice | 187
the standard in both Chinese and Japanese Buddhist monastic circles. In fact, Chikō’s interpretation was so influential that it remained uncontested by priests of the Hossō school until the late tenth century, despite that school’s dominance in the Buddhist academia.40 In the introduction to the Shingyō jutsugi Chikō describes the background of his composition: At age nine, I left the world. Since then I lived in the monasteries and devoted myself to study. For over thirty years and now in the fourth year of Tenpyō Hōshō [752], I have resided in quiet places, concentrated my mind, and disciplined myself in prostration and recitation to study widely the scriptures. Among all that I studied, this sūtra is sublime. The sūtra’s lines are terse, yet it is rich in meaning. It exemplifies laconism; yet its discussion is refined. It contains the profoundest essence of all the sūtras. Therefore, people always cherish it, chant it, revere it, and uphold it. No other sūtras enjoy or exceed such esteem and devotion.41 The main text of the Shingyō jutsugi comprises four parts: (1) a discussion of the sūtra’s title; (2) an explanation of the goal of the sūtra’s argumentation; (3) a list of three different translations in the Chinese canon; and (4) a phrase-by-phrase interpretation of the sūtra passages.42 In the first section, Chikō interprets the term “heart” (Ch. xin; J. shin; Skt. hrdaya) in the title as the essence (chūjitsu) of ˙ the entire prajñā-pāramitā literature. The word “heart” refers to the most essential substance (kenjitsu saiyō). With their texts extremely lengthy and their meaning multifarious, other prajñāpāramitā sūtras often discourage students from memorizing and studying them. To facilitate the mastery of the prajñā-pāramitā for sentient beings, those who compiled the treasury of Dharma collected the essentials of the Buddha’s words to manifest the gist of all the prajñā-pāramitā sūtras, the essentials that make up this sūtra.43 This understanding of the Heart Sūtra as the condensation of other, more voluminous prajñā-pāramitā sūtras appears not to be Chikō’s original idea but an adaptation of the opinions expressed in commentaries by many influential figures in the Chinese doctrinal schools, including Fazang (643 – 712) of the Huayan (Kegon) school and Kuiji (632 – 682) and Yuance (613 – 696) of the Fa xiang school.44 Chikō’s wide-ranging knowledge of the theoretical literature of Chinese Buddhism is demonstrated in the second part of the Jutsugi. Chikō here identifies the goal of the sūtra as that of illustrating the transcendental wisdom (hannya) of unattainability (mushotoku), which is the capacity of the transcendental wisdom, or prajñā, that makes it possible to escape the conceptualization and verbalization that, together, construct external objects. Chikō then divides this goal into two aspects, the primary motif (shō) and secondary motif (bō) of the sūtra text: these are, respectively, prajñā as an enlightening perspective that sees all things as nonarising (hannya mushō shokan) and prajñā as the resultant, perceived
188 | ryŪichi abé
middle way of perfect nondiscrimination (shokan musō chūdō).45 Chikō’s theoretical approach here faithfully follows the example set in the Subtle Meaning of Mādhyamika,46 a major treatise by the Sanlun patriarch Jizang (549 – 623). He also quotes frequently and liberally from Jizang’s other works47 and thus demonstrates the continuity of the Sanlun/Sanron transmission across China and Japan. Yet Chikō never confines himself to the doctrinal developments within the Sanron school. To elucidate the goal of the sūtra, he relies heavily on the theory of the fifty-two progressive stages of the bodhisattva practice, a theory that holds a central place in the Huayan school. In particular, Fazang and his friend Wonhyo of Silla (b. 617) provide inspiration for Chikō. Basing himself on Fazang, Chikō explains the simultaneous identity of and separation between the fifty-two stages as the manifestation of the two aspects of the prajñā: its substance (tai), which, in transcending causality, illuminates the practitioners’ mind in all the fifty-two stages; and its function (yu), which serves as a driving force behind the practition er’s ascent through those stages.48 Chikō’s defense of the Sanron school against criticisms from the Hossō school comprises another major issue in his commentary. In the third part of the Shingyō jutsugi, he briefly describes the three versions of the Chinese translation of the sūtra, first by Kumārajiva (344 – 413), next by Xuanzang (ca. 602 – 664), and most recently by Bodhiruci (572 – 727), identifying the dates and places of each translation.49 The translation by Xuanzang circulated most widely throughout East Asia. Chikō acknowledges that Xuanzang’s translation of the Heart and other major prajñā-pāramitā sūtras encouraged Xuanzang’s Yogācāra-school disciples to produce the most dominant interpretations of the prajñā-pāramitā texts in China. In the Shingyō jutsugi, Chikō refers occasionally to commentaries composed by Xuanzang’s two most famous disciples, Kuiji and Yuance.50 He seems to indicate that even by basing himself on Xuanzang’s translation and by referring to the Chinese Yogācāra commentaries, it is possible for him to reclaim the Heart Sūtra for the Sanron school. He illustrates this point in the following dialogue: Question: Does this sūtra [the Heart Sūtra] belong to those of definitive meaning or those of disputable meaning? Answer: This is a sūtra of truly definitive meaning. For the audience of most advanced capacity, the Buddha taught [in the sūtra] the Dharma of nondiscrimination, in which the mind and its objects are both empty, perfectly interfusing. He also announced the attainment of buddhahood by all living beings based on the permanence of the Buddha nature. Kuiji, Yuance, and others [of the Chinese Yogācāra schools] designated the teachings of prajñā-pāramitā as of disputable meaning. This is ignorant confusion in the extreme. Just like one who is blind running around madly in the darkness vainly attempting to find his way, not only have they not thoroughly studied the Prajñā school, but they have not mastered the discussions of their own school.51
Scholasticism, Exegesis, and Ritual Practice | 189
Generally, in both the Abhi Dharma and Yogācāra literature, the distinction between definitive meaning (J. ryōgi; S. nītartha) and disputable meaning (J. furyōgi; S. neyārtha) is understood as a contrast between the literal and interpretive (which, in turn, is understood generally in modern scholarship as that between the explicit and implicit). The essential scriptures of the Yogācāra school, . such as the Samdhinirmocana (Gejinmikkyō), were written in the language of the definitive, while all the prajñā-pāramitā scriptures, which constitute the textual foundation of the Mādhyamika/Sanlun/Sanron school, are rendered in the less refined language of the disputable.52 In Tang China, Xuanzang’s comprehensive new translation of the gamut of Yogācāra scriptural texts unmistakably established the superiority of the Yogācāra school over the Sanlun school, which was already in decline and unable to counteract the dominant Yogācāra position. In contrast, for Chikō and his fellow Buddhist scholars in mid-Nara Japan, when the Sanron and Hossō schools continued to vie with one another for intellectual supremacy, the debate over definitive meaning and disputable meaning was an ongoing concern.53 Chikō asserts that the Heart and other prajñā-pāramitā sūtras are in fact written in the language of the definitive. However, Yogācāra scholars failed to understand this because they were unable to distinguish between the primary motif (the instantaneity of attaining transcendental wisdom) and the secondary motif (the sequential growth of the nondiscriminatory knowledge of emptiness) in the prajñā-pāramitā literature. In the Yogācāra texts, the order of the primary and secondary motifs is reversed because of their emphasis on gradualism, the diachronic process of the bodhisattvas’ spiritual advancement. This is why, Chikō argues, the language of the prajñā-pāramitā texts tends to be obscure and secretive, and that of the Yogācāra scriptures apparent and pronounced. In short, the difference between the Mādhyamika and Yogācāra texts lies in their phraseology but not in their meaning — meaning as the crucial element that, for Chikō, ultimately determines scriptural texts’ definitiveness, finality, and authority. They are equally advanced as the representatives of Māhāyana philosophy, but in their phraseology, Mādhyamika texts show greater sophistication.54 Chikō points out that this is why the Greater Prajñā-pāramitā Sūtra is in fact providing grounds for an array of quintessential Yogācāra treatises, . such as Madhyāntavibhāgatikā (Chūhen hanbetsuron) and Mahāyānasamgraha (Shōdaijōron).55 In the fourth, final, and lengthiest part of the Shingyō jutsugi, consisting of the phrase-by-phrase analysis of the Heart Sūtra’s text, Chikō first divides the sūtra into two major sections: “the first is the sūtra’s discussion proper, which explicates the correct understanding of the Dharma for practitioners to cultivate it and realize it; the second is the description of the sūtra’s dhāran. ī,56 which is to protect the practitioners and enable them to preserve the sūtra.” Chikō then divides these two parts into the following subsections.
190 | ryŪichi abé
I. Discourse on the correct understanding of the Dharma for the practitioners’ Cultivation and Realization A. Discourse addressed to Avalokiteśvara 1. Presentation of the substance of correct understanding 2. Analysis of the enlightened realm 3. Demonstration of the benefit given to sentient beings B. Discourse addressed to Śāriputra 1. Elucidation of the principle for understanding (a) Outline (b) Extensive discussion 2. Elucidation of the wisdom that makes understanding possible (a) Illustration of the wisdom of bodhisattvas (b) Demonstration of the wisdom of the Buddha II. Discourse on the dhāran. ī and its power of protection A. Praise of its excellent merit B. Illustration of the dhāran. ī’s substance In the first section, Chikō delivers a detailed analysis for each of the sūtra’s passages by quoting liberally from essential Mādhyamika sūtras and treatises, including the Greater Prajñā-pāramitā, Madhyāmika-Śāstra, and Discourse on the Greater Prajñā-pāramitā (Ch. Dazhidoulun; J. Daichidoron). However, a careful reading of his text shows that Chikō occasionally borrows arguments from Kuiji’s and Yuance’s commentaries on the sūtra. That is, Chikō shows his sophisticated skill in advancing the argument by incorporating the discussions of his opponents. In the second section of the fourth part of Shingyō jutsugi, which corresponds to the Heart Sūtra’s brief conclusion and closing dhāran.ī, Chikō divides the sūtra passages into two subsections: the praise of the efficaciousness of the dhāran.ī and the dhāran.ī itself. Maintaining his style of phrase-by-phrase analysis, Chikō presents his explication as follows: Therefore, one understands the prajñā-pāramitā as the great, divine spell, With the right understanding of the unattainability and nonduality of the prajñā, the spell conquers heretics by crushing their obstructing belief in the self. Therefore, the spell is called great, divine. Being lofty and victorious in various ways is called great. Being endowed with limitlessly marvelous working is called divine. What is called a spell is a magical technique. All the gods and sages have revered and upheld the dhāran. ī, the secret of the Bhagavat, the essence of the true Dharma, which destroys evil and manifests good. Its power is unequaled. It is also said in Jing-men’s commentary, “The meaning of the spell is to annihilate evil and generate good.” the spell of the great light of wisdom, With the deep prajñā of unattainability, the spell even lifts the thick karmic darkness of the lowliest beings, the icchantikas. That is why the dhāran. ī is called the great light of wisdom, since there is no darkness that it cannot illumine.
Scholasticism, Exegesis, and Ritual Practice | 191
the unexcelled spell, By means of the true understanding of the prajñā-pāramitā, the dhāran. ī causes the followers of the two Hīnayāna vehicles to realize that they have not reached the ultimate goal — to abandon their fear of the suffering of the world and to give rise to the compassionate mind. Therefore, it is called unexcelled. and the unparalleled spell. Through the deep prajñā, the dhāran. ī encourages bodhisattvas to work to benefit both themselves and others and enables them to bring all their practices to perfection. It is thus called unparalleled.57 Chikō here interprets the four phrases of the sūtra eulogizing the efficaciousness of the sūtra’s concluding dhāran. ī — the great divine spell, the spell of the great light of wisdom, the unexcelled spell, and the unparalleled spell — as corresponding to the prajñā-pāramitā’s guide to salvation for the four classes of beings: the heretic, the icchantikas, the Hīnayānists, and the Mahāyānists. The inclusion of lowliest beings, icchantikas (issendai), can be seen as Chikō’s challenge to Yogācāra, whose theory permanently bars them from attaining enlightenment. However, in contrast to the foregoing parts of the Outline in which Chikō defends the prajñā-pāramitā scriptures against Yogācāra’s criticism, Chikō’s attitude toward Yogācāra in the final section has a conciliatory side. To validate his interpretation of the dhāran. ī in the section quoted above, Chikō openly quotes from another Yogācāra commentary on the Heart Sūtra by Jingmen (fl. 627 – 649),58 a student of Xuanzang who participated in the master’s translation of the prajñā-pāramitā texts. Furthermore, for the sūtra’s passage “This spell of the prajñā-pāramitā has been delivered,” which immediately follows the sūtra passage analyzed in the quotation above, Chikō cites fascicle 45 of the Yogācārabhūmi (Yogajishiron), a seminal Yogācāra treatise translated by Xuan zang, which illustrates the dhāran. ī in four aspects: method, meaning, spell, and perseverance. The dhāran. ī serves as an effective method for practitioners to memorize voluminous scriptural texts. Second, its memorization, in turn, enhances practitioners’ capacity to grasp the meanings of countless sūtra texts. Third, the dhāran. ī also intensifies practitioners’ samādhi, empowering them to end misfortunes and sufferings. Finally, the dhāran. ī gives the practitioners perseverance (J. nin; S. ksānti), which enables them to adhere to the Buddha’s teaching against all adversities in order to maintain their spiritual progress.59 As for the dhāran. ī at the end of the Heart Sūtra itself, Chikō states, Gate gate prāgate pārasamgate bodhi svāhā. The spells that appear in many chapters of diverse sūtras are impregnated with the supernatural powers of all the buddhas and bodhisattvas. Each one of the dhāran. ī’s letters and words embraces many meanings. Any attempt to translate them into local languages causes change in the number of their letters and alteration in their meanings that would deprive them of their efficaciousness at recitation. Therefore, it has become customary for translators
192 | ryŪichi abé
of sūtras not to translate them. Some have said that the spells consist of the languages of spirits, gods, demons, and other nonhuman beings, which differ significantly from the ordinary speech of Indian people. Thus, the principle of not translating the spell applies to all sūtras. Yet Kui-ji and Yuan-ce dared to translate the Heart Sūtra’s spell. This is utterly unacceptable. They are not only negligent of the original intent of the dhāran. ī but are preventing it from manifesting its power through recitation.60 Chikō underscores the nontranslatability of dhāran. ī, which as a universal principle should be observed by all doctrinal schools. The issue of inappropriateness in translating dhāran.ī is the very reason for his citation of Jingmen’s commentary and the Yogācārabhūmi, which demonstrates Chikō’s general agreement with the Yogācāra understanding of what the dhāran. ī is and is not about. Chikō’s insistence on not translating the sutra’s dhāran. ī at the end of his Jutsugi also highlights his criticism of Kuiji and Yuance. It is aimed at demonstrating that their approach to the sūtra’s dhāran. ī deviates even from those established standards in some key texts in their own school. Chikō’s commentary on the Heart Sūtra presents a valuable glimpse into the method and style of textual studies performed by the Nara scholar-priests in the eighth and ninth centuries. The essence of his composition lies in his effort to establish the proper interpretation for every sūtra phrase — “proper,” that is, from the viewpoint of his Sanron school. The differences between schools — as expressed in the polemic over the “disputable meaning” and “definitive meaning” — do not express themselves in the manner in which a certain interpretive text is structured. From Chikō’s point of view, all schools agree that a scriptural commentary functions as an exegesis, as a transducer of meaning. In this sense, Chikō remains faithful to the tradition of exegetic literature in Chinese Buddhism. The manner in which he constructs his exegesis is no different from the exegesis of the Sanlun patriarch Jizang or, for that matter, the Yogācāra masters Kuiji and Yuance. Chikō presupposes that all the words in the Heart Sūtra must have right meanings that are most effectively illuminated by Sanron theories. In the light of a range of Sanron commentaries, these words too provide Chikō with opportunities to establish the superiority of Sanron over other schools, especially Hossō. In short, for Chikō and the Nara Buddhist intelligentsia, to interpret means to determine the correct usage of a sūtra’s words — even seemingly ornamental phrases in the original scriptural context, such as the “great, divine spell, the unexcelled spell, and the unparalleled spell” — by replacing them with other ideas, concepts, and definitions, frequently quoted from authoritative treatises, commentaries, and exegeses from the past, which are considered more accessible, definite, and reliable than the original words of the scripture. For his readers to interpret properly, Chikō expects them to perform this semantic exchange within his own Sanron theoretical framework. In this sense, Chikō’s Outline can best be described as an exegetic text typically of kunkogaku. All the qualities of the kunkogaku texts of Nara and early Heian Japan discussed earlier — such as dividing the main text
Scholasticism, Exegesis, and Ritual Practice | 193
into smaller or minute units (passages, phrases, words, and characters), elucidating the meaning of these semantic units against the texts of established authority, and displaying encyclopedic knowledge of the specialized literature — figure as essential elements in his work. Although the subject of discussion is Buddhism, the mode in which Chikō’s text is constructed is identical with the ritsuryō legal exegeses or other genres of Chinese studies texts of the early Heian period discussed earlier. Indeed, the Yakushiji priest Keikai characterized Chikō in Nihon ryōiki as the nemesis of Gyōki (668 – 749), the charismatic Hossō priest who led a popular anti-ritsuryō Buddhist movement: in the eyes of this hagiographer, Chikō was an exemplary Buddhist scholar who was avidly encouraged and supported by the ritsuryō regime.61 Shingyō jutsugi is thus a reflection of historical conditions peculiar to the Nara and early Heian Buddhist community — for example, the conflict between the Sanron school and the Hossō school. With the establishment of Tōdaiji, the Sanron school that developed separately at Gangōji and Daianji began to integrate to form a stronger academic institution.62 Throughout the late Nara and early Heian periods, this rising institution was the intense rival of the Hossō school, which developed rapidly from the Fujiwara clan’s patronage of Kōfukuji, Hossō’s major academic center. However, this contest can be seen as a game that Buddhists were permitted to play in the framework of the ritsuryō state, which encouraged the clergy to be scholars and compete with each other so as to eliminate the aspects of Buddhism that could be a threat to the state, whose ruling ideology was dominantly Confucian.63 It may be said that the aim of Chikō’s Jutsugi is to “rectify the names” in the Heart Sūtra and to dismiss all other naming proposed by competing schools. In this sense, too, Chikō’s work conforms to the standard of scholarly activities that were encouraged under the ritsuryō codes. However, one aspect of Chikō’s texts distinguishes it from other genres of scholarly works produced within the institutional framework of the ritsuryō state. His work is conspicuously lacking in attention to issues of Buddhism’s relationship with the state. Unlike some other major non-Buddhist texts written in Chinese, in Chikō’s text there is no association between writing and statecraft. The manner in which Chikō’s text constructs itself has no bearing on the affairs of the court because, unlike the principal texts of the ritsuryō age, it gave no attention to the practices in the court, nor even to Buddhist rituals sponsored by the state. A case in point is Chikō’s treatment of dhāran. ī, a Buddhist ritual language. The uniqueness of the dhāran. ī’s linguistic form becomes far more pronounced visually in the Chinese translation than in the original Sanskrit sūtra texts. In the Chinese text, the dhāran.ī presents itself as a row of unrelated Chinese characters strung together without any syntactical regularity. The characters for writing dhāran. ī are appropriated only for the transliteration in Chinese: they are hieroglyphic characters eviscerated of their meanings, standing only for their tonal values. Chikō’s meticulous effort to interpret all scriptural phrases ceases as soon as it reaches the dhāran. ī at the end of the sūtra. There no longer are meanings to be transposed by his textual operation; and no longer are the differences in in-
194 | ryŪichi abé
terpreting the meanings to distinguish his Sanron. For Chikō, Buddhist schools cannot but agree upon the dhāran. ī because there is nothing in it for them to differ over — as long as one abides by the rule of not translating the dhāran. ī into local languages. One can certainly describe dhāran. ī, as Jingmen does in his commentary, as that which suppresses evil and encourages good, or delineate its various functions, as in fascicle 45 of the Yogācārabhūmi. However, these texts as well as Chikō’s commentary lack the ability to analyze the dhāran. ī’s efficaciousness. The language of dhāran. ī in Chinese transliteration may be devoid of meaning, but not, supposedly, of power. How, then, does each of the dhāran. ī phrases relate to one another to manifest its power? What method of recitation enables the practitioners to realize dhāran. ī’s effect? In what way can the dhāran. ī’s power of intensifying memory and protecting practitioners be explained? In a sharp contrast to his discussion of the sūtra text proper, which is marked by rigorous logical consistency, Chikō’s explanation of dhāran. ī seems to degenerate quickly into mystification: “the spells consist of the languages of spirits, gods, demons, and other nonhuman beings, which differ significantly from the ordinary speech of Indian people.” This point illustrates a particular gap that existed in the discourse of Chikō and Nara Buddhism generally. Although a plenitude of theories for interpreting scriptures was available for Chikō, none of these theories explained the relationship between the doctrinal discussion and the working of dhāran. īs in a single scriptural text.64 In the Jutsugi Chikō states that the Heart Sūtra’s concluding dhāran. ī — as a great spell, unexcelled spell, and incomparable spell — has the power to conquer heretics and guide lowliest beings to salvation. But he is unable to explain in what manner and for what reason the act of reciting the dhāran. ī would produce such powers for practitioners. Nor does he show in what way the theoretical discussion in the sūtra on wisdom and emptiness is relevant to the act of chanting the dhāran. ī. This hiatus in Chikō’s exegesis is a manifestation of the epistemic disjunction between theory and practice that resulted from Nara Buddhist scholarship’s reliance on the Confucian mode of interpreting texts in their study of Buddhist scriptures. They did not yet possess a language of their own, a Buddhist mode of describing the relationship between the textual and the ritual. Nara and early Heian Buddhists engaged avidly in writing in Chinese and embraced the mode of textual construction exemplary of their age, but because of this disjunction, their Buddhist texts located themselves on the periphery of the texts produced in the ritsuryō age. In particular, dhāran. ī and Buddhist ritual language remained a linguistic excess outside Nara Buddhist theoretical discourse, relegated to the outermost realm of the landscape of Nara and early Heian texts.
Kūkai and the Reversal of the Textual Center and Periphery As will be discussed further below, the introduction of esoteric Buddhism, or mikkyō, during the ninth century was a powerful motivator for Buddhist scholars
Scholasticism, Exegesis, and Ritual Practice | 195
seeking to integrate problems concerning ritual practices within their textual interpretations. As one of those who initiated such a change, Kūkai (774 – 853) is significant not simply because he was the first Japanese figure who systematically studied esoteric Buddhism, but because of his knowledge of Confucian learning and the manifestly Confucian statecraft of his time. In his youth, Kūkai diligently studied Confucianism, Chinese literature, and history at the state academy where his maternal uncle, Ato no Ōtari (n.d.), was a renowned Confucian scholar. As soon as Kūkai returned from his study in China in 806, his exceptional talent in writing and calligraphy was admired by Emperor Saga and his court. Shigeno no Sadanushi (785 – 852), the principal compiler of the Keikokushū, included seven poems by Kūkai in the imperial poetic anthology — more than any poet other than Emperor Saga and Sadanushi himself.65 In 819, as a means to improve his court officials’ writing skills, Saga appointed Kūkai to an office in the Ministry of Secretarial Affairs (Nakatsukasashō).66 It was about this time that Kūkai composed the Bunkyō hifuron, a comprehensive treatise on Chinese poetics, phonetics, syntactical rules, grammar, and rhetoric.67 Because Kūkai was fully aware of the immediate relevance that Confucian learning and the knowledge of Chinese writing had on the affairs of the early Heian state, he considered as a serious problem the gulf in Buddhist scholarship that divided intellectual inquiries from practical matters, especially matters related to the affairs of the state. In the latter part of his life, Kūkai gained a large number of supporters within the Nara priestly community for his effort to introduce esoteric Buddhism into early Heian society. Several prominent leaders of the Nara Buddhist circles are known to have personally received Kūkai’s initiation into esotericism. They include Sanron Master Gonsō (758 – 827) of Daianji, Kegon master Dōyū (? – 851) of Tōdaiji, Hossō master Kenne (? – 872) of Butsuryūji, and Abbot Enmyō (? – 851) of Tōdaiji.68 In 822, he erected at Tōdaiji the Abhiseka Hall (Kanjōdō), the first temple structure in Japan for performing the Mikkyō initiation ceremony.69 In 824, with the endorsement of eminent Nara priests, he joined the Sōgō, the Office of Monastic Affairs, at the rank of junior priest supervisor (shōsōzu).70 In 829 Kūkai was appointed bettō (chief administrator) at Daianji, a major center of Sanron studies.71 In these years he composed a range of commentaries on exoteric Buddhist texts popularly recited and studied in the Nara scholarly circle, including the Vimalakīrti, the Diamond Cutter, the Lotus, and the Golden Light, the sūtras that were considered particularly efficacious in their power of protecting the state. Kūkai’s goal was to demonstrate the effectiveness of the esoteric Buddhist method of reading, which, he claimed, enabled the readers to expand their interpretive operation to the realm of ritual and religious practice, the area that had a direct relevance for the management of the state. Among Kūkai’s commentaries on popular Mahāyāna sūtras is the Hannya shingyō hiken (the title of which might be translated as “The Secret Key to the Prajñā-pāramitā Heart Sūtra,” hereafter referred to as Hiken).72 In the second month of Jōwa 1 (834), the celebrated Sanron master Dōshō (798 – 875) of Gangōji gave a public lecture on Kūkai’s Hiken at the Abhiseka Hall in Tōdaiji.73 Dōshō’s lecture indicates that the Sanron school played a major role in Kūkai’s effort to
196 | ryŪichi abé
build a cooperative relationship with the Nara temple establishment. It also suggests that already during Kūkai’s lifetime, the Nara scholarly community adopted and engaged in a serious study of esoteric Buddhism. To an eye familiar with the exegetic texts of Tang doctrinal Buddhism and their analogs in Nara and early Heian Japan, Kūkai’s Hiken appears drastically different and, perhaps, even outré. In its textual structure, Kūkai’s commentary has nothing to do with kunkogaku, no resemblance to the phrase-by-phrase interpretation favored by his scholarly peers. His comments on the sūtra’s opening passage demonstrate the essential perspective for his interpretation: Sūtra text: When the Bodhisattva Avalokiteśvara was practicing the deep meditation on the prajñā-pāramitā, he discerned that all the five skandhas are empty and thus saved beings from all sorts of suffering. The great Prajñā-pāramitā Heart Sūtra is the Dharma gate to the samādhi of the great [female] Bodhisattva Prajñā’s heart-mantra. The sūtra’s text is not even long enough to fill a page; its lines number only fourteen. The sūtra is simple, yet essential; brief, yet profound. Every phrase encompasses the prajñā preserved in all the five treasuries; every line embraces all the accomplishments of the seven schools....74 Furthermore, the two letters [of the sūtra’s dhāran. ī] ga te swallow the practices and goals of all the Dharma treasuries, and the two dhāran. ī characters prā sam bear the Dharma of both the exoteric and the esoteric. Even a discussion lasting for eons could not exhaust the meaning of a single letter of the sūtra and its sound. Even the multitude of buddhas, countless as particles of dust or ocean foam, cannot explain away a single word in the sūtra or the reality predicated by it.75 Like Chikō and many Chinese masters who composed exegetic texts on the Heart Sūtra, Kūkai understands the sūtra as the gist of Buddhist teachings. However, Kūkai’s reasoning is not shaped by the theoretical framework of Sanron or other doctrinal schools. The sūtra, Kūkai claims, is a guide to the Bodhisattva Prajñā’s samādhi, or deepest meditative experience, that can be realized by the recitation of her dhāran. ī given at the conclusion of the sūtra — which, according to Kūkai, is her heart-mantra. Bodhisattva Prajñā, the prajñā-pāramitā manifesting itself as a female divinity, resides in the Shingon garbha (womb) mandala (Taizō mandara) as the presiding divinity in the “assembly of the holders of wisdom” (jimyōin), located immediately to the west of Mahāvairocana’s central assembly.76 In accordance with its popular characterization in Mahāyāna literature, Kūkai understands the prajñā-pāramitā as kakumo, the mother of enlightenment, and its female bodhisattva manifestation, in the esoteric context, as butsugen butsumo, the eye of the Buddha, the mother of all the buddhas.77 For Kūkai, therefore, the Heart Sūtra is an unfolding of the inmost realization of the Bodhisattva Prajñā, the progenitor of all the buddhas. Furthermore, based on Subhākarasimha’s Commentary on the Mahāvairocana Sūtra,78 Kūkai interprets Bodhisattva Prajñā as vidyā-rājñī (myōhi), the queen of the light of wisdom, one of the common ways in which mantras are described in the Mahāvairocana Sūtra.79 Elsewhere, in his illustration of semiotic
Scholasticism, Exegesis, and Ritual Practice | 197
rules governing mantra syllables,80 Kūkai quotes the following passage from Subhākarasimha’s Commentary for illustration. The term vidyā means the light of great wisdom. The word rājñī is the feminine form of the Sanskrit word for king [rājan]. The queen here refers to the samādhi of the garbha realm of great compassion. This samādhi is the mother of all the Buddha’s children. The Buddha’s children means the bodhicitta, the banner of purity....The one who realizes great emptiness is Prajñā, the mother of buddhas. She is verily vidyā-rājñī, who in her treasury of great emptiness [the garbha realm of great compassion] nurtures the seeds [bija, the seed syllables] of mantra and securely protects them.81 Mantras, or to use Kūkai’s expression, the treasury of mantras, which preserves the sacred seed syllables, are here equated with the Bodhisattva Prajñā, the mother of all the buddhas, who possesses her womb as the garbha realm mandala, the palladium of bodhicitta, the seed of enlightenment. Kūkai’s Hiken, which aims at revealing the samādhi of the Bodhisattva Prajñā, assumes importance beyond a commentary on a popular sūtra; it is Kūkai’s discourse on the nature of mantra. It bears witness to his effort to introduce the novel linguistic form of mantra into the established discourse of Japanese Buddhism. The Hiken is, in this regard, Kūkai’s manifesto of his new Buddhist transmission, to which he refers by the term shingon/mantra. Kūkai sets the tone of his singular interpretation of the Heart Sūtra by placing it side by side with another scriptural text called the “Heart Sūtra.” In fascicle 3 of the Collected Dhāran. ī Sūtra,82 an anthology of shorter sūtras containing various dhāran. īs, translated in 653 and 654 by Adhigupta, one finds a text titled “Great Prajñā-pāramitā Heart Sūtra.”83 The Collected Dhāran.ī Sūtra was copied in Japan as early as 737.84 However, the “Heart Sūtra” chapter in this anthology had escaped the attention of Chikō and other Nara scholar-priests who specialized in the prajñā-pāramitā texts. The narrative in this Heart Sūtra begins with a scene of Śākyamuni Buddha’s assembly at Anāthapindada Park in the Jetavana Grove in the city of Srāvastī, in which Brahmā beseeches the Buddha to expound the merit of the spells and mudras associated with the prajñā-pāramitā. In his reply, the Buddha says that what Brahmā requested belongs to the utmost secret of the prajñā-pāramitā, which the Buddha once expounded for advanced spiritual beings in the celestial realm Paranirmita-vasavartina (Takejizaiten) — the highest heaven in the realm of desire — and that he is about to disclose it again in a way suitable to the audience in the earthly assembly there at Jetavana.85 The Buddha’s terrestrial lecture consists, first, of the method of painting the image of the Bodhisattva Prajñā;86 second, of nineteen yogic procedures comprising combinations of mudras and dhāran.īs to attain union with the bodhisattva;87 and finally, of some devotional practices addressed to the bodhisattva.88 The sixteenth yogic procedure described in the chapter is the recitation of the dhāran. ī named the “great heart” of the bodhisattva: Gate gate pāragate pārasamgate bodhi svāhā.89 This dhāran. ī is of course identical with the concluding dhāran. ī of the Heart Sūtra. Kūkai states,
198 | ryŪichi abé
The meditative method of the mantra [in the Heart Sūtra] is described in detail in fascicle 3 of the Collected Dhāran.ī Sūtra. The title of the chapter [in the Collected Dhāran.ī Sūtra] is identical with the title of the Heart Sūtra translated by Kumārajiva. Among many dhāran. īs in the chapter corresponding to various aspects of the Bodhisattva Prajñā’s body and mind is the Heart Sūtra’s mantra, which is none other than the spell of her great heart. Because of this heartmantra, the sūtra is entitled the Prajñā-pāramitā Heart.90 According to Kūkai’s reading, the Heart Sūtra is an abbreviated teaching the Buddha prepared for Śāriputra, an outline of the sermon he once gave to Brahmā, as recorded in the Collected Dhāran. ī Sūtra. This is why in the “original” version in the Collected Dhāran. ī Sūtra, the method of meditative practice on the Bodhisattva Prajñā is described in detail. This relationship between the Heart Sūtra and the Collected Dhāran. ī chapter explains the reason that the identical mantra/dhāran. ī appears in both texts. In contrast to Chikō, for Kūkai the Heart Sūtra’s dhāran. ī is not an appendage to the scripture’s main text. It is the apogee in which the sūtra’s discussion culminates. However, as indicated above, according to the “Prajñā-pāramitā Heart” chapter in the Collected Dhāran. ī Sūtra, the Buddha’s preaching to Brahmā at Jetavana on the yoga practiced by Bodhisattva Prajñā was also a reworking of his former teaching given to the celestial beings in Paranirmita-vaśavartina (Takejizaiten or Dairokuten), the heavenly realm of desire reigned by Māra. Within the prajñā-pāramitā scriptures, it is only the Path of the Prajñā-pāramitā Sūtra91 that claims to be the record of the Buddha’s expounding the Dharma in Māra’s celestial realm. In this scripture, which simultaneously belongs to the prajñāpāramitā literature and esoteric ritual texts, the Bodhisattva Prajñā is described as “great desire” personified, the desire for the ultimate emptiness and pureness that enables practitioners to attain the great pleasure of liberation. The Interpretive Guide to the Path of the Prajñā-pāramitā Sūtra,92 a ritual commentary imported by Kūkai, identifies the divinity as Kāmavajri, Queen of Desire, a consort of Vajrasattva.93 The meditative ritual of this divinity is described in detail in the Vajraśekhara Yoga Manual on the Five Secret Meditative Practices on Vajrasattva,94 another ritual commentary on the Path of the Prajñā-pāramitā. In short, Kūkai’s essential interpretive strategy can be located in his effort to create an associative linkage across, or the opening up of an intertextual space between, three scriptural texts: The Path of the Prajñā Sūtra, the “Great Prajñā-pāramitā Heart” chapter in the Collected Dhāran. ī Sūtra, and the Prajñā-pāramitā Heart Sutra (table 8.1). In fact, such an intertextual association can be expanded almost infinitely to other scriptural texts that have any bearing on the Bodhisattva Prajñā, what she conceptually represents, and various aspects of her yogic exercise. What Kūkai sees in the Heart Sūtra’s text is the process of condensation, the process in which the texts of associated scriptures reflect themselves on and become overlapped in the Heart Sūtra’s passages. For Kūkai, the sūtra’s words are essentially polysemic. Each word manifests multifarious meanings as they are placed in contexts of pas-
Scholasticism, Exegesis, and Ritual Practice | 199
Table 8.1. Kūkai’s interpretive strategy Location Text Teaching The Buddha’s first teaching on the Bodhisattva Prajñā’s meditation
Paranirmita- vasavartina, the celestial realm of desire
The Buddha’s second Jetavana Park teaching to Brahma
Ritual commentaries
Path of the Prajñā Sūtra
“Great Prajñāpāramitā Heart” chapter, in the Collected Dhāran. ī Sūtra
The Buddha’s Vulture Peak Prajñā-pāramitā third teaching to Heart Sūtra Sāriputra
Interpretive Guide to the Yoga Ritual on the Five Secret Meditative Practices
sages in other interlinked texts. To interpret the scripture means to appreciate the infinite possibilities through which the sūtra’s words produce diverse meanings to the practitioners who, with their knowledge of all other related texts they have studied, engage with the text of their primary interest. There is no intention of limiting the meaning of each word, phrase, and passage into a single, correct definition. Such an operation, as carried out typically by Chikō and the doctrinal scholars of Kūkai’s time, leads only to the impoverishment in the scripture’s rich texture, which constantly generates new meanings for interpretation. Kūkai’s interpretation of the Heart Sūtra’s title illustrates his approach. Kūkai presents the Sanskrit title in the Siddham (Shittan) script: bu-ddha bhā-sa mahā pra-jñā pā-ra-mi-tā hr-da-ya sū-tram.95 Kūkai renders the term bu-ddha as the perfectly enlightened one; bhā-sa, “speech,” as the Buddha’s act of opening the secret treasury and providing sentient beings with amrita; pra-jñā as wisdom arising from samādhi; pā-ra-mi-tā as completion of activities; hr-da-ya as the center, the heart; and sū-tram as a fabric made of letters, namely, text. Having said this, Kūkai states, The title as a whole, however, is endowed with three aspects: personage (nin), Dharma (hō), and metaphor (yu). The title refers to the great Bodhisattva Prajñā-pāramitā. This is the personage. The Bodhisattva has completed the samādhi developed through the mantra of the Dharma mandala, the mandala consisting of each letter [of the title], which is the Dharma. Each word [in the
200 | ryŪichi abé
title] expresses the profundity of the nature of the Dharma by means of the shallow names of the world. This is the metaphor.96 Kūkai reads in the title of the Heart Sūtra as rendered in Sanskrit characters three metaphorically related levels of signification. First, at the most obvious level, the words of the title, arranged in accord with the syntactical rules of Sanskrit compound formation, can be translated as the Scripture of the Heart of the Perfection of Wisdom Preached by the Buddha. However, Kūkai says that such a rendition is only part of the title’s entirety of meaning(s) because each of these words is a metaphor.97 That is, the title’s words carry semantic values that exceed their ordinary, literal meaning. Moreover, the meaning of these words varies according to the figurative forces generated from their context. For example, the term hrdaya (heart), “by means of the shallow, worldly meaning” for an internal organ,˙refers to the gist of the prajñā-pāramitā, the perfection of the transcendental wisdom, and simultaneously, the meditative mind of the Bodhisattva Prajñā, both of which are located at the level of the “profundity of the nature of the Dharma.”98 Kūkai seems to identify this deep level of signification of these words with the syllabic letters used to write them, which are now understood as seed mantras comprising the Dharma mandala, the mandala of monosyllables presided over by the Bodhisattva Prajñā. When the title’s words are broken first into syllabic units and then into phonic fragments, one attains, for example, by changing the position of the vowels and the aspirant in the word bu-ddha, the letter bhah, the seed mantra expressing the samādhi realized by Śākyamuni Buddha.99 Similarly, from the term hrdaya, by changing the vowel ā to semivowel r(a) and semivowel ˙ obtains hrīh , the seed mantra for the samādhi of the Bodhiy(a) to vowel i, one sattva Avalokiteśvara, another ˙key figure in the narrative of the Heart Sūtra. That is to say, the sūtra’s title, as the mother of an almost unlimited number of seed letters, issues from within their syllabic combinations various buddhas and bo dhisattvas, all of whom are claimed in the prajñā-pāramitā literature as progeny of the Bodhisattva Prajñā, mother of all the buddhas. In this manner, through the semiotic production by means of the title’s syllables of seed mantras, Kūkai creates yet another level of metaphorical reference that sees the sūtra’s title itself as the personage of the Bodhisattva Prajñā.100 The three levels of metaphorical signification that Kūkai observes in the Heart Sūtra’s title are, in short, aimed at transforming the title into a mediating process between writing and practice. The key to this mediation is the (intentionally) ambivalent position in which the seed syllables place themselves between the text and the meditative rituals prescribed by the text. Phonic fragments strewn within the title’s words, the seed syllables constitute part of the sūtra’s written text. At the same time, each of the seed syllables contain two aspects indispensable for esoteric Buddhist meditation: its phonic aspect, which regulates the respiratory and chanting exercise of meditation; and its graphic aspect, the object of meditative visualization. These two combined aspects, in turn, make possible the yogic
Scholasticism, Exegesis, and Ritual Practice | 201
exercise aimed at attaining oneness between the practitioner and the deity. Fascicle 30 of Subhakarasimha’s Commentary on the Mahāvairocana Sūtra, for instance, describes the meditation upon the seed mantra in three stages: first, the recitation of the seed mantra as the sounds of inhalation and exhalation; second, the visualization of the entering and exiting breath as streams of particles of light shaped as the seed letter; finally, the visualization of the heart of the divinity, the source of the divinity’s life, the vital breath, as the letter of the seed mantra.101 Kūkai applies the same intertextual strategy for his interpretation of the main body of the sūtra text, which he divides into five sections. Sūtra text: O, Śāriputra, (A) form is no different from emptiness; emptiness, no different from form. Emptiness is none other than form; form, none other than emptiness. And the same holds true for perception, thought, ideation, and consciousness. O, Śāriputra, (B) here, all things are marked with emptiness: they are nonarising, nonceasing, nondefiled, nonimmaculate, nonexcessive, and nondeficient. (C) Therefore, in emptiness, there is no form, no ideation, no consciousness, there are no perceptions, no thoughts. No eye, ear, nose, tongue, body, and mind. No form, sound, scent, taste, touch, and mind’s objects. No field of eye organ and so forth [of the others in the six sensory fields] ending with the field of consciousness. (D) There is no ignorance, no cessation of ignorance, and so forth [of the others in the twelve chains of causation] ending with no aging and death and no cessation of aging and death. There is [likewise] no suffering, no cause of suffering, no annihilation of suffering, and no path of the annihilation of suffering. (E) There is no knowledge, no attainment, and no nonattachment.102 As he did with the sūtra’s title, Kūkai reads the text itself in light of the three levels of metaphorical signification: “personage,” “Dharma,” and “metaphor.” Kūkai states that the five divided sections are (A) establishment (ken); (B) eradication (zetsu); (C) aspects (sō); (D) two (ni); and (E) one (ichi), which correspond, respectively, to Kegon, Sanron, Hossō, the three Hīnayāna schools, and Tendai. Kūkai draws the associations between the five sūtra sections and the Buddhist schools by understanding each of the five sections as brief philosophical renditions of particular meditative experiences realized by the principal buddhas in the Path of the Prajñā-pāramitā Sūtra. For example, regarding passage (B), Kūkai states, The second section is entitled “eradication,” because it points to the gate of samādhi manifested by the Tathāgata Who Eradicates All Sophistries (Skt. SarvaDharma-aprapañca; J. Issai mukeron nyorai) [in the Path of the Prajñāpāramitā Sūtra]....The Tathāgata Who Eradicates All Sophistries is the secret name of Bodhisattva Mañjuśrī. With his sharp sword of the eightfold negation, Mañjuśrī eradicates the mind of illusory attachments. For this reason, he is known by this name.103 Kūkai’s description here is based on a chapter titled “Mañjuśrī” in the Path of the Prajñā-pāramitā:
202 | ryŪichi abé
At that time, Bhagavat, Tathāgata Who Eradicates All Sophistries, expounded the circle of letter-syllables as the path of the prajñā-pāramitā. All things are empty because they coincide with the nature of emptiness. All things are without marks because they coincide with the nature of no marks. All things are without desire because they coincide with the nature of no desire. All things are the light of wisdom because of the pureness of the prajñā-pāramitā.104 The Interpretive Guide has the following comments attached to the above passages. “the Bhagavat, Tathāgata Who Eradicates All Sophistries” This is another name of Bodhisattva Mañjuśrī. “expounded the circle of letter-syllables as the path of the prajñā-pāramitā” The circle of the letter-syllables refers to his [Mañjuśrī’s] samādhi on the circle of the five-letter mantra [A Ra Pa Ca Na].105 Kūkai’s interpretative operation here aims first to understand literally the Heart Sūtra’s passage (B) as the paraphrasing of Nāgarjuna’s celebrated formulation of emptiness as the eightfold negation, as given in his Madhyamaka-kārikā (“neither arising nor ceasing, neither arriving nor departing, neither identical nor different, neither terminal nor permanent”106) the formulation crystallizing the philosophy of the Sanron school. He then takes the same sūtra passage as an invocation of Mañjuśrī, the patron deity of the Mādhyamika/Sanlun/Sanron school, and finally, as the unfolding of the samādhi experienced by the Tathāgata Who Eradicates All Sophistries, in which all forms of sophistic speculation on the nature of things are eradicated — as intended by Nāgarjuna’s eightfold negation. Using this method, Kūkai establishes the correspondence between the five Heart Sūtra sections, the samādhi experiences given in the Path of the Prajñā-pāramitā, and divinities revealed in the Interpretive Guide (table 8.2). Table 8.2 shows that Kūkai identifies the six Nara schools and Tendai as doctrinal institutions located along the literal grain of the Heart Sūtra’s text. Interpreted literally, Kūkai illustrates the sūtra as a collection of phrases epitomizing the positions of each of these doctrinal schools. Chikō’s interpretation of the Heart Sūtra, for example, grounds itself in such a positioning. On the other hand, Kūkai locates the samādhis described in the Path of the Prajñā-pāramitā and the divinities revealed in the Interpretive Guide as aligned with the sūtra text’s metaphorical texture. When read metaphorically, the sūtra manifests itself as an analog of the Dharma-mandala, the mandala made of letter-syllables. Because the Heart Sūtra’s passages analyzed by Kūkai are in Chinese script, one cannot approach them as the matrix productive of bija, seed mantras, as in the case of Kūkai’s interpretation of the sūtra’s title. However, when approached metaphorically, each passage evokes the presence of principal divinities in the Mahāyāna literature who express through their practice of samādhi their insights into the prajñā-pāramitā. Kūkai’s reading of the Heart Sūtra embodies, therefore, a shift from the literal to the metaphorical, a shift that makes it possible to mediate the theoretical ori-
Scholasticism, Exegesis, and Ritual Practice | 203
Table 8.2. Correspondence between the five Heart Sūtra sections Samādhi Bodhisattva Sūtra passage (metaphor) (Dharma) (personage)
School (literal reading)
A. Establishment Establishing Equality of All things Samantabhadra B. Eradication Eradicating All Sophistries Mañjuśrī C. Aspects Entering All the Buddhas’ Mandalas Maitreya D. Two Mastery by the Two Hīnayāna Paths E. One One Pure Nature of All things Avalokiteśvara
Kegon Sanron Hossō Ritsu, Kusha, Jojitsu Tendai
entation of the six Nara schools and Tendai with the orientation toward praxis of esoteric Buddhism. But Kūkai’s own school, Shingon, does not appear in his analysis. That is because shingon/mantra in his Hiken is the very perspective through which this hermeneutical shift from the literal to metaphorical, from theory to practice, is affected. In Kūkai’s interpretation, Shingon becomes visible only in relation to the discussion of the Heart Sūtra’s concluding dhāran. ī, in which the sūtra’s textual language itself changes from the theoretical to the practical. Regarding the concluding part of the sūtra, which he calls “the summary in the dhāran. ī” (sōki jimyō), Kūkai states, Sūtra text: Therefore, one should know the prajñā-pāramitā as the great divine spell, the spell of the great light of wisdom, the unexcelled spell, and the incomparable spell. Because it eradicates all forms of suffering, genuinely, unfailingly, this spell of the prajñā-pāramitā has been revealed. It says, Gate gate . pāragate pārasamgate bodhi svāhā. This part, the summary in the dhāran. ī, has three components: names, substance, and function. The four descriptions of the spell are the name; [the passage] “genuinely, unfailingly,...” refers to the substance [of the spell]; and [the passage] “eradicates all forms of suffering” points to its function. As for the names, the first “great divine spell” suggests the mantra as it is prepared for the srāvakas; the second [“spell of great light of wisdom”] is the mantra as intended for the pratyekabuddhas; the third [“unexcelled spell”], the mantra as appropriate for Mahāyāna; and the fourth [“unparalleled spell”], the mantra as it belongs to the secret treasury (hizō). However, in terms of their broader meaning, each aspect of the mantra is endowed with all the four names.107 Although Kūkai divides the sūtra passage into smaller units, his intention differs significantly from the kunkogaku approach. As will be discussed below, his goal is to establish a corresponding relationship between the sūtra’s text proper and the dhāran. ī at the end of the text. Kūkai first draws a linkage between the sūtra’s four phrases eulogizing its dhāran. ī and the four yānas, or vehicles, of Buddhism.
204 | ryŪichi abé
They are the two Hīnayāna vehicles (which include the Kusha, Jōjitsu, and Ritsu schools), Mahāyāna (in which the Sanron, Hossō, and Tendai schools are located), and the Mantrayāna (the Shingon school). And here, as elsewhere throughout his work, Kūkai employs the term hizō, or secret treasury, interchangeably with Mantrayāna to refer to his own Buddhist transmission108 In Kūkai’s explanation of the Heart Sūtra’s dhāran. ī, these four laudatory phrases are linked, respectively, to the first four words in the dhāran. ī. The dhāran. ī, the mantra of the secret treasury, has five components. The first “gate” [O, she who is gone] discloses the goal of [the vehicle of] the śrāvakas; the second “gate” elucidates the goal of the vehicle of the pratyekabuddhas; the third “pāragate” [O, she who is gone beyond] points at the most excellent goal of the schools of Mahāyāna; the fourth “pārasamgate” [O, she who is thoroughly gone beyond] unveils the goal of the Mantrayāna and its mandalas perfectly endowed with virtues. The fifth “bodhi svāhā” [O, the awakening, hail!] expounds the enlightenment at which all the various vehicles ultimately arrive.109 As in his explication of the Heart Sūtra’s title, Kūkai resorts again to the Siddham script to render the dhāran. ī, highlighting its linguistic idiosyncrasy. By this means — that is, by not resorting to the established convention in the scriptures imported throughout the Nara period of transliterating the dhāran. ī with Chinese characters — Kūkai underscores that the Heart Sūtra’s dhāran. ī is mantra, the dhāran.ī of the “secret treasury.” Instead of being a dhāran.ī in the exoteric Mahāyāna context — an auxiliary device to facilitate memorization of the sūtra text or to extend protection to the chanters — the sūtra’s concluding dhāran. ī is, for Kūkai, the climax in which the scriptural text culminates. The dhāran. ī understood in this manner is not merely the condensed essence of the sūtra’s text and its related discourses on the six Nara and Tendai schools. It is also a miniature of the entire sūtra text revealing the process of the compression of ordinary language into mantra (in the diachronic order of the text read from the beginning to the end) and the reverse process of mantra unfolding itself into ordinary language (Bodhisattva Prajñā’s mantra-meditation unfolded, first into the metaphorical languages of the Path of the Prajñā-pāramitā, the Collected Dhāran. ī, and the Heart, and then to the doctrinal language of the seven schools). The vantage point of mantric language makes it plain that the distinction between the seven early Heian schools are not inherent but derive from the different ways in which the prajñā-pāramitā manifests itself for audiences of diverse capacities. In this sense, Shingon for Kūkai is both inside and outside the Heart Sūtra’s text. It crystallizes itself as the scripture’s concluding dhāran. ī. But it is also an interpretive perspective, or a hermeneutical technology, which he employs to explain the sūtra’s doctrinal discourse into the realm of ritual practice.
Scholasticism, Exegesis, and Ritual Practice | 205
Writing, Ritual, and Post-ritsuryō Buddhism In the opening of the Hiken, Kūkai proclaims the Heart Sūtra as the disclosure of Bodhisattva Prajñā’s samādhi and the concluding dhāran. ī as her heart-mantra.110 In addition, the female bodhisattva is understood by Kūkai as vidyā-rājñi, the queen of wisdom-light, a designation in the Mahāvairocana Sūtra of what is universally mantra. The function of the sūtra’s dhāran. ī is, therefore, twofold: the unveiling of both the inmost meditative state of the bodhisattva and the inner working of the mantra as sacred language. The dhāran. ī is the manifestation of the female divinity in language and as a language for religious practice. At the same time, it functions as the meta-dhāran.ī demonstrating how a dhāran.ī works as mantra, a sacred language endowed with the power of removing delusions. Because of this work of the dhāran. ī, Kūkai in the Hiken treats the sūtra’s mantra, the samādhi attained by the recitation of the mantra, and the divinity personifying the mastery of the mantra’s meditation as exchangeable and replaceable. Kūkai states this point poetically in the following verse: Beyond thinking, beyond explication is the mantra. Chanted, it swiftly removes the darkness of ignorance. With every one of its letters encompassing myriad truths, Suchness of Dharma arises right within the chanter’s body. Practicers, go further to find the pristine peace, Search deeper and enter the original abode of your mind. Traverse the triple world, the lodging house, Return home, the one mind of the Tathāgatas.111 When read literally, the dhāran. ī, rendered as feminine singular, vocative, “gone, gone, gone beyond, gone thoroughly beyond,” is by itself almost meaningless. The dynamism of the dhāran. ī as mantra unveils itself as soon as the “gone” in the dhāran. ī is understood as a trope for the female divinity whose name is Prajñāpāram-itā, “She Who Has Gone Completely to the Transcendental Wisdom.”112 Because, for Kūkai, the mantra is the divinity, its sound — “gate, gate, pāragate, pārasamgate” — is the progressive movement in which the bodhisattva, the queen of wisdom-light, unleashes her effulgent wisdom to destroy avidyā (darkness, ignorance), the root cause of delusion and suffering. The dhāran. ī’s sound is the resonance of the galloping gait of the bodhisattva with which she drives various vehicles of the Buddha Dharma to pass beyond samsara and reach nirvana. Recitation of the dhāran. ī, then, allows practitioners to participate immediately in the bodhisattva’s act of leading living beings to enlightenment. It is precisely this dissolution of the literal into the metaphorical that illustrates the efficaciousness of the dhāran.ī. That is because such dissolution is tantamount to the shift in the nature of scriptural reading from intellectual speculation to religious practice. The movements of reader’s lungs, of vocal cords, tongue, and lips for reciting and chanting the sūtra text now becomes an integral part of
206 | ryŪichi abé
their reading and practice, the reading that constitutes their physical labor. In other words, the sūtra’s metaphoricity provides the critical moment in this shift through which the scriptural text and the readers’ somaticity integrate themselves into a ritual action. For Kūkai, the expansion as such of the sūtra’s textuality into the realm of the body is most essential for manifesting the power inhering in the scriptural texts — he frequently expressed this through deployments of the metaphor of medicinal power. In the twelfth month of 834, just a few months before his death, Kūkai sent a memorial to Emperor Ninmyō (r. 833 – 850), requesting the inauguration of an esoteric Buddhist ritual for celebrating New Year at the Imperial Palace, later known as the second seven-day New Year ritual (goshichinichi mishuhō). Kūkai’s goal was to perform a ritual ceremony to invigorate the emperor by appropriately chanting the mantras and dhāran. īs contained in the Golden Light Sūtra. This scripture claims that its mantras and dhāran. ī have power to transform a ruler to the cakravartin, the legendary Buddhist monarch who rules the world by his virtuous mastery of the Dharma.113 Kūkai proposed his ritual to be performed as evening services, as an extension of the imperial purification rite (misaie), the seven-day recitation and lecture on the Golden Light Sūtra at the palace performed during the second seven days of the first month by the leaders of the Nara Buddhist schools. Kūkai proposes, The buddhas’ preaching of the Dharma is of two kinds: one revealing shal- low meaning (senryakushu), the other, secret meaning (himitsushu). The shallow meaning is expressed in the prose lines and hymns of the scriptures. The secret meaning is contained in the scriptures’ dhāran. ī. The shallow meaning is like examining the sources of illness and analyzing the nature of medicaments in the textbooks of medical science. The secret method of dhāran. ī is just like compounding medicines and, by providing them to the sick according to prescription, removing illness. Opening a medical text and giving a lecture about it to the patient does no good in curing his illness. One must compound drugs in response to symptoms and provide them to the patient following prescriptions. Only then can one eradicate disease and save lives.114 This statement illustrates Kūkai’s hermeneutics aimed at constructing a new type of Buddhist discourse in which text and action are no longer two immiscible categories — as was the case in the writings of Chikō and his Nara scholarly peers — but intertwined fields, as exemplified in the ritual action of uttering dhāran. īs and mantras. Through such an innovative discourse, Buddhism became immediately relevant to the management of the state. Kūkai’s founding of the New Year ritual service provided the watershed for diverse esoteric rituals to be rapidly incorporated within the services of the imperial court. Cases in point are the integration within the emperor’s coronation ceremony of abhiseka, esoteric Buddhist ordination (sokui kanjō), the ritual services based on the Sūtra of the Virtuous King (ichidai ichido ninnōe), and the ritual worship of the Buddha’s relic (ichidai ichido busshari).115 The rapid growth of esoteric Buddhist rituals at the Imperial Palace and the
Scholasticism, Exegesis, and Ritual Practice | 207
court explains the reason for the establishment of grand Shingon temples in the vicinity of Kyoto as new centers of Buddhist ritual studies. In 851 Shingon master Shinga (801 – 879) received the direct patronage of Emperor Monkotu and Minister of the Left Fujiwara no Yoshifusa (804 – 872) and founded Kashōji at Fukakusa. Daikakuji was built in 876 by the Shingon priest Kōjaku, a son of Emperor Junna, and by Kōjaku’s mother Shōshi; Ninnaji was established in 888 by Emperor Uda (r. 887 – 897); Kajūji was erected in 905 by imperial consort Inshi; and Hosshōji, an extensive monastic complex in the northwest of Kyoto, was completed by 925, under the supervision of the Fujiwara regent Tadahira (880 – 949).116 It must be noted that Kūkai’s proposal quoted above accentuates the significance of doctrinal studies (“medical texts”) as the foundation for esoteric rituals (“application of medicine”). Esoteric rituals were understood as complementary to doctrinal studies, which encouraged the Nara Buddhist temples to actively integrate esotericism within their own institutional structure, a move that ascertained their immediate participation in the crucial affairs of the state. In 848, under the aegis of the imperial dowager, Fujiwara no Junshi, Hossō Master E’un (798 – 869) erected Anshōji in Yamashina. In 874, Sanron Master Shōbō (832 – 909) of Tōdaiji founded Daigoji in Fushimi. These temples soon established themselves as prominent academic centers of the Shingon school that provided the knowledge of esoteric rituals for Nara scholar-priests.117 Unlike the major monasteries founded by the state in the Nara and the early Heian periods, all these new centers of Buddhist ritual studies, built around Kyoto, enjoyed freedom from government supervision. Run by private funds given by imperial and aristocratic family members, these temples received the court’s exemption from the Rules for Priests and Nuns of the Ritsuryō. For example, they were able to acquire ordinands outside the annual allotment given by the state to each Buddhist school. This arrangement encouraged a swift increase from the mid-Heian period onward in the number of priests who specialized in esoteric Buddhist studies.118 At these new centers, the scholar-priests focused their textual productions no longer in doctrinal problems but in writing ritual commentaries, ritual manuals, and liturgical texts — amounting to nothing less than a reversal of the center and periphery of Buddhist textual discourse of the Heian period.119 Many of their works evolved around the analysis and treatment of dhāran. ī and mantras, which they studied in Siddham, in phonetic letters.120 Their emphasis on the phoneticism demonstrates vividly the movement away from the norms established for producing authoritative texts in the framework of kangaku in early Heian society, the norm as it was once exemplified by Chikō’s text.
Notes 1. See Kurozumi Makoto, “Kangaku: sono shoki, keisei, ken’i,” 213 – 256. 2. The texts adopted by Kanmu are Chunqiu gongyangzhuan (Shunjū kuyōden) and Chunqiu guliangzhuan (Shunjū kokuryōden). See Ryō no shūge, fscl. 15, 447 – 448. 3. Momo Hiroyuki, Jōdai gakusei ronkō, 70.
208 | ryŪichi abé 4. Suzuki Kazuo, “Heian jidai no gakusei, kyōiku,” 198 – 200. 5. Inoue Mitsusada, Nihon kodai shisō shi no kenkyū, 137 – 140. 6. Nihon shisō taikei (hereafter NST), 3:786 ff. 7. Ibid. 8. Ikeda Genta, Nara Heian jidai no bunka to shūkyō, 275. 9. Nihon sandai jitsuroku, Jōgan 18 (876) 10/4. 10. Zoku kojidan chūkai, fscl. 2, “shinsetsu,” 24. 11. Kawaguchi Hisao, Heian chō no kanbungaku, 11 – 16. For more on the topic of the role and importance of kanshi, see Ivo Smits’ essay in this volume. 12. Gunsho ruijū (hereafter GR), 8:449a. 13. GR 8:490a. 14. Gotō Akio, “Kyūtei shijin to ritsuryō kanjin,” 34 – 36. Kawaguchi, Heian chō no kanbungaku, 162. 15. Only two fascicles of this vast work survive. Fascicle 864 is preserved in the Seikidō Bunko Collection at Ochanomizu Toshokan, Tokyo; and fascicle 868, in the Sonkeikaku Bunko Collection in Komaba, Tokyo. 16. Translation by Wing-tsit Chan in Chan, trans., A Source Book in Chinese Philosophy, 47. Translation slightly modified. 17. Ikeda, Nara Heian jidai no bunka to shūkyō, 260. 18. Sakamoto Tarō, “Rikkokushi to Montoku jitsuroku,” 3 – 12. 19. Kawaguchi, Heian chō no kanbungaku, 52. 20. Konishi Jin’ichi, Nihon bungaku shi, 49. 21. Tsurezuregusa, episode 122, in Shin Nihon koten bungaku taikei (SNKBT), 39:196. 22. One of the most obvious phenomena illustrative of this change was the formation of kagaku and yūshiki, lineages of particular, and often minor, aristocratic families who specialized in producing scholars and ritual experts and preserved their knowledge through hereditary transmission. See Uesugi Kazuhiko, “Heian jidai no ginō kanjin,” 180 – 192. 23. Inoue, Nihon kodai shisō shi no kenkyū, 232 – 236. 24. Ryūichi Abé, The Weaving of Mantra, 38 – 40. 25. Yoshida Kazuhiko, Nihon kodai shakai to bukkyō, 30. 26. NST 3:219 – 220. 27. Ryō no shūge, 232 – 233. 28. Kanō Naoki, Kanbun kenkyū ho, 89 – 93, 148; Shirakawa Shizuka, Shirakawa shizuka chosakushu, 130. 29. Ikeda, Nara Heian jidai no bunka to shūkyō, 257 – 260. 30. The commentarial texts Naomoto cites range from the oldest extant Ritsuryō exegesis dating from the early Nara period, the Koki — the two commentaries issued during the reign of the Emperor Kanmu, one (Ryōshaku) written by legal experts who served his court and the other (Anaki) composed by a certain State Academy Professor Anata — to two additional commentaries dated from the early Heian period, one (Atoki) preserved in the House of Ato and the other (Shuki) by an anonymous author who added his comments in red on the main Ritsuryō text. 31. David Hall and Roger Ames, Thinking through Confucius, 268 – 275. 32. Shoku Nihongi, fscl. 33, 416. 33. Dai Nihon bukkyō zensho (henceforth DNBZ) 1, no. 1. Itō Ryūju, “Anchō no in’yō seru sho chūsakushō no kenkyū,” 121b – 122b; Hirai Shun’ei, “Nanto sanronshū shi no kenkyū josetsu,” 144 – 145. 34. Elizabeth ten Grotenhuis, Japanese Mandalas, 34.
Scholasticism, Exegesis, and Ritual Practice | 209 35. KBZ, 6:167. 36. Hirai Shun’ei, Chūgoku hannya shisō shi kenkyū, 144. 37. Chūron shoki, in Taishō shinshū daizōkyō (hereafter T), vol. 65, no. 2255. 38. Daijō sanron taigishō, T 70, no. 2296. 39. Sueki Fumihiko, “Chikō hannya shingyō jutsugi ni tsuite,” 70 – 71, 244. 40. The first Hossō counterargument to Chikō’s interpretation of the Heart Sūtra appeared in the Hannya shingyō ryakushō by Abbot Shinkō (934 – 1004) of Kojimaji, a prominent midHeian-period priest renowned for his combined mastery of the Hossō and Shingon. 41. DNBZ 1:169b. 42. The order in which these four parts are arranged differs somewhat among surviving manuscripts of the Shingyō jutsugi. My description here is based on the printed edition in volume 1 of DNBZ. For the variations between the surviving manuscripts and woodblock editions, see Sueki, “Chikō hannya shingyō jutsugi ni tsuite,” 161 – 167. 43. DNBZ 1:169c. 44. See, for example, T 33, no. 1712:552a – b; T 33, no. 1710:523c – 524a; and T 33, no. 1711:543b – c. 45. DNBZ 1:169c – 170a. 46. Sanlun xuanyi (Sanron gengi). T 45, no. 1852:10b. 47. In particular, see Dapin youi (Daihon yūi) and Renwang bore jing su (Ninnō hannya kyō sho). T 33, no. 1696; and T 33, no. 1707. For the influence of these works on Chikō, see Sueki, “Chikō hannya shingyō jutsugi ni tsuite,” 167. 48. DNBZ 1:170a. 49. DNBZ 1:170b – c. 50. See Kuiji’s bore boluomiduo xinjing youzan (Hannya haramita shingyō yūsan) and Yuance’s Fushou bore boluomiduo xinjing zan (Bussetsu hannya haramita shingyō san). T 33, no. 1710 and no. 1711. 51. DNBZ 1:170a. 52. For the Mādhyamika thinkers of the Indo-Tibetan tradition, however, such an understanding of the definitive and the disputable, entailing the Yogācāra’s conceptualization, was not acceptable. Mādhyamika’s counterargument focuses on refuting as naïve and simplistic Yogācāra’s identification of the definitive as literal. See Robert Thurman, “Buddhist Hermeneutics,” 28 – 34. 53. See Sone Masato, “Heian shoki Nanto bukkyō to gokoku taisei,” 655 – 718. 54. For dhāran. ī, mantra, and their linguistic characteristics, see Abé, The Weaving of Mantra, 159 – 168, 240 – 246. 55. DNBZ 1:170a – b. 56. Ibid., 170c. 57. Ibid., 174c – 175a. 58. Bore xinjing su (Hannya shingyō sho), in Dai Nihon zokuzōkyō, case 1, vol. 41, bk. 3, p. 218; henceforth NZ 1:41:3, 218. 59. T 30, 542c. 60. DNBZ 1:175c. 61. Keikai relates a story in which Chikō was consumed by jealousy at the imperial appointment of Gyōki to the post of grand priest master (daisōjō) in 744. Protesting what he considered an unjust appointment, Chikō returned from Gangōji to the remote temple of Sukidadera in southern Kawachi Province. In disgust, Keikai continues, Chikō declared, “I am a man of wisdom, while Gyōki is a petty priest. How was it possible for the emperor to overlook my erudition and to promote, instead, a mere worthless priest?” Instead of a request
210 | ryŪichi abé from the emperor that he return to the capital, Chikō was immediately summoned by King Yama’s court, where the pundit was found guilty of temerity and ignorance and was sentenced to horrendous punishments in hell (NKDZ 6:167 – 168). See Katsuura Noriko, Nihon kodai no sōni to shakai, 336 – 339. 62. Hirai, Chūgoku hannya shishō shi kenkyū, 150. 63. Hayami Tasuku, “Ritsuryō kokka to bukkyō,” 3 – 17. 64. In his Summary Discussion on the Lotus Sūtra (Hokke rakusho), Hossō master Myōitsu (728 – 798) of Tōdaiji explains the sūtra’s dhāran. ī chapter (chap. 28; T 8, 58b – 59b), showing the doctrinal scholar’s insensitivity to ritual language: “These are called dhāran. ī in Sanskrit and are translated as zongchi (sōji) because they manifest an incomparable spiritual power of protection by means of a small number of secret, meaningless letters. These letters destroy evil and establish good. Therefore, they are called dhāran. ī. This chapter is called the dhāran. ī chapter because it indicates such effects of the dhāran. īs” (T 56, 143b). 65. Ichikawa Mototarō, Nihon kanbungaku shi gaisetsu, 78. 66. Takagi Shingen, Kōbō Daishi no shokan, 86 – 87. 67. Katō Seiichi, Kōbō Daishi Kūkai den, 95 – 97. 68. Abé, The Weaving of Mantra, 45. 69. Tōdaiji zoku yōroku, ZZG 11:287a – b. 70. Sōgō bunin, DNBZ 65:9a. 71. Tōji chōja bunin, GR 4:622b. 72. Kōbō Daishi zenshū (hereafter KDZ), 1:555. 73. Tōdaiji yōroku, fscl. 4, ZZG 11:69a. 74. The Six Nara schools and the Tendai school. 75. KDZ 1:555. 76. Taizō genzu mandara, divinity no. 10, in Sawa Takaaki, ed., Shingon jiten (hereafter SJ), appendix 32. The description of the Bodhisattva Prajñā in the garbha mandala is based not on the Mahāvairocana Sūtra itself but on Subhakarasimha’s commentary on the sūtra, which explains the location of the Bodhisattva Prajñā in the mandala as the seat for the ordinand in the garbha abhiseka (T 39, 612b – c, 622b, 673b – c). For the problem of discrepancy between scripture and mandala images, see Robert Sharf, “Visualization and Mandala in Shingon Buddhism.” 77. See KDZ 1:554, 551. 78. Dapiluzhe’na jing shu (Daibirushana kyō sho). T 39, no. 1726. 79. Dapiluzhe’na jing (Daibirushana kyō, a.k.a. Dainichi kyō). T 18, fascl. 3, fascl. 4. 80. KDZ 1:551. 81. T 39, 673b – c. 82. Tuolouni jing (Darani jikkyō). T 18, no. 901. 83. T 18, 785a – b. 84. Dai Nihon komonjo, 7:75. 85. T 18, 804c – 805a. 86. Ibid., 805a – c. 87. Ibid., 805c – 808a. 88. Ibid., 808a – 812b. 89. Ibid., 807b. 90. KDZ 1:557. 91. Prajñā-pāramitā-naya-satapañcasatika. It is also known as Adhyardhasatika-prajñāpāramitā. Bore boluomiduo liqui jing (Hannya haramita rishu kyō). T 8, no. 243. 92. Bore boluomiduo liqui shi (Hannya haramita rishushaku). T 19, no. 1003. 93. T 19, 616b – 671a.
Scholasticism, Exegesis, and Ritual Practice | 211 94. Wumimi xiuxing niansong yigui (Gohimitsu shugyō nenju giki). T 20, no. 1125. 95. Siddham is an Indian script most commonly used in premodern East Asia for writing Sanskrit words. The Siddham script as it now appears in the text of Shingyō hiken in KDZ 1:556 reads, Buddhabāsa mahāprajñā pāramitā hrdā sūtram. It is unknown whether this was ˙ a wrong spelling given by Kūkai himself or resulted from the corruption through the process of copying by those later priests who lacked the knowledge of Sanskrit. 96. KDZ 1:556 – 557. 97. In this sense, Kūkai’s understanding is similar to that of many of the pioneers of contemporary theories on metaphor, such as I.A.Richards (The Philosophy of Rhetoric, 108 – 109) and Max Black (“Metaphor,” Proceedings of the Aristotelian Society, 273 – 249), who see metaphor not as a rhetorical ornament, but as a seminal mode of human cognition generative of new perspectives on reality. Also see Paul Ricoeur (The Rule of Metaphor, 247 – 256) and George Lakoff and Mark Johnson (Metaphors We Live By, 156 – 184). 98. KDZ 1:556 – 557. 99. Mahāvairocana Sūtra, fscl. 2. chap. 4. T 18, 16c; SJ, no. 628. 100. Compare Kūkai’s operation here with Saussure’s analysis of anagrams in his study of ancient Vedic mantras. See Julia Kristeva, Sèmeiōtikè — Recherches pour une sémanalyse, 174 – 207; and Jean Starobinki, Words upon Words, 11 – 19. 101. T 39, 785c. 102. KDZ 1:558 – 559. 103. Ibid. 104. T 8, 785a – b. 105. T 19, 613 b – c. 106. T 30, 1b. 107. KDZ 1:560. 108. Abé, The Weaving of Mantra, 191 – 200. 109. KDZ 1:561. 110. KDZ 1:555. 111. KDZ 1:561. 112. To follow Kūkai’s reading and to emphasize the parallelism between the dhāran. ī and Bodhisattva’s name, I am interpreting the term prajñā-pāramitā here as an accusative tatpurus.a compound with the feminine ending of the past participle of the verb i (to go). For an elucidation of the accusative tatpurus.a with past participles, see The Siddhānta Kaumudhi (SK) 1:406. 113. Kanaoka Shūyū, Konkōmyō kyō no kenkyū. 114. KDZ 3:518. 115. Abe Yasurō, “Hōshu to ōken,” 140; Hashimoto Hatsuko, “Busshari kankei kaidai,” 209. 116. For a detailed discussion of these temples as external to the ritsuryō institutional structure, see Hiraoka Jōkai, Nihon jiin shi no kenkyū, 483 ff.; and Abé, The Weaving of Mantra, 367 – 370. 117. Hiraoka, Nihon jiin shi no kenkyū, 287ff. 118. Abé, The Weaving of Mantra, 371 – 376. 119. Hayami Tasuku, Heian kizoku shakai to bukkyō, 3 – 29. 120. See, for example, Ono rokuchō by Ningai (no. 2473), Betsugyō by Kanjo (no. 2476), Atsuzōshi by Genkai (no. 2483), Hishō by Shōken and Shukaku (no. 2489), and many other Heian-period ritual manuals collected in volume 78 of the Taishō shinshū daizōkyō.
9
d Mik ael Adolphson
Institutional Diversity and Religious Integration The Establishment of Temple Networks in the Heian Age
T
he cultural accomplishments of the Heian age, most notably in literature and arts, have been admired, reproduced, and reinterpreted for centuries. Although not quite as enduring, political and religious institutions as well as ideologies created and developed in the ninth and tenth centuries similarly lasted much beyond the Heian age itself, some even until the late sixteenth century. For example, warriors vied for court titles in the fourteenth century, and legitimacy from the imperial court was still of tremendous importance to sixteenth-century daimyō.1 Granted, while the imperial court did not wield much actual power in national politics during these centuries, the monastic complexes, acting as the last defenders of a court-centered system, if only to protect their own privileges and independence, were still beyond the direct control of most members of the warrior class until Oda Nobunaga. Yet despite their lasting influence, the process by which Enryakuji, Kōfukuji, Onjōji, and Tōdaiji emerged as religious elites with large communities of monks and secular laborers, tracts of tax-exempt estates, and networks of branch institutions and political influence in the late Heian age, and the conditions that allowed this development, remain remarkably elusive. One of the least explored aspects is their astonishing ability to incorporate other temples and cloisters into a cohesive temple network. Why, for instance, did religious institutions such as Tōnomine, Kiyomizudera, Gionsha, and Daisenji, to mention just a few, become affiliated with elite temples as branches despite their comparatively independent origins? The answer to this question has implications beyond the confines of religious institutional history, since a scrutiny of the establishment of individual cloisters and chapels, their role within the imperial state, and their eventual incorporation or development into larger monastic complexes will also enable us to examine more far-reaching issues of center and periphery in the religious and political worlds of Heian Japan. To this end, this essay will explore the religious policies of the Nara and early Heian ages, paying particular attention to the court’s efforts to control the Buddhist establishment despite strong centrifugal tendencies even among the nobles themselves. Early proclamations against private cloisters were soon replaced by various denominations to differentiate between official and private temples, but it was, in the end, the largest elite monasteries in central Japan that came to reverse the trend through the establishment of temple networks
Institutional Diversity and Religious Integration | 213
that came to recenter the religious hierarchy. These small, and often overlooked, cloisters and provincial temples are the main focus of this essay.
Religious Policies and Politics in Nara and Early Heian The establishment of Buddhist temples and the role of Buddhism in the seventh and eighth centuries ranged from severe curtailing to unrestricted founding of new cloisters and halls with generous grants and the promotion of monks to unprecedented heights. On the one hand, Emperor Tenmu (r. 673 – 686), in his quest to create a court-centered polity in the late seventh century, made it a priority to control Buddhism and its institutions. Tenmu’s order to restrict the land grants, administration, and names of temples in 680 reflected an extensive effort to bring temples supported by individual clans and aristocrats under the control of the court. On the other hand, there were also times when temples were freely established, and large numbers of new monks were ordained every year, as in the mid-eighth century, during the rule of Emperor Shōmu (r. 724 – 749) and Empress Kōken/Shōtoku (r. 749 – 758; 764 – 770).2 Following the reign of Shōtoku, whose patronage of Buddhism allowed the infamous Dōkyō (? – 772) to become an influential member of the court in the 760s, her successor, Emperor Kōnin (r. 770 – 781), was determined to contain the political influence of Buddhist monks.3 Consequently, directives were issued admonishing the illegal behavior of monks, and Kōnin even used the destructive results of a lightning strike on Yakushiji and Katsuragidera to rebuke the Buddhist clergy in a famous edict of 780.4 Sharing his father’s concerns, Emperor Kanmu (r. 781 – 806) continued to curb the privileges, misappropriations, and unrestricted establishment of private temples. Thus, an edict was issued in 783 prohibiting the private construction of new cloisters, specifically out of concerns for the vast tracts of land that they might obtain and thus remove from the registers of taxable rice fields.5 In addition, Kanmu’s decision to establish a new capital with a less distinct religious presence was in part motivated by concerns of the growing influence on politics of monks and temples.6 In 795, the year after the move to Kyoto, Kanmu repeated his ordinance, adding that land previously donated to temples had to be registered with the government, and he had the governor of Yamato Province investigate the behavior of the clergies in Nara three years later.7 These efforts notwithstanding, there are also records of noble patrons and exalted monks establishing smaller temple halls in various locations even during Kanmu’s reign. For example, Fujiwara no Kiyokawa’s (n.d.) residence was turned into a cloister named Saion’in in 792, and Sakanoue no Tamuramaro (758 – 811) built Kiyomizudera east of Kyoto in 798, in gratitude for and celebration of his successful campaign against the Ezo in northeastern Japan.8 Founded to commemorate and perform religious ceremonies for an ancestor, these temples were by necessity promoting a specific family or lineage, making such temples potential obstacles to a court-controlled and unified hierarchy of Buddhist temples. In the end, Kanmu himself contributed to the court’s failure to restrict the
214 | mikael adolphson
growth of private temples. Seeking to broaden the religious foundation of the early Heian state, Kanmu promoted, as is well-known, young monks, such as Saichō (767 – 822) and Kūkai (774 – 835), who offered alternatives to the established Nara schools. Although the monks of Tōdaiji and Kōfukuji in Nara remained influential in religious matters, the court’s subsequent support of Saichō’s Tendai and Kūkai’s Shingon schools in effect worked against Kanmu’s own efforts to curtail the establishment of smaller cloisters and chapels outside the control of the government. It was esoteric Buddhism, promulgated above all by Tendai and Shingon, with its offerings of a wide range of benefits through a range of specific rituals that best responded to the social, political, and religious needs of the capital elites. As Stanley Weinstein has noted, the Heian aristocrats became “intoxicated with Mikkyō [esoteric] rituals, which were performed to achieve specific material or spiritual ends.”9 Thus, Kanmu’s regulations notwithstanding, the introduction of new schools and the move of the capital liberated monks and patrons to establish new cloisters beyond the institutional control of the Nara schools and the imperial court. The sponsoring of new temples was not only a question of religious fervor or a desire to broaden studies of Buddhism in various schools; it could also serve to enhance the status of the patron himself. For example, the alliance between Fujiwara no Morosuke (908 – 960) and the Tendai monk Ryōgen (912 – 985) was as much a political creation as a reflection of Morosuke’s religious fervor. Morosuke sponsored the construction of cloisters in Ryōgen’s section of Enryakuji — the Tendai center on Mount Hiei — and had several of his sons become ranking monks to oversee the new centers, creating, in effect, a new cultic network for his own subbranch separate from the clan’s main religious center at Kōfukuji. As a result, Ryōgen’s career took off, and he became one of the most important head abbots on Mount Hiei, while Morosuke and his descendants managed to gain control of the headship of the Fujiwara family. In short, with the advent of esoteric rites, Buddhism became useful for the individual in his lifetime in addition to being used to sacralize the imperial state.10 This building spree created problems for the bureaucratic state, since many of the new temples established after the court’s move to Kyoto followed Enryakuji’s lead in seeking independence from the Office of Monastic Affairs (Sōgō). Still depending on the support of nobles, however, these temples served the courtiers in a more direct capacity while still claiming their role as “protectors of the state.” In this way, an important link was created and maintained between the benefactor, the temple itself, and the state despite the private initiative that lay behind the founding of new institutions. This complex and interdependent relationship between religion and politics compels us to look more closely at both sides of the coin: to examine the religious trends within the political hierarchy at the imperial court and the political setting behind the development of Buddhism and the locales devoted to its worship. Kanmu’s forceful leadership belied a continued and intense competition between the leading families at the imperial court. During the short reign of Kan mu’s successor, Heizei (r. 806 – 809), Fujiwara no Kusuko (? – 810) and her brother
Institutional Diversity and Religious Integration | 215
Nakanari (774? – 810) of the Ceremonial Branch were involved in an incident in 807, in which their rivals of the Southern Branch were suppressed.11 Two years later, Heizei became ill, and abdicated in favor of his younger brother, Saga. However, once recovered, Heizei made an attempt to reascend the throne and move the capital back to Nara with the help of Kusuko and Nakanari, but the insurgence failed, resulting in the elimination of the Ceremonial Branch from the high end of politics and a shift in favor of the Northern Branch (hokke) within the Fujiwara clan.12 Yet both Emperor Saga (r. 809 – 823) and his successor Junna (r. 823 – 833) retained a high degree of independence and led the court successfully for the next three decades. Although both were favorable to esoteric rites and particularly supportive of Shingon, the increase of Buddhist institutions was modest. However, a substantial shift occurred from the 840s, coinciding with the rise of the Northern Fujiwara under Yoshifusa (804 – 872). Only a week after Saga’s death in the seventh month of 842, the incumbent crown prince (Tsunesada) was deposed in favor of Yoshifusa’s own nephew in the so-called Jōwa Incident. The new crown prince, the future Emperor Montoku, was Emperor Ninmyō’s son by Yoshifusa’s sister, while Tsunesada was the son of Junna (figure 9.1). Prince Tsunesada retired from court politics, eventually taking Buddhist vows and the name Kōjaku in 849, whereas his most influential supporter, Tachibana no Hayanari (? – 842), died on the way to his exile in Izu.13 The importance of this incident can hardly be overstated. The Northern Fujiwara now saw themselves in an unprecedented position of power, while Junna’s descendants were eliminated from imperial succession. As indicated by Tsunesada’s retirement into monastic orders, the religious world offered a decent retreat for those who had been eliminated in court competition. Notably, the aforementioned Emperor Heizei took Buddhist vows in 810 following his failed comeback attempt, and he lived as a monk until his death in 824, whereas his accomplices were executed or committed suicide. In addition, his third son, Prince Takaoka (799 – 865), who was the crown prince at the time of the Kusuko Incident, lost that designation and shortly thereafter retired to become a monk, taking the name Shinnyo.14 There was thus a distinct pattern for disillusioned and outmaneuvered competitors at court to retire into monkhood. This trend is even more apparent following the Jōwa Incident. When Tsunesada took Buddhist vows, he chose as his teacher none other than Shinnyo (Prince Takaoka), who had become one of the most prominent aristocratic monks of the time. Furthermore, Princess Seishi, the mother of Tsunesada, became a nun in 844 and began sponsoring Buddhist rituals for her late father (Saga) and husband, Junna, who had died in 840. Most of the ceremonies took place at Daikakuji, a temple that was originally founded by Junna at the location of Saga’s detached palace. Seishi rebuilt the temple with material from Junna’s own mansion in 864, although it was her son Tsunesada (named Kōjaku as a monk), who was recognized as the temple’s founder. In the early tenth century, Daikakuji became part of a larger complex, Ninnaji, under Emperor Uda’s sponsorship.15 Although the retirement of Yoshimine no Munesada (816 – 890) in 850 was
216 | mikael adolphson
Figure 9.1. Imperial genealogy of the early Heian age
similarly related to the Jōwa Incident, its effect on court politics was far different. Munesada was a secularized grandson of Kanmu, and thus also a paternal cousin of Prince Tsunesada. He was particularly close to Junna, and favored by Ninmyō as well. Following Emperor Ninmyō’s death in the third month of 850, Munesada suddenly left the court, supposedly out of grief for the loss of his patron. However, political reasons may also have played a large part in this decision, since he was part of the faction opposing the Northern Fujiwara in the Jōwa Incident. Although not directly involved himself, a fact that allowed him to continue his services under Ninmyō, he was bitterly opposed to Emperor Montoku, the nephew of Yoshifusa. Thus, in what was probably an act of protest, Munesada entered the monastic order of Tendai and became a disciple of Ennin. The prominent Japanese historian Tsuji Zennosuke argued that the choice of Tendai was above all politically motivated inasmuch as the Northern Fujiwara were more noted as patrons of Shingon and the Hossō schools. However, it is also important to take into account Ennin’s rising popularity at court (he returned from China in 847) and a tie of friendship that appears to have existed between the two before Munesada became a monk. In fact, Munesada’s brother was the captain of the ship that took Ennin to China in 838. In either case, Munesada took the name Henjō and came to enjoy a more prosperous career as a monk than he had as a courtier, in part because he continued to reside in Kyoto close to the political action. Contrary to other courtiers who had retired, Henjō did
Institutional Diversity and Religious Integration | 217
not disassociate himself or his temple Gangyōji, founded in the 860s in the hills just east of Kyoto, from the court, but rather attempted to reassert the influence of the Office of Monastic Affairs in appointments of abbots, while limiting their terms to four years to avoid monopolization of abbotships by certain lineages and misappropriations of exempt estates.16 Henjō was, in other words, as active in political matters as he had been as a courtier, promoting strategies that would seem to go against the interests of most members of the ranking clergy. Although untypical for ousted members of the imperial court, his continued involvement in court politics and efforts to reestablish the authority of the Office of Monastic Affairs is evidence of both his political inclinations and the overlapping nature of ninth-century politics and religion. After Ninmyō’s death, his successor and son Montoku, Yoshifusa, and the Shingon monk Shinga (801 – 879) transferred Ninmyō’s residential palace to the Fukakusa area south of Kyoto and made it into a temple, where services were held for the late emperor. Shinga had been Ninmyō’s palace monk, but he was also favored by Yoshifusa, and there can be little doubt that there was a guiding Fujiwara hand behind the establishment of the new temple, Kashōji, in 851. In 859, Shinga was granted funds for three yearly ordinands and for services promoting peace and tranquility in the realm, establishing Kashōji as a goganji (imperially vowed temple).17 Montoku sponsored the construction of a chapel, with the support of Yoshifusa and Shinga, within the same area in 851 to pray for the safe birth of a son. The prayers were apparently effective, because the future Emperor Seiwa (r. 858 – 876) was born soon thereafter, and both Yoshifusa and Montoku subsequently sponsored new buildings under the guidance of Shinga. What was originally known as the western cloister (saiin) of Kashōji eventually came to outgrow the main temple and became independent as Jōganji in 862, as Seiwa’s imperially vowed temple.18 These goganji represent important aspects of both politics and religion in the ninth century. Shingi shiki, an early tenth-century record of ceremonial proceedings, explains that even though these temples may have been founded by individual monks or nobles, they were granted the status of goganji for their services to the emperor, and in extension the state itself.19 They were, in other words, privately funded and promoted, but they assumed an official function because of their services to the ruling members of the imperial family. In reality, though, the goganji served the political ambitions of the Northern Fujiwara as well. It was Ninmyō’s line that the Fujiwara chieftain wanted to promote as the proper successors to the throne, and the importance of designating a temple honoring Ninmyō as imperially vowed was not lost to anyone in mid-ninth-century Kyoto. In a similar fashion, the cloister of Jōshin’in was granted the status of goganji some eleven years after its founding in 846 by Ninmyō within the monastic complex of Enryakuji, with the specific purpose of “protecting the state and pacifying the subjects.” Funding for ten monks through rice taxes from Ōmi Province was assigned, so that they could read sūtras and perform daily services for the state, although they mostly resided in the capital while sending copies of sūtras to the temple for storing.20 Moreover, Montoku endorsed the construction of Sōji’in
218 | mikael adolphson
within Enryakuji in 850. It was designated a special chapel to offer prayers for the tranquility of the realm and the emperor’s own success in his first year as the sovereign of Japan.21 Both Ninmyō and Montoku were accordingly staunch supporters of esoteric rites, as also evidenced by their patronage of Ennin and by an unprecedented large number of kaji kitō (prayers designed to bring benefits to the believer) performed during their reigns.22 By contrast, we find few examples of cloister and temple construction by imperial patrons prior to Ninmyō, underscoring the importance of this spurt of sponsorship in the mid-ninth century. In 858, Yoshifusa’s grandson, Emperor Seiwa, ascended the throne, and the Fujiwara chieftain was granted the prestigious title of regent (sesshō) eight years later. The appointment was precipitated by a factional dispute following the destruction of the Ōtenmon, the main gate to the Imperial Palace, in a mysterious fire. In the end, two families with a long history at the court, the Tomo (Ōtomo) and Ki, were eliminated, further strengthening the position of Yoshifusa and the Northern Fujiwara. After Yoshifusa’s death in 872, his adopted son Moto tsune was no less successful in dominating matters in Kyoto, even though he encountered challenges along the way. After controlling politics under the young Seiwa (who founded Enkakuji for his retreat at an old mountain residence of Mototsune) and Yōzei (r. 877 – 884), he encountered problems under Emperor Kōkō (r. 884 – 887), who was in his fifties at the time of his accession. One of Kōkō’s supporters was the aforementioned Tendai monk Henjō, who was given an opportunity to disrupt Fujiwara dominance by serving the emperor in various administrative tasks. For his efforts, he was awarded additional funding for ordinands and ranking monks at his Gangyōji, as well as 153 chō of land in Ōmi Province.23 He was promoted to the highest-ranking post (sōjō) within the Office of Monastic Affairs in 886 and allowed rare privileges and influence at the imperial court. In a manner reminiscent of the way that Empress Shōtoku promoted Dōkyō, Henjō’s promotion was a means for Kōkō to escape the influence of the Northern Fujiwara. Mototsune’s position improved with the accession of Emperor Uda (r. 887 – 897), who granted the Fujiwara chieftain the title of kanpaku (chancellor). Still, frictions eventually tainted their relationship, as Uda seems to have been determined to control matters at court after he retired. Interestingly, he did so as abbot of Ninnaji, which developed into the prime center for Shingon rituals in the tenth century. Ninnaji was originally established as a religious retreat in the northern part of the capital for Emperor Kōkō, but was transformed into a much more substantial Shingon center by Uda, who took Buddhist vows in 899. Headed by members of the imperial family for most of the Heian age, Ninnaji accumulated wealth in the form of land donations and religious prestige through its possession of important Shingon treasures. Following the same pattern, Emperor Daigo (r. 897 – 930) patronized Daigoji, southeast of Kyoto, although he did so while remaining emperor until his death. Daigoji was first built in 874 as a chapel by a Shingon monk named Shōbō (832 – 909), but was transformed into a goganji in 907, leading to the construction of several important buildings within the expanding area. In addition to being an imperially vowed temple, Daigoji was also
Institutional Diversity and Religious Integration | 219
granted the status of jōgakuji (certified temple; see below), undoubtedly to secure more funding, in 913. As Montoku had done at Jōganji, Daigo sponsored prayers for the birth of healthy sons at his temple in an effort to retain his line on the imperial throne. As can be expected, both of Daigo’s sons, who succeeded him consecutively as Suzaku (r. 930 – 946) and Murakami (r. 946 – 967), continued the lavish imperial support of Daigoji, and all three sovereigns were eventually buried close to the temple.24 The esoteric schools broadened the spectrum of the Buddhist discourse, but they did not lessen the political importance of monks and temples. Rather, as the usages of ceremonies and prayers became more specific, the interdependence of the two spheres increased. Whereas the traditional ceremonies of the Nara schools were geared toward peace and prosperity for the state, esoteric rites were additionally used for more personal purposes, such as the safe delivery of a child, to cure a prominent member of the family, or simply to protect the family against particular enemies. Furthermore, the political factionalism that characterized the Heian court from the outset was prominent in the religious sector as well. Smaller cloisters became increasingly important in intra- and interfamily competition. Prayers for a family ancestor, for example, were not only a religious service for the deceased; they also served to strengthen that particular line’s position. If the cloister in question additionally received official recognition and funding, the family’s fortunes would naturally be further enhanced. This overview has been confined to temples founded by the leaders of the imperial state, reflecting the central political setting that made up the framework within which a number of new temples were founded in the capital region. But the sources also reveal that individual monks, lower-ranking nobles, and regional gentry beyond the central sphere of the imperial court founded temples. To further delve into the relationship between centers and peripheries in a politico- religious context of the early Heian state, it is pertinent to explore the circumstances behind the founding, patronage, and official position of such temples in a broader spatial context.
Private and Official Temples in the Early Heian State Private cloisters and chapels established during the ninth century created a paradox in the imperial bureaucratic state. Whereas private Buddhist temples were established outside the parameters of the bureaucratic state and thus not recognized by it, the leading courtier families were dependent on them in sacralizing the succession of the imperial family and the position of their own clan or lineage. By the same token, the religious world was an indispensable alternative for outmatched nobles and ranking members of the court who were no longer active, for a variety of reasons, in the secular affairs of the imperial court. Adding to this central patronage, local gentry sponsored temples in the provinces, benefiting in similar ways. Official recognition of these regional temples served as a means for the provincial gentry to create a connection with the imperial court while reinforcing their own status in the local community. There were also
220 | mikael adolphson
economic advantages; by retaining control of the temple and its administration, local patrons could pocket yields from exempt rice fields and taxes levied for the temple.25 For the imperial court, the reading of sūtras, in particular those that sanctified the state, established and confirmed the presence of central authority in the provinces. The granting of public recognition of temples privately funded by regional nobles seems to have been a small price to pay for such a crucial tie with local powers. The larger problem for the political center was how to control and incorporate religious outposts within the parameters of the imperial bureaucratic state. Originally, the state powers constructed and promoted particular temples, culminating with the completion of the Great Buddha statue and Tōdaiji in the mideighth century, while assigning an official status of government temples (kanji) to temples in its network of Buddhist institutions. As state-sponsored institutions, such temples commonly followed certain architectural patterns with a prescribed number of buildings, including a “golden hall” (kondō), main lecture hall (kōdō), dining hall, pagoda, and sūtra depository, with public support for repairs and maintenance.26 But there were no official designations for those established outside these parameters, and as individual chapels and temples were founded at an increasing rate from the mid-ninth century, the imperial court attempted to find ways to interrelate them. For instance, the Engi shiki, a set of ritual proceedings of the early tenth century, refers to several categories of officially recognized temples, although not all temples mentioned were given specific designations. First, and at the very top, we find great temples (daiji), which increased from two or three at the time of Emperor Tenmu to fifteen by the mid-tenth century. Originally reserved for temples sponsored by emperors in the name of the imperial state, such as Daianji, Yakushiji, and Tōdaiji, this category also came to include Kōfukuji. The amount of land the daiji were granted varied, but each possessed substantial holdings for their sustenance, ranging from 500 to 10,000 chō. Second, there were provincial temples (kokubunji) in the temple network system established during the Nara age. Although an important rank at the time, they became less and less meaningful as outposts of the state because the coordinated system Emperor Shōmu had envisioned never fully materialized. Still, the court sponsored rituals at provincial temples in the early Heian age, and several important regional temples later obtained this status in recognition of their central role in their respective province. The third group consists of temples designated as certified temples, of which the Engi proceedings lists a total of forty-six. Although founded outside the parameters of the bureaucratic state, these temples obtained official recognition and funding in return for their performing religious services for the benefit of the imperial state.27 A fourth category is the imperially vowed temples, which were crucial in establishing an elite layer of temples established and supported by the various imperial and noble factions. The goganji, situated as they were almost exclusively in the capital region, were, in other words, an important element of politics at the center. By contrast, jōgakuji can be found across Japan, and, as such, more directly address the issue of central and peripheral trends in the religious policies of the early Heian age.
Institutional Diversity and Religious Integration | 221
With origins in China’s Sui and Tang dynasties, the jōgakuji was a means for the central authorities to restrict the number of private temples in the realm. Although one can infer that Tenmu’s edict of 680, aiming to limit the construction of private temples, suggests an early usage of this designation in Japan, the term itself first appears in Shoku Nihongi in an entry for 749. No specific temples are listed by name, but in an attempt to place an upper limit for jōgakuji assets, an edict was issued stating that the land possessions of jōgakuji were limited to 100 chō of rice fields for their sustenance. By the same token, it also indicates that such temples amassed estates, expanded their land holdings, and continued to enrich themselves, resulting in repeated edicts from the imperial court.28 By the ninth century, smaller temples and halls competed to be designated jōgakuji, but it is not surprising to find that the popular schools of Tendai and Shingon were the most successful. Still, there was a due process to becoming a jōgakuji. As already noted, it was originally a means of incorporating already existing temples into the religious hierarchy and assigning them assets for their sustenance. Thus, the sponsor and the abbot needed to show a need to qualify for funding while at the same time demonstrating the temple’s importance to the imperial state itself. For example, Shinganji, a temple originally sponsored and built by Wake no Kiyomaro (733 – 799), obtained the status of jōgakuji in the early 800s to perform prayers in support of the realm. However, the funding appeared inadequate, and the temple declined. It was revived when Kūkai made it a Shingon center, and it was subsequently renamed Takaodera. In 824, it was regranted the status of jōgakuji as a Shingon monastery and supplied with funding for seventeen monks.29 In another case, a Shingon monk named Shinshō (n.d.) founded a small chapel in the Otagi district of Yamashiro Province. However, the monk soon found it difficult to maintain the compound in the mountainous area despite its relative proximity to the capital, and instead bought a piece of land formerly owned by a noble in the eastern hills of Kyoto, where he rebuilt his private chapel. He admitted to breaking the imperial codes by building his own private chapel and to violating a rule that prohibited monks from buying property from nobles. Still, Shinshō appealed to the imperial court in 859, stating that his private enterprise was in fact in accordance with the wishes of the late Emperor Ninmyō for the protection of the realm. His wishes were granted, and the chapel was named Zenrinji, receiving the status of jōgakuji.30 Jōgakuji in the capital region were typically founded by a noble or a monk with central connections, through which their temples obtained official recognition. Although the largest concentration can be found in the Kinai, members of the local gentry also founded a substantial number of jōgakuji in the provinces. In fact, of the eighty jōgakuji that are known today, half were outside of the capital area, with six as far away as Dewa Province in northeastern Japan (table 9.1).31 For the imperial court, the jōgakuji served an important purpose both in the political center and peripheries by incorporating privately founded temples into the state structures and extending the state’s influence through the use of state funding. The increased public role of these temples is perhaps most telling in the performance of Buddhist rituals ordered by and performed for the state, which
222 | mikael adolphson
increased from the 830s. In addition, these local temples served as the gateway for the central elites to local shrines, thus creating a concrete link between court, certified temples in the districts, and village shrines.32 The character of the jōgakuji bespeaks a system designed to solve a contradiction in the bureaucratic state. Founded as these temples were on private initiative, their administration was originally under the management of the benefactor (dan’otsu). However, a desire to secure steady funding and be acknowledged in the religious state hierarchy induced the benefactors or the abbot to seek the status of jōgakuji. Although the benefits varied somewhat depending on location, a few common characteristics can be noted. First, jōgakuji were allowed tax-exempt land, as already noted. Anshōji in Yamashiro Province, for example, had paddy and dry-land fields of about 100 chō in several provinces close to the capital, according to a land record of 871.33 Second, jōgakuji received funding for repairs and maintenance. They were permitted to act as “rice banks,” lending out rice plants supplied by the court to farmers in exchange for interest, which was used to purchase material for Buddhist services, in particular oil, and for the general maintenance of the temple. In central Japan, the government could grant stipends of five hundred or a thousand rice seedlings from public storages to kokubunji and jōgakuji, although in reality the amounts varied greatly. By contrast, jōgakuji outside of the capital region were not allowed to use government funds, so the plants were frequently provided from private fields, by local powers or the benefactor. Such strategies were nothing short of “rice laundering” since the rice was ostensibly donated to the temple, but the interest and the principal would eventually return to the lender’s pockets. In either case, the lending business was undoubtedly a popular practice, as it could offer hefty profits for the temples and their benefactors. Indeed, it was also a privilege granted to Tōdaiji, Enryakuji, and Tōnomine in the mid-Heian period.34 Third, jōgakuji were allowed public support for a set number of monks as well as yearly ordinands. For example, Takaodera was granted funding for fourteen monks in 824, when it joined the ranks of certified temples, and Jōganji was allowed sixteen, while Daigoji received funding for ten. Although not all jōgakuji received such funding, it was undoubtedly part of the package that monks and nobles submitting appeals were hoping for.35 Temples such as Kashōji, Anshōji, and Gangyōji, for instance, had no specific funding for monks, but were granted three new ordinands every year.36 Records of temples that simultaneously held other designations further reinforce the primary role of the jōgakuji as a means to secure funding. For example, Jōganji, as noted, was originally a small hall for prayers, established by Shinga under the patronage of Fujiwara no Yoshifusa to pray for the birth of a son for Emperor Montoku. Following the accession of Emperor Seiwa in 858, for whom the prayers had been designated, it became an imperially vowed temple, although the administration remained with the Fujiwara chieftain. However, it was also designated jōgakuji in 874, only two years after it became independent from Kashōji, thereby indicating that imperially vowed and certified temples had different functions. Ninnaji, Gangyōji, and Fujiwara no Tadahira’s Hosshōji (built in 925) are examples of other high-ranking temples that similarly earned double
Table 9.1. Jōgakuji Designated
Jōgakuji
Province
Early 800s 820 824 828 828 828 833 835 836 837 842 843 847 847 851 855 855 856 857 859 860 862 863 863 863 863 864 865 866 866 866 866 866 866 866 866 867 867 867 869 860s?
Shinganji Kokushōji Takaodera (Shinganji) Bodaiji Shōsuiji Mirokuji Kinshōji Kongōbuji (Jōgakuji) > Hokkeji Kashima jingūji Jūeiji Daikōji Hasedera Tsubozakadera Kaiinji Anshōji Daikōji Hōryūji Kyokurakuji Byōdōji Kibi jingūji Hōonji Hōshōji Butsunanji Tōaji ? Zenrinji Ensanji Kannonji Jakkōji Myōrakuji Nishikiodera An’yōji Yadaiji Yugadera Sōōji Reizanji Chōanji Myōhōji Saishōji Kanshinji Katsuodera
Yamashiro Ōmi (kokubunji) Yamashiro Mino Higo Iyo Ōmi Kii Izu Hitachi Tajima Noto (kokubunji) Yamato Tajima Yamashiro Yamashiro Izu Dewa Mutsu Yamashiro Echizen Yamashiro Suruga Tanba Tōtōmi Yamashiro Mino Dewa Shinano Shinano Shinano Shinano Shinano Dewa Yamashiro (kokubunji) Dewa Dewa Ōmi Ōmi Kawachi Settsu
Table 9.1. (Continued) Designated
Jōgakuji
Province
870 870 872 873 874 876 877 878 878 881 884 885 885 887 889 890 890 896 899 905 905 913 918 900s 934 935 949 969 986 990
Kankūji Anryūji Shōrinji Kangaji Shittanji Daikakuji Gangyōji Kashōji Gokokuji Enkakuji Gankōji Mirokudera Jinnōdera (Jōgaku amadera) Enseiji Butsuryūji Nanmuji Jōfukuji Kyokurakuji Tōkōji Kanshūji Daigoji Makiodera Kagonji Hosshōji Kangyōji Muiji Saga tōin Enryakuji Eshin’in Enryakuji Myōkain
Yamashiro Dewa Owari Sagami (kokubunji) (881) Settsu Yamashiro Yamashiro Yamashiro Ōmi Yamashiro Owari (kokubunji) Kaga Yamashiro Mino (kokubunji) Yamashiro Yamato Ōmi Yamashiro Yamashiro Yamashiro Yamashiro Yamashiro Izumi Mino Yamashiro Yamashiro Yamashiro Yamashiro Ōmi Ōmi
Sources: A majority of the jōgakuji in the list can be found in Ruijū kokushi, Sandai jitsuroku, Nihon kiryaku, and Shoku Nihon kōki. The works by Takeuchi (Ritsuryō sei to kizoku seiken) and Hayami (Jōgakuji no kenkyū), and to some extent Usami (Jōgakuji no seiritsu to henshitsu), contain lists of jōgakuji that were helpful to me in my own compilation.
Institutional Diversity and Religious Integration | 225
honors.37 There are moreover records of at least six jōgakuji (Sōōji in Yamashiro, Kokushōji in Ōmi, Daikōji in Noto, Gankōji in Owari, Kangaji in Sagami, and an unknown nunnery in Mino) that were additionally designated kokubunji, and a few cases where jōgakuji became official temples (kanji), such as in the case of Myōhōji and Saishōji in Ōmi Province.38 Supported by the local gentry, regional jōgakuji were frequently better maintained than kokubunji, and some came to take the place of the state-supported provincial temples that had been allowed to decline. This trend is evident already in a court edict of 809: “The Daihannya Sūtra shall be copied and sent out to the kokubunji. However, if there are no kokubunji, then they should be kept at a jōgakuji.”39 In 820, the governor of Ōmi Province asked and received permission to have Kokushōji designated kokubunji to replace Konkōmyōji, which was destroyed in a fire in 785. It is unclear how state rituals were performed in Ōmi in the thirty-five years before Kokushōji was elevated to kokubunji, although it is likely that the designation in 820 was merely a post facto confirmation of existing conditions.40 In this way, the jōgakuji performed services for the state and could become converted into fully official temples, despite being founded and patronized privately by ranking monks and nobles. These complex official temple designations must be understood in the setting of ninth-century Japan. As the court was unable, or unwilling, to reconstruct and maintain a network of provincial temples, it turned to those that had survived and likely prospered under private management. The jōgakuji was a means out of necessity to officially acknowledge such temples and to bring them into the umbrella of the imperial state in exchange for financial support. The imperial court gained a religious and ideological presence in the provinces, but the selection of jōgakuji was clearly not random. For example, in 866, the very same year that the Ōtenmon Incident occurred, five temples (Jakkōji, Myōrakuji, Nishikiodera, An’yōji, Yadaiji) in Shinano Province were granted the designation of jōgakuji. The sources do not reveal a direct connection between the incident in the capital, which was a factional dispute centering on blame for the fire that destroyed the main imperial gate in the south, and the granting of certified temple status for the Shinano temples. However, Fujiwara no Yoshifusa, who undoubtedly gained the most from the incident, was likely behind the creation of the new jōgakuji just a month earlier: both developments were a result of the Fujiwara chieftain’s political strategies. The historian Ushiyama Yoshiyuki has convincingly argued that the granting of jōgakuji status was an effective means for Yoshifusa to recruit support from the provincial gentry and gain more control of land in Shinano.41 It was the local notables who constructed private temples in order to reclaim new paddies, and by having them recognized as certified temples, they were in a position to exempt them from taxation and use them for rice loans. Indeed, given that six jōgakuji were also designated in Dewa Province during Yoshifusa’s era, we can conclude that they were used by the court leaders to tighten the tie with various districts. Significantly, Yoshifusa’s strategies represent an effort to increase the center’s presence, not by extending the bureaucratic system of the
226 | mikael adolphson
Nara age, but by creating a network based on considerably more private connections in response to local conditions. Since the state assumed responsibility for the financial well-being of the jōga kuji, it only seems natural that it wanted a say in the administration of its assets. The solution was an impractical compromise, according to which “the assets of the jōgakuji of the various provinces shall be administered by the provincial governor in cooperation with the temple’s benefactor and the temple management [sangō].”42 At first, governors were required to submit asset accounts every year to the imperial court, but from 825, reports were expected only once every six years — that is, after each appointment of a new governor. In 868, the interval was cut down to four years when the appointment period for governors was similarly adjusted.43 Accordingly, the jōgakuji might best be described as semiprivate and semiofficial temples with private patronage and support but also having public roles and responsibilities. As can be expected, however, this arrangement worked poorly in practice. For example, Engi shiki states that the benefactor should initiate the selection process of a new abbot by sending a letter of recommendation to the local official (gunji), who would deliver it to the provincial head monk (kōdokushi). He, in turn, took the letter to the provincial governor, who submitted it to the Grand Council (Daijōkan) for a formal appointment.44 Yet, in actuality, most appointments were made by the benefactor himself, or restricted to monks from a certain lineage, a practice that was frequently acknowledged by the imperial court. For example, it was decided by imperial decree in 878 that appointments of abbots for the aforementioned Kashōji were to “follow the example of Jōganji, without permission or interference from the Office of Monastic Affairs.”45 In Daigoji’s case, a document dating to 914, in which the temple was granted jōgakuji status, stipulates, When it comes to the appointment of the abbot, a qualified person should be selected and appointed from the lineage of [the founder] Grand Master Shōbō. There will be no restrictions [in terms of his rank], and he does not have to be a member of the Office of Monastic Affairs or a provincial head monk, as long as he belongs to that lineage. Nobody from outside this lineage shall ever be appointed abbot of this temple.46 Another problematic area was that of financial responsibility. Benefactors were interested in increasing their revenues and land holdings through their jōgakuji, but they were much less willing to shoulder the burdens for the upkeep of the temple itself. In fact, by 806 there were reports that benefactors abandoned their temples, and edicts imploring the benefactors to maintain their temples were repeatedly issued from the imperial court.47 In the end, the burdens of funding and repairs were regularly left to the provincial officials, who nevertheless had little influence over the temple. By the end of the tenth century, jōgakuji had become an outdated category, and in 990 Enryakuji’s Myōkain was the last cloister to earn that status.48 In particular, the joint management of taxes and assets proved to be an impractical arrangement, especially from the tenth century, when many provincial governors began
Institutional Diversity and Religious Integration | 227
to use their term-limited offices as a means to income and the imperial court’s bureaucratic apparatus in itself proved insufficient to control the countryside. And, most important, the state was no longer able to provide the funds that came with the status. In fact, Engi shiki lists only fifteen of the eighty jōgakuji that we know of from historical records by the time of its compilation, indicating that the designation was not an effective means for control or funding.49 In short, jōgakuji strengthened the financial foundation of a selected number of religious institutions in geographical and religious peripheries, but not the political presence of the central authorities. The timing of this change is far from coincidental. The early tenth century presented the Kyoto elites with challenges of a ferocity not encountered before. In the 930s, the uprising by Taira no Masakado and the Fujiwara no Sumitomo piracy in the Inland Sea rocked the court. Religious developments in the provinces were also unsettling. In Miyoshi Kiyoyuki’s famous memorandum of 914, we learn of private chapels that were erected in the provinces and of locals who took Buddhist vows to escape taxation.50 At the same time, governors appointed to the provinces frequently abused their powers by trying to enrich themselves during their brief four-year terms, as evidenced by the appeal from the farmers of Owari Province treated by Charlotte von Verschuer in this volume. As the jōgakuji themselves prove, the court had made adjustments in its organization and handling of government matters since the move to Kyoto, but those made in the tenth century, spearheaded in particular by the Northern Fujiwara, appear far more substantial and far-reaching. In short, finding the bureaucratic structures unable to sustain the control of and a steady flow of income from the provinces, the ruling elites came to use their own private assets in managing the business of the state. The Northern Fujiwara’s “teeth and claws” — warrior retainers of the Minamoto clan — for example, became indispensable both to the Fujiwara themselves and for the peace of the realm.51 The creation of shōen, private estates, which began to gain momentum in the late tenth century, is another aspect of this trend. Although the state structures thus failed to sustain the centrality of the realm, the elites themselves successfully utilized their social, economic, and political powers to maintain, if not strengthen, the center’s control of the geographical and political peripheries. It is as part of this process and these adjustments that regional and central temples became less dependent on official designations.
Betsuin: Affiliating with the Religious Centers The politico-religious setting of the early Heian age, spearheaded by noble patronage of various Buddhist schools and ceremonies, facilitated the establishment of new monastic centers without prohibitive state intervention. The flip side of this freedom was that funding could be precarious, and it became difficult for a monk from a small temple without affiliation to a larger complex to compete for the most coveted titles, even if he was of noble origin. Only the most prominent temples, such as the goganji, whose importance increased from the late tenth
228 | mikael adolphson
century as the jōgakuji designation ceased to fulfill its function, and those with the support of leading courtiers and aristocratic monks could secure financial and political support to sustain themselves in the long run. Still, there was a separate track available for new chapels and smaller temples through the creation of ties with more influential temples as betsuin (detached cloisters). Although sources pertaining to its beginnings are scarce and scattered, it appears that betsuin originally indicated a separate location, as in the use of the term bessho (separate place), not necessarily implying a distance from the main complex to which it belonged. A document of 798 refers to the Amida betsuin within Tōdaiji, and a record from the late twelfth century lists twelve cloisters within Tōnomine as betsuin.52 In another case, a chapel built in Yasaka village in Yamashiro Province became a betsuin of Yasakadera by virtue of its proximity to the larger complex in the 830s. It remained a separate cloister, but was for all practical purposes part of Yasakadera, as its popular name Yasaka Tōin (Eastern Cloister of Yasaka) indicates.53 Thus, the term originally denoted cloisters and chapels that were founded by individuals while not being part of the standardized architectural layout as envisioned by the eighth-century court leaders. Betsuin gained momentum in the ninth century, and we find, as in the case of jōgakuji, a dramatic increase from the reign of Emperor Ninmyō. Assuming that extant sources somewhat accurately represent historical circumstances, we note that the schools of Tendai and Shingon dominate (table 9.2), reflecting the popularity of esoteric rites and thought. At least five Shingon betsuin were created between 836 and 860, and Enryakuji established ties with no fewer than twelve temples between 839 and 881 in provinces such as Ise, Harima, Ueno, Kaga, Shinano, and Mutsu. Moreover, Mudōji, which later became a prominent section of the Enryakuji complex, first achieved recognition when it became a Tendai betsuin in 866 through the request of the monk Sōō (n.d.) and the support of Fujiwara no Mototsune.54 Mudōji itself became the patron institution for Tōnomine later. Another substantial spurt in the creation of Tendai affiliations took place during the late tenth century, beginning with the successful leadership of head abbot Ryōgen (912 – 985).55 In explaining the emergence and success of betsuin, comparisons with jōgakuji are unavoidable. They were both a response to limitations of the bureaucratic system, which had little room for nonsanctioned temples, and to what can best be called an overestablishment of private temples. Furthermore, their origins are often very similar, since the earliest betsuin were normally small chapels (dōjō) or temples established by noble patrons, individual monks, or local gentry. For example, a local gentry family originally patronized Saiganji in Yamashiro Province before it was converted into a cloister of Enryakuji’s Myōkain.56 To mention one more example, a capital noble named Kamishigeno no Sadanushi (n.d.) built a private chapel on his estate in southern Kyoto. As it bordered on Saiji, he appealed to have it recognized as a detached cloister, renaming it Jion’in. Still, he retained his family’s rights over the temple and the estate by stipulating the benefactor’s control over future abbot appointments.57 Another common pattern in the founding of new betsuin was that of monks sent to the provinces as lecturers, sponsoring
Table 9.2. Betsuin First reference
Betsuin
Honji
Province
780 798 836 837 839 839 840 840 840 840 844 849 849 855 860 863 866 867 878 881 881 882 882 884 931 935 937 945 post-950 post-950 956 959 963 974 990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990
Ryūenji Amida betsuin Gakuanji Yasaka Tōin Tado Jingū Takaosanji Daidōji Seimyōji Kannonji Jōgakuji Jion’in Tado Jingū Hōunji Seiryūji Daikōji Ninchōji Shittanji Onjōji Onjōji Shikanji Kannonji Kōryūji Enshōji Mudōji Unrin’in Tōanji Shōosanji? Kitain Fudarakudera ? ? Myōrakuji Mirokudera Rokuhara Mitsuji Kanjin’in (Gion) Hōkōin Saiganji Rakuonji Fugenji Hōrakuji Jōon’in Shūjōji
Saidaiji Tōdaiji Shingon Yasakadera Tendai Shingon Tendai Tendai Tendai Tendai Saiji Shingon Enryakuji Kaiinji Shingon Tendai Tendai Enryakuji Tendai Tendai Tendai Tendai Tendai Tendai (Gangyōji) Daigoji Sōjiji Tendai Mudōji Tendai Tendai Mudōji Ninnaji Tendai Tendai Tendai Myōkain Myōkain Myōkain Myōkain Myōkain Jōon’in
Yamato Within Tōdaiji Yamashiro Yamashiro Ise Kaga Harima Harima Harima Iyo Yamashiro Ise Ueno Izu Settsu Settsu Ōmi Ōmi Kaga, Ishikawa-gun Shinano Mutsu Ōmi Ōmi Yamashiro Ise ? Echizen Yamashiro Dewa Shinano Yamato Kinai Yamashiro Yamashiro Yamashiro Yamashiro Yamashiro Yamashiro Yamashiro Yamashiro Yamashiro
230 | mikael adolphson
Table 9.2. (Continued) First reference
Betsuin
Honji
Province
Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-990 Post-991 Post-991 Post-998 Post-1005 1080 1080 1080 Post-1086 1101
Hōkaiji Sōji Kūtadera (?) Utōji Hōzōji Hōdōji Jōrinji Tokuryūji Gannōji Kawaidera Gakuzenji Sanshiji Fumimurodera Hōonji Jōgakuji Sanmaiin Kannon’in Shūgakuin Daiunji Gedatsudera Hōjōji Myōōin Ichijōji Entokuin Sūkeiji
Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Myōkain Miidera (Onjōji) Miidera Miidera Miidera Miidera Miidera Tendai Tōdaiji
Ōmi Ōmi Ōmi Ōmi Ōmi Ōmi Ōmi Ōmi Tanba Tajima Tajima Harima Echizen Kii Kii Iyo Yamashiro Yamashiro Yamashiro Ōmi Kinai? Ōmi Ise Ōmi Kii
Sources: In compiling this table, I have above all relied on Shoku Nihon kōki, Sandai jitsuroku, Mon’yōki, Sō Myōtatsu soseki chūki, and Tōnomine ryakki. I have also benefited from the works of Takagi and of Toyoda, and from Kageyama Haruki’s Hieizan. Still, the table represents only the betsuin that I have been able to locate and to some extent identify, and should not be considered exhaustive.
chapels with a focus on the doctrine of their choice in the local community. In other cases, betsuin were established as small monastic retreats by retiring monks or nobles. However, many experienced problems as soon as the founder was no longer in this world. Even if small grants or land had been arranged, there were no guarantees that these assets would remain. The imperial court could revoke tax-exempt land, and did so on occasion if it found that too much was escaping its taxation, as indicated by edicts in the tenth century. More serious, however, was the abuse that provincial governors frequently engaged in. As noted earlier, governorships were limited to four years for most of the Heian age (from 868), and
Institutional Diversity and Religious Integration | 231
many nobles viewed them as opportunities for personal enrichment, especially since they rarely had a strong tie to the province. Thus, local temples were under heavy pressure not only to sustain themselves but also to defend their assets. In such circumstances, the doctrinal focus of the founding monk could be used as a foundation for linking the smaller temples to influential monasteries, a link that would secure some measure of official status for the detached cloister. The initiative in transforming a temple into a betsuin lay with the founder or a subsequent abbot or sponsor. For example, Hōkōin, sponsored by Fujiwara no Kaneie (929 – 990), was made into a Tendai betsuin through a donation, as were Myōkain’s Hōkaiji, Saiganji, and Rakuonji.58 Still, it was not a tie that could be established at will, at least not in the ninth century, since each affiliation or donation also required the permission of the imperial court. Both Kaga’s Shikanji and Ōmi’s Enshōji became Tendai betsuin “in accordance with an imperial order.”59 In addition, Enryakuji’s Mudōji section was granted the same status through the recommendation of Fujiwara no Mototsune, and the aforementioned Henjō was behind Unrin’in’s transformation into a detached cloister of his own Gangyōji. The official character of such transformations is further underscored by the usage of governmental edicts (daijōkanpu).60 Since the government retained the right to allow betsuin status, it also had the power to disallow it, even if only one such case is known. Ise’s Tado jingūji was made into a Tendai cloister in the first month of 839, only to have that affiliation revoked in the twelfth month the following year. In 849, it became a Shingon betsuin through the efforts of the monk Jūchū (n.d.), when he dedicated it to prayers for the imperial state.61 Although the affiliation between the detached cloisters and the main institution was essentially a private one, the status of betsuin itself thus required the acknowledgment of the state, bearing further evidence of the court’s desire to control the designation and status of temples in the ninth century. At the same time, it was not a designation that necessarily excluded others. Indeed, we know of several instances where jōgakuji were also detached cloisters. The aforementioned Enshōji was granted jōgakuji status in 866, and became a Tendai detached cloister in 882, and there is one Tendai betsuin in Iyo that is simply known as Jōgakuji in the records. In the case of Izu’s Daikōji, it became both a jōgakuji and a betsuin of Kaiinji in 855.62 These cases require a more thorough discussion of the role and character of the various designations. Did jōgakuji additionally become betsuin to protect their assets? Did they serve different purposes? Although such double honors may indicate a weakness in the jōgakuji designation, it is more important to recognize the areas where they did not overlap. In other words, affiliated cloisters gained above all an important tie with a leading religious center, making it possible for its monks to maintain an exchange of learning and learned monks, in addition to the security of being attached to one of the more prominent temples close to the capital. Moreover, whereas the jōgakuji represent the diversity of teachings and institutions that were gradually allowed in early Heian Japan, the betsuin seem to represent a trend toward integration. This difference is further reinforced by the existence of third-level betsuin, as evidenced by all of Myōrakuji’s (Tōnomine’s) detached cloisters (third level). Myōrakuji (second level) was in turn a betsuin
232 | mikael adolphson
of Mudōji (first level), which, as already noted, had the same status toward Enryakuji.63 Such hierarchical structures can only point to the importance of religious and lineage affiliations, and a trend toward religious integration. Considered from a more political perspective, however, the establishment of betsuin also represents a weakening of the imperial bureaucratic state, since detached cloisters were an admission of temples established outside the official central structures. This admission is indicated in a petition from the monk Enjun cited in the chronicle Nihon sandai jitsuroku: The sixteenth year of the Jōgan era [874], twelfth month, twenty-fifth day. Shittanji in Shimanokami District, Settsu Province became an official temple (kanji). Previously, Master Transmitter of the Law (dentō daishi) Enjun appealed: “Master Transmitter of the Law Anpi built this chapel for the protection of the state, and in the fifth year of the Jōgan era [863], he received an imperial edict allowing it to become a Tendai betsuin. I now wish to have the temple converted into an official temple (aratamete kanji ni nasan koto wo).” The request was granted.64 It is unclear whether Shittanji lost its Tendai affiliation in becoming an official temple, but the petition nevertheless illustrates the distinctively nonofficial status of the betsuin. In contrast to the more hybrid character of the jōgakuji, the betsuin were fully private affiliations of their patron institutions. Although Enjun aimed to increase the status of Shittanji through an official designation, his request may also reflect a desire to erase the subordinate status of betsuin. This private nature of detached cloisters is further demonstrated by the relationship between the main institution and the betsuin. The authority over abbot appointments for detached cloisters generally rested with the main temple, which on occasion dispatched its own monks to head the betsuin. For example, when Unrin’in was made into a detached cloister of Gangyōji, Henjō, the founder and abbot of the main institution, stated that monks from his own lineage should manage the cloister.65 And in the early tenth century, the Tendai head abbot Zōmei (843 – 927) appointed his disciple abbot of a Tendai betsuin in Dewa Province.66 The main institution might additionally control appointments of members of the detached cloister’s main administrative organ (the sangō), as was the case at Saiji’s Jion’in in Yamashiro (844) and Gangyōji’s Unrin’in. Some betsuin were even listed in the main temple’s records of assets, indicating a higher level of control.67 By contrast, betsuin that maintained more independence in the appointment of its ranking monks did so by virtue of the founder’s or benefactor’s stipulations that the headship should be passed on within a particular lineage, a wish that the main institution normally would respect. In either case, there can be no doubt as to the inequality of status between the two parties. One particularly telling case involves Onjōji, which is noted as a betsuin of Enryakuji in the 860s. As the Enchin lineage at Onjōji gained strength and status, however, it came to compete with the monks of the Ennin branch for the abbotship of Tendai itself, resulting in a split in the late tenth century. The Enryakuji clergy nevertheless continued to label Onjōji
Institutional Diversity and Religious Integration | 233
a detached cloister in the Heian age, using the designation as rhetoric to prove Onjōji’s inferiority.68 The betsuin proved to be a more successful creation than the jōgakuji. While the former were still being created in the eleventh century, the latter became an outdated category by the mid-tenth. In hindsight, this difference seems easy enough to understand. The jōgakuji was the imperial state’s attempt to maintain some control over an increasing number of private temples and smaller cloisters. As hybrid creations with a vision of joint management by the residing monks, the lay benefactor, and the provincial governor, it was an impractical solution from the outset. And as public administration was increasingly performed through private connections and personnel during the Northern Fujiwara ascendancy, affiliations with large religious centers became a more practical and safe means of support and protection for peripheral temples. Betsuin and jōgakuji worked jointly to dismantle the bureaucratic center in that they both served to weaken the state’s direct control of individual cloisters. Yet the creation of ties between smaller temples and larger institutions closer to the political center reveals a centripetal pull as well, albeit not to the secular leaders of the court, but to the central temples of Kinai. This process continued in a more accelerated pace from the eleventh century, when the elite temples’ control over the religious peripheries reached a new level through the establishment of another form of affiliation.
Toward the Center: Matsuji and the Creation of Temple Networks By the late eleventh century, matsuji, or branch temples, emerged as the predominant form of temple-to-temple affiliation (table 9.3). This development reflects not so much an abrupt change as a continued trend of attaching formerly independent and politically peripheral (that is, without official recognition or a central sponsor) temples to the most influential monasteries in the capital region. Indeed, both the betsuin and matsuji served identical purposes as religious outposts for major temples such as Enryakuji, Kōfukuji, and Tōdaiji. The patron temple (honji) held the same kind of administrative control over the branch, with rights to appoint abbots and members of the main administrative organ, a right that not even the sovereign could overturn. An illuminating example can be found in 1105, when Enryakuji protested against Retired Emperor Shirakawa’s appointment of a personal retainer as abbot for Kamadoyama, a Kyushu branch shrine of Daisenji in Hōki Province, in turn an Enryakuji branch. The clergy argued that Shirakawa was violating the rights of the honji and managed to have the appointment reversed, emphatically confirming Enryakuji’s judicial control of Kamadoyama.69 Although this example dates to the early twelfth century, it reflects a custom that was not new at the time, as indicated by the support it received among courtiers familiar with precedents. A closer examination shows that the patron temple’s control was a notch or two stronger over branch temples than it was over detached cloisters. While the main institution’s authority over betsuin was based on the right to appoint
Table 9.3. Matsuji First reference
Matsuji
Province
Honji
740 780 869 945 974 977 978 978 990s Late 900s Late 900s Late 900s Late 900s Late 900s 1000 1000 1013 1025 Mid-1000s 1058 1072 1074 1075 1084 1086 1091 1093 1093 1094 1097 1101 1101 1101 1101 1102
Yakushiji Ryūenji Onkokuji Kumedera Kanjin’in An’unji Kokugenji Tōnomine Hōkōin Hosshōji Gangyōji Kyokurakuji Kannon’in Jitokuji Zentsūji Mandaraji Ogitadera Kawaidera Kanzeonji Chikurinji Eizanji Kōfukuji Hōunji Hakusan Heisenji Yakushōji Seiganji Kinpusen Daiyōji Jakureiji Santakudera Sukeiji Fukurinji Jōrinji Myōanji Chinkōji
Shinano Yamato Yamato Yamato Yamato Settsu Yamato Yamato Yamashiro Yamashiro Yamashiro Yamashiro Yamashiro Yamashiro Sanuki Sanuki Yamato Yamato Chikuzen Yamato Kii Yamato Ise Echizen Kii Yamashiro Yamato Mikawa ? Chikuzen Ise Omi Yamato? Yamato? Yamashiro
Tōdaiji Saidaiji Tōnomine Tōnomine Enryakuji Ninnaji Tōnomine Enryakuji Tendai Tendai Tendai Tendai Tendai Tendai Tōji Tōji Tōnomine Tōnomine Tōdaiji Kōfukuji Kōfukuji Tōji Tōji Enryakuji Ninnaji Jōganji Kōfukuji Tendai Tōnomine Daigoji Enkōin Tōdaiji Tōji Hōryūju Hōryūju Tōji
Sources: A large number of matsuji were created from the eleventh century, and the ones included here represent those that I have been able to find, identify, and date from before the early twelfth century. Most of the matsuji in the table can be found in the Dai Nihon shiryō, Heian ibun, and Tōnomine ryakki, and many are noted in the works by Toyoda and by Satō Kenji.
Institutional Diversity and Religious Integration | 235
ranking monks at the cloister, its jurisdiction over matsuji was frequently more complete, since the latter were consistently listed in the main temple’s asset rec ords. Furthermore, branches were required to pay annual dues to the patron institution. The matsuji served, in other words, a function similar to that of shōen in the Heian age, and the honji played the same role as the secular honke (patron lineage) did for private estates. They guaranteed the estate or branch an exemption from provincial taxes and protected its assets from intrusions by third parties, in return for yearly dues and judicial rights. The dues were based mostly on rice, but there is evidence that silk, white cloth, paper, and ink were submitted as well.70 Moreover, historical records reveal that monks from the patron temple were dispatched to take over the administration of the branch directly if the dues were not delivered as promised. Supported by their social and religious status, and possibly by their own retainers, these monks often managed to straighten out the branch, although there was naturally an inverted correlation between the success rate and the distance to the main temple.71 The difference in degree of control between the betsuin and the matsuji is further indicated by the substantial number of detached cloisters that were transformed into matsuji to increase the honji’s control over the affiliate. The complicated history of Tadō jingūji, in Ise Province, includes such a transformation. After becoming a detached cloister of Shingon in 849, Shirakawa designated it a matsuji of Tōji in 1075, despite protests from and obstructions by some local powers and Enryakuji monks.72 It is thus tempting to interpret the matsuji as a replacement of the betsuin, but that is problematic. First, betsuin continued to exist throughout the Heian age, and such cloisters were still created by the warrior class in the Kamakura age. Second, there is a periodical overlap between the two types of affiliations in the early Heian age as well. For example, Onkokuji in Yamato Province was transformed into a branch of Tōnomine already in 868, at a time when betsuin ties were the more common form of affiliation.73 Although this record is contained in a later source, other branches established in the late ninth and early tenth centuries confirm the overlapping existence of the two types of subordinate ties. Third, there are examples of temples that were both. Tōnomine was listed as a matsuji of the Enryakuji complex while also being a betsuin of Mudōji, one of the major sections on Mount Hiei.74 Moreover, in a dispute over an estate in 1102, Tōdaiji’s administrative organ refers to Sūkeiji as both betsuin and matsuji in its edicts and appeals.75 Fortunately, this complicated situation can be untangled by looking at the history of Tōnomine. According to Tōnomine ryakki, written in 1197, a Mudōji monk named Genkan began to perform Tendai rituals at Tōnomine in the early tenth century while training his own monks. In 919, Genkan, who later became Tendai head abbot, was allowed to appoint a chief administrator for a temple he erected there. Then, following his death in 956, the appointments of chief administrators for Tōnomine became a Tendai prerogative, effectively transforming Tōnomine, also known as Myōrakuji from about this time, into a Mudōji branch.76 However, as the clan temple of the Fujiwara, Kōfukuji also had claims on Tōnomine, which was in the Hossō center’s home province of Yamato, and attempted to
236 | mikael adolphson
assert its own influence over it, but failed. Mudōji’s own main institution, Enryakuji, stepped in as the main patron, while Tōnomine retained its betsuin tie with Mudōji. Thus, the essential difference between the betsuin and the matsuji was that of the role the main institution played in the affiliate’s administration. Whereas the detached cloister status indicated above all an affiliation of religious lineage, the branch designation was primarily one of political protection and jurisdiction. In general, one can distinguish two patterns through which temples became branches of larger institutions, depending on where the initiative for the establishment can be found. In many cases, the affiliation was created on local initiative through donations, quite the way that shōen were donated to powerful patrons by influential local landholders. For example, Kokugenji was transformed into a branch of Tōnomine by the provincial governor Fujiwara no Kunimitsu (n.d.), who had sponsored the construction of a chapel within that temple. And Ogitadera and Kawaidera in Yamato were both constructed by head administrators of Tōnomine and were subsequently established as branches of that complex. In the case of Jakushōji, it was made into a branch of Tōnomine by one of its abbots late in the eleventh century, about fifty years after its founding. And it is noteworthy that the head administrator of Sakadadera included both an asset record and an official document listing the domain when he donated his temple to Tōnomine, confirming the proprietary nature of the affiliation.77 Although dating to the twelfth century, another informative case involved Enryakuji and Heisenji in Echizen Province. Heisenji was actually Onjōji’s branch, but when that temple’s abbot — Kakushū (n.d.) — performed various “cruel acts,” the Heisenji clergy decided to affiliate their own temple with Enryakuji. The court denied the request in the fourth month of 1147, undoubtedly owing to Kakushū’s connections with influential nobles. However, such ties apparently did not extend beyond the realm of this world because the transfer was granted only a month later, following Kakushū’s death.78 The particular circumstances behind each temple’s founding and location thus played an important role in their transformation to a matsuji, but some general patterns can be distinguished. One common reason for a regional temple to affiliate itself as a branch with one of the elite temples was an inability to maintain the compound and defend its estates against external threats. Mandaraji in Sanuki Province on Shikoku was probably founded already during Kūkai’s time, but had declined considerably by the eleventh century. In 1062, a local monk managed to restore the deserted dwellings while securing rice and dry fields for the temple’s sustenance. However, the provincial government was reluctant to acknowledge the tax exemptions for the fields in question, thereby also making them a target for intrusions by local warrior-managers. To avert these threats, Mandaraji became a branch of the nearby Zentsūji, which in turn was a branch of Tōji.79 Similar problems of maintenance were also behind the transformation of Kiyomizudera in eastern Kyoto into a branch of Kōfukuji. It suffered repeated fires during the eleventh century, and a Kōfukuji monk headed the reconstruction effort in 1094, indicating that this prestigious temple, founded by the re-
Institutional Diversity and Religious Integration | 237
nowned general Sakanoue no Tamuramaro in the late eighth century, had been transformed into a Hossō branch. Indeed, Fujiwara no Munetada (1062 – 1141) noted in his diary in 1106 that “Kiyomizudera is a branch temple of Kōfukuji, and letters of recommendation for appointments of new abbots will accordingly be submitted by the patron temple.”80 The motives appear to have been similar for Kanzeonji in Kyushu, which experienced funding problems in the mid-eleventh century. It was under pressure from local officials regarding its tax-exempt land in 1058, and when lightning caused severe damage in 1064, the temple became vulnerable. Most buildings were in fact reconstructed with the help of the Dazaifu governor by 1066, but the problems with local officials continued. At this point, Tōdaiji dispatched a lecturer to perform ceremonies at Kanzeonji, eventually resulting in a honmatsu (patron-branch) relationship between the two temples.81 Although the process of this transformation is not known, we can surmise that the Kanzeonji monks were willing to make their temple into a Tōdaiji branch to protect it from incursions. Indeed such concerns were behind the creation of a large number of matsuji. To mention one example from the twelfth century, when the need for protection against local intrusions was even greater, Kongōji in Kawachi Province made itself into a branch of Ninnaji “to gather strength” against repeated abuses by provincial warriors.82 The second important pattern of affiliations is cases when the initiative was taken by the patron temple as part of an effort to expand its influence and income. Eizanji, a temple in Yamato Province under the patronage of the Southern Fujiwara, offers an unusually detailed story. By the mid-eleventh century, it experienced problems in protecting its estates. In particular, Higashiya-no-shō, from which the clergy derived most of its sustenance, became increasingly difficult to control as provincial officials attempted to expand their influence in the region. Eventually, in a somewhat twisted development, a Kōfukuji monk named Kahan (n.d.), who was close to Kii governor Shigetsune (n.d.) of the Southern Fujiwara, managed to get his hands on the original estate documents and take possession of the estate. The Southern Fujiwara requested that Kahan return the documents, but he claimed that they had disappeared while he was at Kōfukuji. Shortly thereafter, Eizanji abbots were appointed from among Kōfukuji monks and edicts were issued from the Hossō center to Eizanji, indicating a branch tie. Besides a shift in favor of Kōfukuji, this affiliation also came to benefit the Northern Branch at the expense of the Southern Branch, as many ranking Hossō abbots were of hokke origin.83 The absorption of Eizanji is merely one example of Kōfukuji’s attempts to expand its control in Yamato, for the Hossō center transformed a substantial number of established temples into branches there. Few temples succeeded in refuting such advances, but Tōnomine managed to defend its independence vis-à-vis Kōfukuji, if only with the support of Enryakuji. By contrast, Kinpusen, in the southern part of Yamato, faced a much more difficult battle. Despite hard-nosed resistance, which on several occasions led to armed conflicts, it was forced to acknowledge Kōfukuji’s superiority by the late eleventh century, when the Hossō
238 | mikael adolphson
center had obtained enough leverage to appoint abbots at Kinpusen.84 These developments reflect not only general trends in eleventh-century Japan, but also particular circumstances at Kōfukuji. First, it was one of the most influential elite temples in the mid-Heian age and could successfully resist excessive taxes levied by the provincial government. Many monks of smaller temples undoubtedly saw advantages in the protection that patron temples could offer, even if some temples resisted aggressive takeovers.85 Second, Kōfukuji was damaged by fire no fewer than seven times in the eleventh century. Although local fields were assessed taxes by the government to help with the reconstruction, Kōfukuji was also forced to extract levies from its own fields as well and was actively looking for opportunities to increase its own revenues.86 In fact, the need for additional funds came to serve as extra leverage against the imperial court in its claims to exemptions and control of other temples in Yamato. Needless to say, Kōfukuji’s special status as one of the most privileged and patronized temples of the imperial court was crucial in this respect. Third, it was precisely in the eleventh century that sons of the Northern Fujiwara and their allies (such as the Minamoto) began to extend their influence within Kōfukuji as a privileged class of noble monks. By entering the Fujiwara clan temple, these aristocratic abbots served as an extension of their noble lineage, exerting influence over important religious ceremonies while adding human and land resources to their own factions. Enryakuji was no less aggressive in creating new branches. Its successful takeover of Gionsha during the ambitious and resourceful head abbot Ryōgen’s tenure in the late tenth century is well-known although the particular circumstances are unfortunately poorly documented. Somehow, Enryakuji managed to snatch Gion away from Kōfukuji, and the only textual explanation we have is an incredible story about jealousy over a tree and its blossoms between neighboring monks recounted in Konjaku monogatari, a twelfth-century collection of stories with Buddhist implications.87 It seems likely, though, that this transformation took place as a result of Ryōgen’s influential connections at court and the successful transformation of a Tendai betsuin associated with Gionsha into a branch. Other prominent branches include Daisenji in Hōki Province, a substantial monastic complex that became a forceful Tendai outpost. Although later legends state that the connection dates to Ennin (794 – 864), who is said to have visited the mountain and sponsored a temple there in thanks for a successful passage to and from China, other sources do not support this claim.88 We can only know with certainty that Daisenji had become a branch of Enryakuji some time before 1094, when its monks traveled to the capital to protest a lack of support from the incumbent Tendai head abbot — a clear indication of a matsuji connection. In either case, Enryakuji continued to expand its branch network well into the Kamakura age, eventually numbering more than 370 branches in the fourteenth century.89 The creation of branches and networks was above all part of a centering process that progressed from the late tenth century and culminated in the late Heian age. On the one hand, unattached and politically peripheral temples voluntarily subjugated themselves to central temples to secure more yield from land by mak-
Institutional Diversity and Religious Integration | 239
ing it completely tax exempt, to find protection and support against threats to its property, and to be included in, rather than excluded from, the power blocs of the capital. On the other hand, the elite temples were also active in trying to incorporate more branches in their network, but material assets were not the only thing on their mind. As indicated by Enryakuji’s and Kōfukuji’s efforts to incorporate important religious institutions such as Tōnomine, Daisenji, Gionsha, and Kiyomizudera, prestigious regional centers with long histories of their own were particularly attractive as branches. Such affiliations undoubtedly raised the patron temple’s own religious status, while the assigning of monks from the main temple as abbots of branches made it possible to promote the teachings of the patron temple. The success and the spread of the matsuji can also, as in the case of shōen, to a large extent be explained by their flexibility. Each branch was attached to the main institution according to its own conditions and history, and the “contracts” were flexible enough to allow both parties room for individual variations. The products and fees due to the main temple thus varied, and could even be adjusted according to circumstances. Still, in all cases, the flow of income went to the center, reversing the trend from the ninth century, when the granting of funds for jōgakuji frequently kept revenues away from the capital region. In that sense, the matsuji represent a trend toward centering in the Heian state, reflecting the emergence of a system of power, authority, and rulership, where the top layer was broader but the vertical ties to the center stronger.
Conclusion The political and religious centers did not always pull in the same direction during the early and mid-Heian age. In the ninth century, leading members of the imperial court sought to maintain a unified hierarchy of temples by distinguishing officially recognized temples from those that had been privately founded and patronized. The most extensive attempt in this direction were the jōgakuji, whose status as certified temples included important grants, exempt land, and other opportunities to earn income. From an economic standpoint, this meant that the flow of income actually favored the peripheral temples and their secular patrons. The situation was further complicated by the court’s attempts to have the patron, residing monks, and provincial governor manage the jōgakuji jointly, an arrangement that was anything but practical. More important, the court itself was utterly unable and unwilling to effectively curtail the establishment of new temples, as esoteric rites became integral to the success of the court’s own leading families, as well as to the welfare of the state. The economic nature of the jōgakuji proved insufficient to the designated temples as well. Not only was the state unable to provide any substantial benefits by the tenth century, but the jōgakuji were eventually forced to rely on their own assets and connections to survive. Alternative means of funding and recognition already existed beyond the confines of the imperial state in the form of exalted goganji or as detached cloisters of the larger monastic complexes. Ironically, this process also marked the beginning of a “recentering” of the Buddhist schools,
240 | mikael adolphson
since the betsuin designation created and maintained an important doctrinal affiliation with the most influential temples in Heian Japan. Furthermore, the judicial rights that the patron temple held over the detached cloisters indicate a strong hierarchical relationship that was not evident before the ninth century. In short, while the ninth century marks a tremendous spurt of institutional diversity, a trend toward religious integration under the umbrellas of the most influential religious centers had already begun. As the central elites faced and eventually dealt with serious provincial challenges in the tenth century, new modes of control and administration developed. Among these, the creation of private estates and the recruitment and use of warrior retainers to protect their assets and quell provincial disturbances stand out as the most significant adjustments of what can only be called a process of privatization that served to maintain if not strengthen the central elites’ control of the provinces. But a simultaneous and parallel development also took place in the religious sphere. In terms of institutional developments, the temples in the center came to head networks of branches, which might best be described as the logical development of the betsuin, as they were fully private assets and religious outposts of their patron institutions. Still, the leeway that allowed for regional variations in these contractual affiliations contributed to a practical arrangement in which, through private ties, the peripheries became more attached to the centers in the Kinai. Moreover, it is important to recognize that the creation of branch ties was not a process that merely reflected the ambitions and power of the central elites. In fact, in most cases, branches were established through the initiative of a local monk or leader and were born out of a desire to be part of, not separated from, the power structures of the capital. Indeed, only through such ties could local temples and shrines maintain control over their estates and have access to legal and political protection. It was thus through a broadening of the center that it became possible for the capital elites to retain, and even improve, the centrality of Kyoto in the Heian polity. It was a sociopolitical system that could adjust to the changing conditions much more effectively than the early Heian bureaucratic structures, allowing for new classes and elites to share in the powers and privileges of state rulership without destroying or threatening the privileged position of the members of the imperial court or the monastic centers of Heian Japan.
Notes 1. For a recent treatment of the continued vitality of the court in the fifteenth and sixteenth centuries, see Lee Butler, Emperor and Aristocracy in Japan, 1467 – 1680. 2. For an illuminating survey of the religious institutional developments in the Nara and early Heian age, see Stanley Weinstein, “Aristocratic Buddhism,” 451 – 452. For Tenmu’s edict, see Ruijū kokushi, Tenmu tennō 8 (680). 3. The scholarly interpretation of Dōkyō and his rise to power varies from that of a malicious usurper who used the weakness of a sickly female ruler and directly caused the court to leave Nara (Ross Bender, “The Hachiman Cult and the Dōkyō Incident”) to that of an ally
Institutional Diversity and Religious Integration | 241 in Empress Shōtoku’s efforts to rid herself of aristocratic rivals (Joan Piggott, “Tōdaiji and the Nara Imperium,” 72 – 79). Whatever his true role or intentions, it is clear that later sources, patronized by the Fujiwara, portrayed Dōkyō as the ultimate representative of an evildoing and ambitious Nara clergy. 4. Shoku Nihongi, Hōki 11 (780) 1/20, 456. See also Weinstein, “Aristocratic Buddhism,” 458; and Abé Ryūichi, The Weaving of Mantra, 24. 5. Shoku Nihongi, Enryaku 2 (783) 6/10; Ruijū kokushi, same date; Weinstein, “Aristocratic Buddhism,” 458. 6. Ronald Toby’s “Why Leave Nara?” remains the most comprehensive analysis in English on Kanmu’s motives for moving the capital. 7. Ruijū kokushi, Enryaku 14 (795) 4/27, Enryaku 17 (798) 7/28; Weinstein, “Aristocratic Buddhism,” 458 – 460. 8. Ruijū kokushi, Enryaku 11 (792) 11/2; Fusō ryakki, Enryaku 17/7/2; Nishidera Shikibu, “Heian chō jiin kikō no seiritsu to katei,” 39. 9. Weinstein, “Aristocratic Buddhism,” 479. 10. Neil McMullin, “The Enryaku-ji and the Gion Shrine-Temple Complex in the Mid-Heian Period,” 170 – 171; Paul Groner, Ryōgen and Mt. Hiei, 75 – 77; Tsuji Zennosuke, Nihon bukkyō shi no kenkyū, 1:371; Nishidera, “Heian chō jiin kikō no seiritsu katei,” 39 – 41. 11. The Fujiwara clan was divided into four main branches — the Southern, Northern, Ceremonial, and Capital branches (or houses) — stemming from the four sons of Fujiwara no Fubito (659 – 720), the son of the patriarch Kamatari. 12. For an informative account of this incident in English, see William McCullough, “The Heian Court, 794 – 1070,” 33 – 34. 13. Ibid., 36 – 37; Tsuji, Nihon bukkyō shi no kenkyū, 384; Groner, Ryōgen and Mt. Hiei, 270 – 272. 14. Tsuji, Nihon bukkyō no kenkyū, 387. Shinnyo became an influential monk following his tonsure. In 855, he headed the reconstruction of the Daibutsu, whose head had come tumbling down, and he visited the T’ang Empire in 861. Unable to meet the Chinese master of his choice, he continued to India with three followers, but died before reaching his destination. See also Groner, Ryōgen and Mt. Hiei, 270. 15. Nihon sandai jitsuroku (hereafter Sandai jitsuroku), Gangyō 3 (879) 3/23; Hiraoka, Nihon jiin shi no kenkyū, 494 – 497, 510 – 517. The works of Hiraoka and Tsuji are still the most extensive treatments of the early Heian era’s religious institutions. 16. Shoku Nihon kōki, Jōwa 9 (842) 7/17; Nihon Montoku tennō jitsuroku (hereafter Montoku jitsuroku), Kashō 3 (850) 7/17; Sandai jitsuroku, Gangyō 1 (877) 12/9, Gangyō 6 (882) 6/3; Hiraoka, Nihon jiin shi no kenkyū, 491 – 507; Tsuji, Nihon bukkyō no kenkyū, 383 – 385. 17. Sandai jitsuroku, Jōgan 1 (859) 3/19, Gangyō 3/1/3; Hiraoka, Nihon jiin shi no kenkyū, 483 – 485. For an informative survey of the goganji, see Hiraoka, “Goganji no seiritsu to sono seikaku.” For a brief account in English, see Groner, Ryōgen and Mt. Hiei, 35 – 37. 18. Sandai jitsuroku, Jōgan 14 (872) 7/19, Jōgan 16 (874) 3/23, Jōgan 18 (876) 8/29; Hiraoka, “Goganji no seiritsu to sono seikaku,” 485 – 487; Tsuji, Nihon bukkyō no kenkyū, 388 – 389. 19. Shingi shiki, in Gunsho ruijū, 5:75. 20. Shoku Nihon kōki, Jōwa 13 (846) 8/17, 12/29; Jōwa 14 (847) 2/18. 21. Sandai jitsuroku, Jōgan 6/1/14, 124. 22. Tsuji, Nihon bukkyō no kenkyū, 370. 23. Sandai jitsuroku, Gangyō 8 (884) 9/17; Ninna 1 (885) 5/23, 9/4, 10/22; Ninna 2 (886) 3/14; Tsuji, Nihon bukkyō no kenkyū, 369, 385 – 387. One chō equaled about 1.13 hectares or 2.45 acres.
242 | mikael adolphson 24. Daigoji yōsho, 27 – 30; Weinstein, “Aristocratic Buddhism,” 501 – 502; Takeuchi Rizō, Ritsuryō sei to kizoku seiken, 546. 25. Hayami Tasuku, “Jōgakuji no kenkyū,” 57, 60 – 62, 70. 26. Nakai Shinkō, Nihon kodai bukkyō seido shi no kenkyū, 177. 27. Engi shiki, in Shintei zōho kokushi taikei, 26:533 – 535; Takeuchi, Ritsuryō sei to kizoku seiken, 515, 519 – 522, 539; Hayami, “Jōgakuji no kenkyū,” 62, 64; Nakai, Nihon kodai bukkyō seido shi no kenkyū, 192; Morita Tei, “Jōgakuji to Shinano,” 27. 28. Shoku Nihongi, Tenpyō shōhō 1 (749); Ruijū sandai kyaku, 116 – 118; Takeuchi, Ritsuryō sei to kizoku seiken, 510; Hayami, “Jōgakuji no kenkyū,” 56, 64; Usami Masatoshi, “Jōgakuji no seiritsu to henshitsu,” 88 – 91; Weinstein, “Aristocratic Buddhism,” 457 – 458. 29. Ruijō kokushi, Tenchō 1 (824) 9/27; Takeuchi, Ritsuryō sei to kizoku seiken, 512. 30. Sandai jitsuroku, Jōgan 5/9/6. 31. Sandai jitsuroku, Jōgan 8 (866) 2/2; Ruijū kokushi, 259, 261; Takeuchi, Ritsuryō sei to kizoku seiken, 513; Hayami, “Jōgakuji no kenkyū,” 58 – 62; Ushiyama, “Jōgan hachi-nen no Shinano jōgakuji rekkaku wo megutte: je,” 36 – 37. While it is impossible to know how many temples were given the status of jōgakuji, the sources suggest that the court intended for one in each district (gun), although it was never fully accomplished. In either case, it would amount to a substantial number beyond the jōgakuji we know today (Nakai, Nihon kodai bukkyō seido shi no kenkyū, 191; Harada Kazuhiko, Jōgakuji seishiki ron, 39 – 40). 32. Shoku Nihon kōki, Jōwa 10 (843) first month, Kashō 2 (849) 2/25; Sandai jitsuroku, Jōgan 7 (865) 2/10, Jōgan 15 (873) 2/23, Jōgan 17 (875) 12/13; Harada, 41 – 44. 33. Heian ibun, vol. 1, doc. 164, Anshōji garan engi shizai chō, Jōgan 13 (871) 8/17; Hayami, “Jōgakuji no kenkyū,” 63. 34. Ruijū sandai kyaku, 120 – 121; Nihon kōki, Daidō 1 (806) 8/27; Takeuchi, Ritsuryō sei to kizoku seiken, 514, 540 – 541; Hayami, “Jōgakuji no kenkyū,” 64; Nakai, Nihon kodai bukkyō seido shi no kenkyū, 206; Harada, Jōgakuji seishiki ron, 30. 35. Ruijū kokushi, Tenchō 1/9/27; Daigoji yōsho, 29 – 31; Nakai, Nihon kodai bukkyō seido shi no kenkyū, 207. 36. Sandai jitsuroku, Gangyō 1/12/9. 37. Ibid., Gangyō 1/12/9; Kyūreki, Tenryaku 2 (948) 4/23; Takeuchi, 546; Satō Kenji, Chūsei kenmon no seiritsu to kasei, 230; Nihon kiryaku, Shōhei 4 (934) 10/10. 38. The entry for Kokushōji can be found in Nihon kiryaku, Kōnin 11 (820) 11/22; for Kangadera, in Sandai jitsuroku, Gangyō 5 (881) 10/3; for Gankōji, in Sandai jitsuroku, Gangyō 8/8/26; for Daikōji, in Shoku Nihon kōki, Jōwa 10/12/1, and Montoku jitsuroku, Saikō 3 (856) 9/13; Hayami, 58 – 59; Usami, Jōgakuji no seiritsu to henshitsu, 94 – 95; Nakai, Nihon kodai bukkyō seido shi no kenkyū, 178 – 183. 39. Shoku Nihon kōki, Daidō 4 (809) 1/18. 40. Nihon kiryaku, Kōnin 11 (820) 11/22; Nakai, Nihon kodai bukkyō seido shi no kenkyū, 178; Harada, 376 – 377. 41. Ushiyama Yoshiyuki, “Jōgan hachi-nen no Shinano jōgakuji rekkaku wo megutte: jō,” 6, 18 – 19; Usami, 92, 98; Harada, 46 – 48. The five temples were granted jōgakuji status on 866/2/2, while the Ōtenmon Incident occurred on 866/int. 2/10. 42. Ruijū sandai kyaku, Daijō kanpu, Enryaku 15 (796) 3/25. See also Takeuchi, Ritsuryō sei to kizoku seiken, 537; Harada, Jōgakuji seishiki ron, 35. 43. Ruijū sandai kyaku, Daijō kanpu, Jōgan 10 (868) 6/28; Sandai jitsuroku, Jōgan 10/6/26; Takeuchi, Ritsuryō sei to kizoku seiken, 553; Harada, Jōgakuji seishiki ron, 35 – 36. 44. Engi shiki, 541; Takeuchi, Ritsuryō sei to kizoku seiken, 528; Harada, Jōgakuji seishiki ron, 33. 45. Heian ibun, vol. 1, doc. 174, Kawachi no kuni Kanshinji engi shizai chō, Gangyō 7 (883)
Institutional Diversity and Religious Integration | 243 9/15; Sandai kyaku, 63; Daigoji yōsho, 28 – 30; Sandai jitsuroku, Gangyō 2 (878) 2/5; Hayami, “Jōgakuji no kenkyū,” 72; Satō Kenji, Chūsei kenmon no seiritsu to kasei, 228; Harada, Jōgakuji seishiki ron, 379. 46. Daigoji yōsho, 28 – 30; Takeuchi, Ritsuryō sei to kizoku seiken, 552. 47. Nihon kōki, Daidō 1 (806) 8/22; Shoku Nihon kōki, Jōwa 8 (841) 5/20; Sandai jitsuroku, Jōgan 1 (859) 7/15; Hiraoka, Nihon jiin shi no kenkyū, 424, 445; Toyoda Takeshi, Nihon shūkyō seidō shi kenkyū, 30 – 31; Harada, Jōgakuji seishiki ron, 36. 48. Sanmon dōsha ki, 514 – 515; Takeuchi, Ritsuryō sei to kizoku seiken, 513. 49. Takeuchi, Ritsuryō sei to kizoku seiken, 522 – 523. 50. Iken jūnikajō, 127. For an English translation of the relevant article, see Groner, Ryōgen and Mt. Hiei, 8 – 9. 51. For a detailed analysis of warrior retainers serving the Fujiwara, see Karl Friday, Hired Swords. 52. Tōdaiji monjo, in Dai Nihon kumonjo, iewake, doc. 592, Fun’ya no Mahito hasera butsuzō narabi ni issaikyō nado senyū ganmon, Enryaku 17/8/26, 41 – 42; Tōnomine ryakki, 504 – 507; Takagi Yutaka, Heian jidai hokke bukkyō shi kenkyū, 44. 53. Shoku Nihon kōki, Jōwa 4 (837) 2/27; Nishidera Shikibu, “Jōdai makki kara chūsei ni itaru betsuin, bessho no matsuji e no tenkai,” 54. 54. Hiraoka Jōkai, Ronshū Nihon bukkyō shi, 3:32. 55. Groner, Ryōgen and Mt. Hiei, 200 – 201. 56. Takagi, Heian jidai, 50. 57. Shoku Nihon kōki, Jōwa 11 (844) 4/30. 58. Takagi, Heian jidai, 49. 59. Sandai jitsuroku, Gangyō 2/8/13, Gangyō 6/12/10. 60. Sandai jitsuroku, Ninna 2/4/3; Takagi, 47 – 48. 61. Shoku Nihon kōki, Jōwa 6 (839) 1/26, Jōwa 7 (840) 12/7; Kashō 2 (849) 1/26; Sandai jitsu roku, Jōgan 16/12/25. 62. For Enshōji, see Sandai jitsuroku, Jōgan 8/4/7, and Gangyō 6/12/10; for Daikōji, see Montoku jitsuroku, Saikō 2 (855) 9/28; Takagi, 48. 63. Tōnomine ryakki, 22 – 24. 64. Sandai jitsuroku, Jōgan 16/12/25. 65. Sandai jitsuroku, Gangyō 8/9/10, Ninna 2/4/3. 66. Sō Myōtatsu soseki chūki, 305. 67. Toyoda, Nihon shūkuō, 32; Nishidera, “Jōdai makki kara chūsei ni itaru betsuin,” 55 – 56. 68. Heian ibun, vol. 1, doc. 153, Onjōji shoshira ge an, Jōgan 9 (867) 2/28; Heian ibun, vol. 9, doc. 4497, Enryakuji mandokoro chō an, Jōgan 8/10/11; Daijō kanchō, Jōgan 8/5/14, in Shiga ken shi, 2:191. 69. Adolphson, The Gates of Power, 112 – 113. 70. Heian ibun, vol. 8, doc. 3293, Kōfukuji kumon dokoro kudashibumi an, Chōhō 2 (1164) 7/28; Heian ibun, vol. 10, doc. 5098, Seishōji nenchū sōsechi chō; Takagi, 56. 71. Kuroda Toshio, Jisha seiryoku, 156 – 157. 72. Heian ibun, vol. 3, doc. 1115, Kanzenji an, Shōhō 2 (1075) 5/12. 73. Tōnomine ryakki, 507. 74. Tōnomine ryakki, 487; Takagi, Heian jidai, 52; Nishidera, “Jōdai makki kara chūsei ni itaru betsuin,” 57 – 58. 75. DNS 3:5, 977; 3:6, 433 – 453. 76. Tōnomine ryakki, 482 – 483. 77. Tōnomine ryakki, 507 – 509.
244 | mikael adolphson 78. Hyakurenshō, Kyūan 3 (1147) 4/7, 5/4; Honchō seiki, Kyūan 3/4/12; Toyoda, Nihon shūkyō, 36. 79. Heian ibun, vol. 3, doc. 983, 984, 986, 1077, 1088; vol. 9, docs. 4631, 4641. 80. Chūyūki, Kajō 1 (1106) 2/23. For entries mentioning the fires and repairs, see Chūyūki, Kanji 5 (1091) 3/8, and Kahō 1 (1094) 10/22. 81. Hiraoka Jōkai, “Echizen no kuni Kanzeonji no Tōdaiji no matsuji ka ni tsuite,” 194 – 196. 82. Kamakura ibun, vol. 2, doc. 967, Kenkyū 9 (1198) second month. 83. Yoneda Yūsuke, Fujiwara sekkanke no tanjō, 350 – 351; Heian ibun, vol. 4, doc. 1385, Kōfukuji mandokoro kudashibumi an, Eichō 2 (1097) 10/16. 84. Chūyūki, Kahō 1/3/6. For an English account of the local conflicts between Kōfukuji and Tōnomine, see Adolphson, Gates of Power, 93 – 96, 144 – 146. 85. Heian ibun, vol. 9, doc. 4639 – 4640, Kōfukuji Yamato no kuni zatsuyaku men tsubo, Enkyū 2 (1070) 9/20; Yoneda, Fujiwara sekkanke no tanjō, 353. 86. Yoneda, Fujiwara sekkanke no tanjō, 353. 87. Toyoda, Nihon shūkyō, 35; McMullin, “The Enryaku-ji,” 161 – 163, 167, 169; Groner, Ryōgen and Mt. Hiei, 203 – 205. 88. Numata Ruisuke, Daisen zakkō, 9 – 10; Daisen chō shi, 148 – 149. 89. Taiheiki, in Nihon koten bungaku taikei, 35:430; Toyoda, 40.
10
d D.Max Moerman
The Archeology of Anxiety An Underground History of Heian Religion
T
he Heian period has often been characterized as a golden age of Japanese religious culture, one in which the Buddhist literature, painting, sculpture, and architecture produced at court are considered to have reached unprecedented heights. It is thus all the more surprising that in this time of cultural florescence many Buddhists, monastics and aristocrats alike, understood themselves to be living in an age of decay. According to their interpretation of Buddhist chronologies, the eleventh century marked the beginning of the end: the onset of mappō, the final degenerate age of the Dharma, or Buddhist Law, in which both the availability of the teachings and the ability of people to realize them would reach their lowest points. The ineluctable decline of the Dharma presented soteriological problems for both the tradition and the individual. The death of the Dharma challenged, of course, the very existence of Buddhism and required acts of protection and preservation to ensure its survival. But mappō also had implications for Buddhist practitioners for whom individual salvation became increasingly difficult as the source of teachings receded into an inaccessible past and the spiritual capabilities of humans diminished. The Heian period was thus a time when history itself represented a profound religious problem and when Japanese Buddhists began to formulate specifically religious responses to the problem of history. This essay examines one religious practice commonly cited as evidence of mappō consciousness: the interment of specially consecrated copies of Buddhist scriptures in the earth at sacred mountains, shrines, and temples. The process required that sūtras be transcribed according to strict ritual protocols, enclosed within reliquary-shaped containers, and buried underground to protect and preserve their teachings until the arrival of the next Buddha, Miroku (Skt. Maitreya), some 5.67 billion years in the future. The practice is referred to both by the objects interred, maikyō (buried sūtras), and by the tumuli that marked their presence, kyōzuka (sūtra mounds). The contents, dedicatory inscriptions, material form, and locations of these burials appear to address the twin religious challenges of mappō by establishing a link between the death and salvation of the Buddhist tradition and that of the individual believer. The primacy that mappō thought has been afforded in the historiography of Heian religion, however, has not gone unchallenged. Inasmuch as much of the discourse on mappō is to be found in literary texts, one may indeed wonder
246 | d. max moerman
whether it was more of a rhetorical trope than a motivational conviction. Indeed, invocations of mappō were deployed for a range of polemical reasons. They were used to justify the loosening of monastic precepts, for example, as often as to buttress the case for their stricter enforcement. Moreover, as the theory of mappō was largely the product of religious professionals, one may also wonder to what extent these notions were more broadly shared. Taira Masayuki has argued that although mappō discourse was born of a crisis in the economic and political authority of temples and shrine institutions, it soon spread beyond such specific ideological origins to constitute a far more widely held form of historical consciousness.1 Through the analysis of sūtra burials, I aim to question the significance of mappō to the historical and religious consciousness of Heian-period Buddhists and to examine the geographical range of that consciousness. How central was the concern with mappō? How closely tied was it to a social elite occupying the geographical center? Do sūtra burials reveal religious concerns other than an anxiety about the end of the Dharma? Does the location of sūtra burials suggest significant centers of religion and culture other than Heian-kyō? In short, how might the evidence of sūtra burials contribute to a more complex understanding of the spatial and cultural boundaries of Heian religion? The chronological and geographic distribution of sūtra burials, I would like to suggest, might offer a sort of subterranean cartography of the Heian religious imagination as well as chart the connections between the institutions, individuals, and ideologies of the capital and the provinces. Such an archeology of Heian religion provides an alternative “spatial history” of the religious centers and peripheries of the age.
Chronicle of a Death Foretold The idea that the Dharma, the Buddha’s teaching, will pass through successive stages of degeneration and eventually disappear is found throughout the Buddhist tradition.2 Although the Pali and Sanskrit texts that refer to this historical plan rarely agree on the precise details — the cause, nature, or timetable of decline — Chinese translations and commentaries address the issue with far greater specificity. Working from Indian sources, and perhaps influenced as well by native traditions, Chinese Buddhists of the sixth century first articulated the scheme of three distinct historical periods following the Buddha’s death.3 This cyclical chronology comprises the periods of True Dharma (Ch. zheng fa; J. shōbō), during which time Buddhist teachings and practices are available and enlightenment accessible; Semblance Dharma (Ch. xiang fa; J. zōbō), when teachings and practices are maintained but humanity’s spiritual capacity has seriously diminished; and Final Dharma (Ch. mofa; J. mappō), when true practice has disappeared and only the teachings remain, destined themselves soon to vanish. In this last age the capacity for enlightenment is at its nadir, and the world will continue in its decline for some ten thousand years. The fact that this final age has little or no Sanskrit textual basis does not seem to have posed a problem for Chinese Buddhists.4 Whether canonical or apocryphal in origin, they found convincing proof
The Archeology of Anxiety | 247
of this prophecy in their immediate surroundings. Buddhist persecutions in the sixth century helped to make mofa a bitter reality, and from the Sui and Tang dynasties, Chinese Buddhists believed they were living in the age of the Final Dharma and began to formulate teachings appropriate to it. The concept of mappō was part of Japan’s early continental inheritance but did not gain popular currency until the mid-Heian period. Fueled in part by the growing influence of Tendai Amidism — a tradition that in Japan as in China presented itself as the most expedient path in the Final Age — many Japanese monks and aristocrats began to see their era as one of acute historical and religious crisis. The precise date for the Final Age, however, was a matter of some debate. The four most common Chinese versions of the chronology measured the periods of True Dharma, Semblance Dharma, and Final Dharma as lasting 500, 500, and 10,000 years; 500, 1,000, and 10,000 years; 1,000, 500, and 10,000 years; and 1,000, 1,000, and 10,000 years. Variant dates for the Buddha’s death — usually 609 or 949 B.C.E. — also had implications for the onset of mappō.5 By the tenth century, however, 949 B.C.E. was generally accepted as the year of the Buddha’s death, and one thousand years were allotted for each of the intervening periods of shōbō and zōbō. The age of mappō was thus most commonly understood to begin in the year 1052.6 The Fusō ryakki, the chronicle of Japanese history written by the monk Kōen (d. 1169), recognized this inaugural date by noting in the entry for 1052/1/26 that “this year marks the beginning of mappō.”7 Although there was less than universal agreement on the chronological details of the overall scheme, this tripartite model of religious history was acknowledged by numerous Buddhist authors of the early Heian period. Saichō (767 – 822), Anchō (763 – 814), and Annen (841 – 884) of the Tendai school; Gen’ei (ca. 840) of the Sanron school; and Zan’an (ca. 776 – 815) of the Hossō school were all concerned with the end of the Dharma and its soteriological implications.8 In 984 Minamoto no Tamenori (d. 1011) voiced the immediacy of the threat in the preface to his Buddhist tale collection, the Sanbōe: “One thousand nine hundred and thirty-three years have passed since the Buddha Śākyamuni left this world. We may now be in the period of Semblance Dharma, but surely only a few years of this interim period remain to us.”9 One year later, the Tendai monk Genshin (942 – 1017) composed his Ōjōyōshū, or Essentials for Rebirth, one of the more influential Buddhist texts of the Heian period, which began with the claim that “the teaching and practice that leads to birth in paradise is the most important thing in this final age of defilement (matsudai).”10 Matsudai, the term Genshin uses for the Final Dharma, can also be translated as “final generations” and, together with the term masse, “final age,” was used often interchangeably with mappō. The political connotations of the expression “final reigns” were not lost on Genshin’s aristocratic audience, who were to later see a growing equivalence between political chronology and Buddhist eschatology.11 Courtier diaries from the late eleventh century often invoke mappō as both lamentation and historical explanation for the current social and political state of affairs.12 As in China, Pure Land faith, promising rebirth in Amida’s heavenly Gokuraku
248 | d. max moerman
realm, offered one of the more compelling solutions to the religious problems of the age. One must be careful, however, not to read back into Heian practice the theological positions of later Pure Land leaders. One simply does not see in the Heian period a populace paralyzed in the face of irreversible religious decline nor many who would find solace in Shinran’s position that meritorious deeds are counterproductive to salvation. Rather, the Heian period witnessed a great deal of individual effort aimed at both counteracting the effects of mappō and ensuring, in the face of such a threat, one’s own heavenly rebirth. It was a period of profound religious activity among the aristocracy in which the production of Buddhist ritual, art, and architecture seemed to have no limits. It is in this context of escalating ritual exertion seeking protection and preservation, of both the individual and the Dharma, that the practice of sūtra burial became common.
Last Rites Buried sūtras were transcribed on a variety of materials, most often on paper or silk scrolls in black, gold, or vermilion ink (the latter was occasionally mixed with blood).13 Yet there are also numerous examples of sūtras inscribed on more permanent materials such as stone, ceramic tiles, or copper plates, signaling perhaps an even more explicit concern with the preservation of the teachings. The silk or paper sūtras would be placed into cylindrical stupa-shaped containers (kyōzutsu, “sūtra tubes”) fashioned out of bronze, iron, ceramic, or stone that were often in turn encased in a second outer vessel of ceramic or stone. They were then buried in small underground chambers lined with stones and occasionally packed with charcoal to aid in preservation. The chambers were sealed with stone and marked, like a grave, with an earthen mound and a stone stupa, lantern, or stele (figure 10.1). The sūtra tubes themselves exhibit a great variety of styles from the detailed miniature treasure pagoda (hōtō) to the simple lidded cylinder. Yet however elaborate or plain, all share the basic form of the stupa, a reliquary housing the remains of a Buddha (figure 10.2). The stupa symbolizes at once the death of the founder and the genesis of the religion. It represents both the architectural structure housing the Buddha’s cremated remains and the sacred vessel enshrining his relics. As death rituals for the Dharma body, sūtra burials were in keeping with the origins of the tradition and were understood within the vocabulary of Japanese Buddhist practice as kuyō (memorial services). The repertoire of death rituals were put to a wide range of uses at the Heian court, and aristocrats often sponsored their own premortem memorial services (gyakushu). The Genji monogatari describes an elaborate gyakushu performed during Murasaki’s final illness in which one thousand copies of the Lotus Sūtra were presented as “her final offering to the Blessed One” in a service so compelling that “one felt that Amida’s paradise could not be far away.”14 The Eiga monogatari also mentions Fujiwara no Michinaga’s performance of “forty-nine days of service on his own behalf” during which forty-nine copies of both the Lotus Sūtra and the Amida Sūtra
Figure 10.1. Diagram of sūtra burial
Figure 10.2. Copper sūtra container and outer ceramic case. Late Heian period (11th – 12th century). Kyoto National Museum Collection. Courtesy ofKyotoNational Museum.
250 | d. max moerman
were offered.15 As kuyō, sūtra burials produced a great deal of symbolic value, yet the beneficiary of this merit — the Dharma, the sponsor, the sponsor’s family members — was by no means fixed. The scripture most often interred in sūtra mounds was the Lotus Sūtra (Skt. Saddharma pun.d.arīka sūtra; J. Myōhō renge kyō, T 262) which in the Tendai tradition included the Sūtra of Innumerable Meanings (Skt. Amitārtha sūtra; J. Muryōgi kyō, T 276) and the Sūtra of Meditation on the Bodhisattva Universal Virtue (Skt. Samantabhadra bodhisattva sūtra; J. Kanfugen bosatsu gyōbō kyō, T 277) as its opening and closing chapters. The presence of the Lotus Sūtra in the vast majority of such burials should come as no surprise. It was perhaps the most important scripture in Heian religious culture and the principal text of the Tendai school, which in China had most explicitly articulated the theory of the Final Dharma Age and in Japan was credited with the origins of sūtra burial. It was also a text that explicitly encouraged its own transmission, enshrinement, and veneration. Like a number of other early Mahāyāna sūtras, the Lotus Sūtra claims that because of its status as the textual corpus of the Buddha’s teachings, it supersedes the corporal relic of the Buddha himself (Skt. sárīra) as the true body of the Buddhadharma.16 “Whatever place a roll of this scripture may occupy, in all those places one is to erect a stupa of seven jewels....There is no need to even lodge a sárīra in it. What is the reason? Within it there is already a whole body of the Thus Come One.”17 The enshrinement of the Lotus Sūtra within stupa-shaped reliquaries is thus entirely in keeping with the sacramental logic of Mahāyāna sūtra cults. The transposition of the text for the relic, or rather the “translation” (in the true etymological sense) of the relic into text, which took place in Indian Buddhism around the turn of the millennium, was part of Japanese Buddhism from its very beginnings. As early as the Nara period, relics traditionally deposited beneath the central posts of pagodas were replaced by sūtra texts.18 Such examples of bibliolatry, following the injunctions of the sūtras themselves, underscore the significance of the materiality and performativity of religious texts in the Heian period, an understanding of the function of scripture that informs the practice of sūtra burial as well. Indeed, the Lotus Sūtra deserves the highest praise for those “who shall receive and keep, read and recite, explain, or copy in writing a single verse of the Scripture of the Blossom of the Fine Dharma, or who will look with veneration on a roll of this scripture as if it were the Buddha himself.”19 In carrying out these scriptural instructions, the sponsors of sūtra burials enjoyed the combined merit of copying and protecting the sūtra together with that of building a stupa in which to enshrine and venerate it. These practices were understood to be particularly timely as the Buddha of the Lotus Sūtra emphasized that such methods of textual preservation and devotion are to be undertaken specifically “after my extinction, in an evil age.”20 To judge from their devotion to the Lotus Sūtra, Heian aristocrats must have seen themselves as the chosen few referred to in the sūtra as those “who in the latter age can receive and keep this scripture.”21 Although preeminent, the Lotus Sūtra was neither the only text nor the only
The Archeology of Anxiety | 251
cult represented in late Heian sūtra mounds. As the Final Age was understood to portend political as well as religious troubles, the Lotus Sūtra was also often joined by two additional scriptures held to protect the state, the Golden Light Sūtra (Skt. Suvarn. aprabhā sūtra; J. Konkōmyō kyō, T 663 – 665) and the Benevolent Kings Sūtra (Skt. *Kārunikā rājā sūtra; J. Ninnō hannya haramitsu kyō, T 245 – 246 [*indicates reconstruction]). The other scriptures most commonly buried, and usually accompanying the Lotus Sūtra, were the three Amida (Skt. Amitābha) sūtras — the Larger Pure Land Sūtra (Skt. Sukhāvatīvyūha sūtra; J. Muryōju kyō, T 360), the Smaller Pure Land Sūtra (Skt. Sukhāvatīvyūha sūtra; J. Amida kyō, T 366), and the Sūtra of Meditation on Amida Buddha (Skt. Amitāyur dhyāna sūtra; J. Kanmuryōju kyō; T 365) — and the three Miroku (Skt. Maitreya) sūtras — the Sūtra on Maitreya Achieving Buddhahood (J. Miroku jōbutsu kyō; Ch. Mile chengfo jing, T 456), the Sūtra on Maitreya’s Rebirth Below [on Earth] (J. Miroku geshō kyō; Ch. Mile xiasheng chengfo jing, T 454), and the Sūtra on Maitreya’s Rebirth Above in Tus.ita (J. Miroku joshō tosotsu kyō; Ch. Mile shangsheng dou shuai jing, T 452). The Amida sūtras describe the Buddha Amida’s Gokuraku Pure Land and his vow to guarantee rebirth there for all who call on him. The Miroku sūtras describe the Bodhisattva Miroku’s practices in the Tosotsu (Skt. Tuśita) heaven where, with the accumulation of sufficient merit, his devotees may also be reborn. They tell as well of a future golden age, 5.67 billion years after the death of the Buddha Śākyamuni, when the Dharma will rise again to the apex of its historical cycle. At that time Miroku, also known as Jison in his role as the future Buddha, will descend to the earth and expound the Dharma at three assemblies to be held beneath the legendary Dragon Flower Tree (ryūge san’e). The Lotus Sūtra, Amida, and Miroku cults were in no way mutually exclusive. The Lotus Sūtra, for example, guarantees rebirth in Amida’s Pure Land (notably “after the extinction of the Thus Come One, within the last five hundred years”) to women who revere the sūtra’s twenty-third chapter.22 Elsewhere it promises that any male devotee “at life’s end...shall straightway ascend to the top of the Tosotsu [Tuśita] heaven, to the place of the Bodhisattva Miroku [Maitreya].”23 Yet while included within the ecumenicism of the Lotus Sūtra, Miroku and Amida could claim their own traditions in Heian Japan. According to the Nihon shoki (Chronicle of Japan), compiled by the state in 720, the cult of Miroku appears to have been the earliest form of Buddhism received by the Japanese elite in the mid-sixth century from the Korean Peninsula. Members of the Soga clan embraced the veneration of Miroku, spread by monks from Paekche and Silla, in the sixth century, and by Japanese rulers of the seventh century, as evidenced by midto late seventh-century images of Miroku at Hōryūji, Yakushiji, and Taimadera.24 By the end of the Nara period, images of Miroku were enshrined at all of the seven great temples of the capital (Daianji, Kōfukuji, Yakushiji, Gangōji, Tōdaiji, Tōshōdaiji, and Saidaiji), and the Miroku cult had developed strong traditions beyond the capital at Kasagidera in Yamashiro, Sūfukuji and Miidera in Ōmi, and Katsuodera in Settsu.25 The cult of Miroku remained active into the Heian period. The Tendai patriarch Saichō (767 – 822) wrote to a former disciple that “we are both waiting to see
252 | d. max moerman
Miroku.”26 Kūkai (774 – 835) had the Jison’in (the chapel of Miroku) constructed as a mausoleum for his mother and wrote of her aspiration for rebirth in the Tosotsu heaven.27 Kūkai’s close association with Miroku, however, was to come with the later development of the patriarch’s posthumous cult. The Goyuigon, Kūkai’s apocryphal final testament, has him saying, “After I close my eyes, I will without fail be reborn in the Tosotsu heaven, where I can serve Miroku Jison. In more than 5.6 billion years, I shall descend to earth along with Jison and honor him.”28 Written by others long after Kūkai’s time, this forgery is not without historical value. It represents mid-Heian-period understanding of the relationship between Kūkai and Miroku. By the late eleventh century, proponents of the cult of Kūkai (Kōbō Daishi) developed the tradition that he had not really died but rather had entered a meditative state until the coming of the future Buddha. Mount Kōya, the site of Kūkai’s monastery and tomb, was also by this time both identified with the Tosotsu heaven and understood to be the place of Miroku’s descent. As such it became a major center for sūtra burials. The mid- to late Heian period, however, is generally identified less with the cult of Miroku than with that of Amida. Texts such as Genshin’s Ōjōyōshū and the hagiographic genre of ōjōden (accounts of those reborn in the Pure Land) exemplify the profound concern among Buddhists of the time with rebirth in Amida’s Pure Land and with the practices by which it could be achieved. Yet the growth in popularity of Amida within the Tendai tradition did not necessarily mean the eclipse of Miroku faith. Genshin, following the Chinese Pure Land patriarch Shandao (613 – 681), argued that the Gokuraku Pure Land was superior to the Tosotsu heaven and that Amida’s vow was more embracing than that of Miroku.29 The written prayers of prominent monks and aristocrats, and the hagiographic literature of ōjōden, however, reveal that faith in Miroku did not disappear.30 Many who prayed for rebirth in Amida’s Gokuraku, for example, added the proviso that they be able to return at the time of Miroku’s advent. One significant difference nevertheless remained between the two cults. Miroku faith emphasized that heavenly rebirth could be gained through religious works, the most common of which in the Heian period was copying the Lotus Sūtra. This emphasis on scriptural production and displays of piety suggests another reason, beyond the eschatological, for the connection between the Miroku cult and the burial of transcribed sūtras. The origin of sūtra burial is traditionally ascribed to the Tendai patriarch Ennin (794 – 864), who copied the Lotus Sūtra and enshrined it in a small stupa on Mount Hiei in 831. Ennin’s method of copying, known as nyohō or “according to prescribed method,” was itself a major ritual undertaking and set the standard for sūtra transcriptions throughout the Heian period. Although sūtra copying as a religious practice goes back at least to the Nara period, Ennin’s efforts were of a different ritual scale. He is said to have retreated to the Yokawa section on Mount Hiei for three years to prepare for and carry out the transcription. There he grew the hemp to make the paper on which the sūtra was to be written, made his own brush of twigs and grass rather than animal hair, and made his own ink
The Archeology of Anxiety | 253
from graphite rather than using ink sticks containing animal glue. Throughout this period he also carried out the four forms of meditation central to Tendai praxis.31 Combining ritual and writing, Ennin performed three full prostrations with the transcription of each character. He placed the completed sūtra within a small wooden stupa, presented to it ten kinds of offering, and installed it in a hall at Yokawa later known as the Nyohōdō.32 In 998, Genshin, whose concern with the Final Age informed his Ōjōyōshū, placed Ennin’s wooden stupa of 831 containing the sūtra inside a larger bronze stupa. Yet the sūtra was not in fact buried until 1031, a full two centuries after Ennin’s transcription, when the monk Kakuchō (960 – 1034) and a small circle of fellow monks at Yokawa enclosed it within a third bronze reliquary and buried it beneath the Nyohōdō. According to the Nyohōdō dōtōki, In the Final Age (masse), the head monk commanded that the gilt bronze sūtra tube in the hall be moved and that the sūtra be buried beneath the earth and stones of the mountain to await the coming of Miroku [Jison]. This is in accordance with the Master’s original vow (Daishi no hongan)....[The sūtra] dwelling inside a seven-jeweled stupa will assuredly be transmitted to the age of Miroku and thus Śākyamuni’s Dharma will save people. People will rely on this sūtra until the age of Miroku arrives.33 It was exhumed 140 years later and then buried again in 1171. Although Ennin cannot thus be credited with performing the first sūtra burial, he was nevertheless in the late Heian period seen as the originator of the practice. The postscript to a Lotus Sūtra buried in 1158, for example, refers to “the power of [Ennin’s] vow [to preserve the sūtras] until the coming of the Buddha Miroku.”34 The inscription on a sūtra tube bearing the date 1186 similarly describes its contents as a “nyohō kyō [transcribed] in accordance with the method [of Ennin].”35 Ennin’s choice of scripture, method of copying, and location of practice served as a guide for later practitioners. Even when found with other scriptures, the Lotus Sūtra is the single most common sūtra buried, the method of its transcription is based on Ennin’s prescribed manner, and the burial sites are often at or near Tendai mountain temples. The question that remains, however, is whether these locations represented central or peripheral sites of Heian religious culture.
Spatial Practices The Heian age was both the earliest and the most active period of sūtra burials. Although sūtra mounds were constructed from the mid-Heian through the late Edo periods, more than half date from the eleventh and twelfth centuries.36 Heian sūtra burials are more extravagant than those of later periods and include many examples of gold ink transcriptions on indigo paper. The period is also distinguished by the greater number of sūtras interred at a single site and by the inclusion of other smaller items such as mirrors and swords.37 Yet these burials speak to more than the historical and soteriological anxieties of the age. They also locate the sites where such anxieties were expressed and where, it was hoped,
254 | d. max moerman
Map 10.1. Late twelfth-century sūtra burial sites
they could be conquered as well. In an age so closely identified with the imperial court, it is significant that the majority of these sites were located outside Heiankyō. Some of these sites were relatively close by, such as Daidōji and Mount Inari to the south, Mount Kurama to the north, and Hatogatake and Mount Hiei to the northeast of the capital.38 Others, such as Makiosan in Izumi Province, Mount Kōya and Kumano in Kii, and Mount Asakuma in Ise, were somewhat farther from the capital. Numerous other Heian-period sūtra burials have been found throughout the northeast from Mount Fuji to as far north as Dewa and Mutsu. By the early twelfth century, sūtra burials were being carried out in every province (map 10.1).39 Western Japan, however, reveals perhaps the most surprising examples. In the late eleventh and early twelfth centuries, when the practice was at its height, more than 60 percent of known sūtra burials took place in northern Kyushu (map 10.2).40 The prominence of sūtra burials in Kyushu is yet another index suggesting that although geographically distant from the imperial capital, Kyushu was more a center than a periphery. The Dazaifu, the government’s outpost in northern Kyushu, indicates the political, military, and economic significance of the region. The power of its local clans, the kingdom of Silla not far from its shores, and the foreign trade afforded by its proximity to the continent were all matters of great concern to the court. Northern Kyushu represented, to the courtiers of Heiankyō, the farthest margins of the state — for some the site of exile but for others a land of riches to be enjoyed by governing aristocrats. The much-sought-after post of “leading official” (kanchō) of Dazaifu serves as an index of the reigning power at the Heian court. During the tenth century and the early eleventh, the post was assigned to followers of Fujiwara no Michinaga; with the ascendancy of retired sovereigns (in) in the late eleventh century, the office was held by a member of the retired sovereign’s circle (in no kinshin).41 The importance of Kyushu to the economic and political history of the Heian period is now becoming increasingly recognized, as shown in Bruce Batten’s essay in this volume. The evidence of sūtra mounds might also reveal that the region was equally significant to the religious culture of the age.
Map 10.2. Sūtra burial sites in Kyushu
256 | d. max moerman
The sūtra mounds of northern Kyushu differ from those of other regions in a number of respects. Many of the tubes housing the sūtras are made of stone rather than bronze, and those made of bronze show a far greater variation in form. But it is the ceramic examples that are most distinctive. Although ceramic containers were used throughout Japan for the outer casings of metal sūtra tubes, the use of Chinese ceramics is unique to the island. Brown-glazed earthenware and porcelain sūtra containers from western China are found nowhere else in Japan (figure 10.3). These ceramic vessels, moreover, were not limited to the outer containers of sūtra tubes. Porcelain stupa-shaped jars were often used on their own, in place of metal reliquaries, to house the scriptures. These were imported from China to Kyushu for the sole purpose of sūtra burials; they were put to such use neither in China nor elsewhere in Japan. They represent but one item in the complex network of exchange, both cultural and economic, between Japan and China during the Northern Song (960 – 1126).42 Heian-period sūtra mounds have been discovered throughout Kyushu, yet the vast majority of these burials — including sūtras inscribed on stone, tile, and copper — have been unearthed in the northern provinces of Buzen and Chikuzen.43 For example, 73 percent of the sūtra burials performed in Kyushu in the century between 1060 and 1160 took place in these two provinces.44 Within this region of northern Kyushu, the three areas with the greatest number of sūtra mounds are Usa, home of the powerful clan of the same name and site of the Usa HachimanMirokuji complex; Dazaifu, the political center of the island; and the mountains extending from Hikosan to the Kunisaki peninsula, which served as the ritual arena of Tendai-affiliated mountain ascetics. All three loci represent points of connection and negotiation between northern Kyushu and the court and will be examined in turn. The Usa Hachimangū, shrine of the Hachiman deities, was originally significant mainly to the local Usa, Ōga, and Karashima clans. Yet in the Nara period, this distant and obscure shrine became essential to the protection and legitimacy of the Nara court,45 which made offerings to the shrines in 737 at a time of military threats from the kingdom of Silla and again in 740 on the occasion of the rebellion of Fujiwara no Hirotsugu (d. 740).46 In recognition of Hachiman’s divine assistance in putting down Hirotsugu’s revolt, the court presented the shrines with gold ink transcriptions of the two scriptures used to protect the state, the Lotus Sūtra and the Sūtra of Golden Light, and granted it ten monks, five horses, and a three-story pagoda. Hachiman was also credited with aiding in the construction of the Daibutsu at Tōdaiji, the central image and central institution of the Nara Buddhist state. Hachiman’s support of the project was recognized through offerings and promotions in rank of the deity and of the shrine’s priests and priestesses. After the Daibutsu was completed, an image of Hachiman, accompanied by a retinue of courtiers and soldiers, was transported from Usa to Nara and was installed as protector of the temple. The replication of the Usa Hachiman at Tōdaiji is emblematic of the larger movement of a regional cult on the geographic periphery of the state to one based at its symbolic center.47 Hachiman became even more closely tied to the fortunes of the imperial court
The Archeology of Anxiety | 257
Figure 10.3. Chinese porcelain sūtra container. Song dynasty (12th century). Excavated from Shitennōjisan sūtra mound, Fukuoka Prefecture. Courtesy ofKyoto National Museum.
in 764 when it appeared that the Hossō monk Dōkyō sought to usurp the throne. Claims were made that an oracle from Hachiman justified the unprecedented action of putting him on the throne. Yet when an emissary of the court traveled to Usa to test the veracity of the claim, Hachiman is said to have proved it false. This was neither the first nor the last time that Hachiman’s oracular powers served to legitimize political authority. In 750 Fujiwara no Otomaro (d. 752) was appointed governor of Dazaifu on the instruction of the deity, and in 939 Taira no Masakado (d. 940) claimed that his revolt was the fulfillment of the deity’s wishes.48 Yet the Dōkyō Incident was perhaps the first time that the legitimacy and indeed the very survival of the political center relied so heavily on what was only a few decades earlier a marginal cult. Usa Hachiman was also a center for the cult of Miroku, linked as it was physically and institutionally with Mirokuji, the temple of Miroku, built beside the Usa shrine in the early eighth century. The late thirteenth-century Hachiman Usagū gotakusenshū (Compendium of the Oracles
258 | d. max moerman
of Usa Hachiman) compiled by Jin’un (1231 – 1314), a scholar-monk of Mirokuji, makes the connection between the Usa deity and the coming of Miroku explicit in declaring that Hachiman’s sacred mountain will “protect the country by housing the sūtras through [the three ages of] shōbō, zōbō, and mappō.”49 Hachiman was granted the title of great bodhisattva (daibosatsu) in the early Heian period, perhaps the first of the Japanese kami to receive such an appellation. The establishment of the Hachiman cult at the center of the Heian state, however, came in the mid-ninth century with the construction of the Iwashimizu Hachimangū in Kyoto and its dedication to the protection of the imperial state. The shrine was built at Otokoyama, the site of what may be, in the opinion of Chijiwa Minoru, the earliest sūtra burials in Japan.50 The remains of the sūtra tiles are fragmentary, but the years of their dedication are clearly recorded: one is dated 958 and the other 960.51 The earliest fully documented sūtra burial was performed by Fujiwara no Michinaga (966 – 1028) on Kinpusen in 1007. The Iwashimizu tiles, dated some fifty years earlier, suggest that the Usa cult from distant Kyushu may have actually provided the inspiration for the practice of the Kinpusen cult in Yamato. Dedicated to the political and religious protection of the Heian state, Dazaifu was the official point of connection between the imperial center and its geographic borderlands. It represented at once the state’s western terminus and the conduit to a greater transmarine East Asia. The ninth-century Montoku jitsuroku calls the Dazaifu “the capital of the western periphery,”52 a phrase that encapsulates the dual qualities of constituting its own political center and serving as a provincial arm of the imperial government. The mountains and temples of this region were the sites of numerous sūtra burials. Indeed, half of the sūtras buried in Kyushu in the Heian period have been excavated from Mount Shiōji and its immediate environs.53 The Shiōji (Temple of the Four [Heavenly] Kings) was built in 775, and statues of these four protective deities were enshrined on the mountain’s peaks in the four cardinal directions. As the name of the mountain and temple suggests, this was a religious site dedicated to maintaining the centralized power of the court. Ever since the founding of Shitennōji in Naniwa, attributed to Shōtoku Taishi after the Four Heavenly Kings helped him vanquish his political opponents, these fierce martial figures represented the interdependence of Buddhism and the state. Like the Hachimangū in Usa, the Shiōji in Dazaifu was a religious establishment for the protection of the country, and as such reveals the institutional and ideological bonds between the political center and the geographic periphery. The mountain temples of the Dazaifu region — Shiōji, Kanzeonji, Buzōji, and Hōmanzan — were the sites of state-protecting rituals, mountain ascetic practices, and numerous sūtra burials concerned with preserving the Dharma until the coming of Miroku. A sūtra mound excavated from the Bishamondō on Shiōjizan’s northern peak housed a steatite image of Miroku twice the size of the freestanding Buddha images usually found in sūtra mounds.54 Two sūtra burials performed at Buzōji, one in 1094 and the other in 1103, contain inscriptions expressing the desire to be present at “the descent of Miroku.”55 Another sūtra mound from
The Archeology of Anxiety | 259
Buzōji contained one sūtra tube with jewels hanging from its lotus-form lid, ceramic jars that housed the tubes, remains of the sūtras themselves, and another sūtra tube, inscribed with the date of 1126 and name of a member of the Fujiwara clan who had received Buddhist ordination.56 A Lotus Sūtra inscribed on ceramic tiles was buried at the Hiei Shrine, a branch shrine of the Mount Hiei deity, adjacent to Kanzeonji, which stood at the foot of Mount Shiōji.57 Hōmanzan to the northeast of Dazaifu protected the region from the baleful forces of that inauspicious direction (kimon shugo no yama), just as Mount Hiei protected the Heian capital. The presence of Tendai institutions and members of the Fujiwara house suggests that the leading religious and political figures of northern Kyushu may not have been so distant from the capital after all. The monk Gikai (870 – 946) exemplifies such continuities. A disciple of Ennin and a member of the Usa clan, Gikai was successively overseer of the Iwashimizu Hachiman shrine and chief abbot of Enryakuji.58 Figures such as Gikai reveal the difficulty in locating centers and peripheries in Heian religious culture and suggest instead the presence of multiple centers with complex patterns of interaction between them. Relations between religious centers and peripheries are perhaps most significant to the third area of sūtra mound construction in northern Kyushu: the region of Mount Hiko, or more accurately, the mountain ranges extending north and northwest from Hikosan and the mountains of the Kunisaki peninsula to the east.59 This region, which included the mountains of Fukuchi, Kokura, Kawara, as well as a number of the previously mentioned religious centers of Daizaifu and Usa, was dotted with temples, Buddha images carved into the sides of cliffs, and numerous sūtra mounds. The monks who undertook austerities in these mountains were organized under the institutional umbrella of Tendai-affiliated Shugendō. The mountain ascetics of Kyushu were thus institutionally linked with the Tendai centers of Mount Hiei, the Kumano mountains on the Kii peninsula, and the temple of Shōgoin in the capital. By the twelfth century these ties had become established economically as well. The Shugendō complex of Kunisaki, also known as Rokugōzan, commended temple lands to Enryakuji, thereby placing themselves under the protection of Mount Hiei. In 1181 Retired Emperor Go-Shirakawa (1127 – 1192) granted income from the domains of Mount Hiko to Ima-Kumano, the Kumano branch shrine in the capital.60 As at Usa Hachimangū, the cult of Miroku was a prominent element at Hikosan. The inscription on a sūtra tube buried there in 1110 refers with anticipation to the “three sermons of Miroku.”61 The forty-nine caves of Mount Hiko were homologized in the Heian period with the forty-nine chambers of Miroku’s inner palace in the Tosotsu heaven.62 According to the Hikosan ruki, a religious history of the mountain written in 1213, the mineral formations found on the mountain were “used to erect the great lecture hall in which Miroku would give his sermons upon becoming the future Buddha” and were themselves said to constitute “Miroku’s metaphoric body, whose function it is to protect the Dharma [throughout the age of mappō].”63 Indicative as well of an anxiety over the permanency of the teachings are two sets of the Lotus Sūtra engraved on bronze plates excavated from Chōanji in Kunisaki and from Kubotesan, northeast of
260 | d. max moerman
Mount Hiko.64 The two sets, made of a material designed to outlast the Final Age itself, were created within a year of each other, the Chōanji sūtra in 1141 and that of Kubotesan in 1142. They also reveal the interrelations between the different cultic centers of northern Kyushu. The bronze for the Chōanji set, for example, was provided by the Rokugōzan of Kunisaki and the Iwashimizu Hachimangū.65 The plates of both sets, moreover, were inscribed by Ki no Shigenaga, head priest of the Usa Hachiman Shrine.66 The Kubotesan set comprises thirty-three bronze plates engraved with the complete texts of the Lotus Sūtra and Heart Sūtra. The sūtra is housed within a gilt bronze box engraved with images of Amida flanked by two bodhisattva attendants on one side, the Buddhas Śākyamuni and Taitō (Skt. Prabūtaratna) of the Lotus Sūtra on another, the fierce guardian deity Fudō myōō (Skt. Acalanātha) on the third, and Bishamonten (Skt. Vaiśravan. a), one of the Four Heavenly Kings, on the fourth. The creation of such an elaborate and expensive object was necessarily a collaborative effort. The dedicatory inscription lists the names of eight individuals including those of the major and minor fund-raisers (daikanjin and shōkanjin) and the scribe (shippitsusō). The individuals involved provide evidence for the close connections that existed between Usa, Hikosan, and Kubotesan. For example, the monk Raigon, who served as the major fund-raiser, hailed from Usa and had traveled to Mount Hiei, where he studied Shugendō and Tendai practices before returning home to restore the provincial temple (kokubunji) of Kubotesan. Raigon had also, two years previously, joined another priest in a transcription of the Lotus Sūtra, each copying alternate chapters, which they then buried. The monk Seijitsu, listed as the project’s minor fund-raiser, had also participated in previous sūtra burials in the years 1111, 1113, and 1141. And Gonson, named as the sūtras’ scribe, was also the copyist of the Hikosan ruki.67
Central Concerns Scholars have offered a variety of motivations for late Heian sūtra burials. Some see the practice linked to the rituals of Shugendō, some to the hope for rebirth in Amida’s or Kannon’s Pure Land, some to the desire for enlightenment and benefits for oneself and others. Nearly all, however, agree that the primary motivation was the concern to preserve the sūtras throughout the age of mappō until the coming of Miroku, who will use the buried sūtras in his three inaugural sermons beneath the Dragon Flower Tree.68 The locations of major sūtra mounds — such as the mountains of Hiei, Kōya, and Kinpusen — were, like Hiko, believed to be the sites of Miroku’s future descent. Moreover, dedicatory inscriptions (ganmon) included with the deposits appear to support the claim that an anxiety over the Final Age constituted the central motivation for sūtra burials in the Heian period. The term mappō appears often in inscriptions from Heian sūtra mounds as if attesting to the timeliness of the rites. It seems to function both as chronological notice and theological rationale for the burials. A Lotus Sūtra inscribed on tile and buried in 1071 at Dainichiji in Tottori Prefecture is dated “the third year of
The Archeology of Anxiety | 261
the Enkyū era in Final Age of the Buddha’s Dharma” (Shaka nyorai mappō Enkyū 3).69 In 1082, thirty years after the calculated beginning of mappō, the monk Jōe buried a Lotus Sūtra that he dated “at the beginning of Śākyamuni’s Final Dharma Age” (Shaka mappō sho). As he wrote in his ganmon, “I have erected a three-shaku[90 cm]-high statue of Miroku on Yasugamine, one of the seven great mountains of Japan, and have transcribed a Lotus Sūtra to be buried there. My prayer is that it will be used when Miroku comes to preach the Dharma beneath the Dragon Flower Tree.”70 Miroku’s role of protecting the sūtras throughout the age of mappō is represented even more literally in another burial from 1071, at Hachigatamine in Hizen Province, in which a large stone image of Miroku himself serves as the sūtra container. Numerous other sūtra burials are dated “in the age” and “in the time of the Final Dharma of Śākyamuni” (Shaka mappō dai; Shaka mappō shi ji).71 Some are even more specific, counting off the exact number of years that have elapsed since the Buddha’s death. The inscription on a sūtra tube buried in 1103 at Shitori Shrine in Tottori Prefecture recognizes its date of Kōwa 5 (1103) as marking “2,052 years after the death of the Buddha Śākyamuni.”72 Although a preoccupation with mappō clearly informs the practice of sūtra burial in the Heian period, the preservationist impulse was not necessarily the sole motivating factor. Inscriptions also express the hope that, as a result of this meritorious act, the donor (or another individual to whom the merit is being transferred) will be reborn, in the interim, in Amida’s Pure Land or in Miroku’s Tosotsu heaven. These two goals, one concerned with the salvation of the Dharma and the other with the salvation of the individual, address the dual challenge of mappō and were often combined in the logic of practice. Even the burial of sūtras inscribed on tile — a medium intended to withstand the test of time — reveals such multiple intentions. Perhaps the most significant burial of tile sūtras was performed in the years 1143 and 1144 at Gokurakuji in Harima Province by six monks under the leadership of Zenne, the abbot (bettō) of the temple.73 Nearly five hundred ceramic tiles were inscribed with some thirty different sūtras and with images of buddhas and bodhisattvas; in addition, various mandalas were interred to last throughout “the ten-thousand-year period of mappō.”74 As Zenne explained, “20,160 years have passed since Śākyamuni entered nirvana, and it is still 5.67 billion years before Jison’s advent.”75 Zenne asked for “tranquility in this life, good health, and longevity...rebirth in the upper realms of the Goku raku Pure Land and presence at the coming of Jison.”76 In addition he prayed for a felicitous rebirth for his ancestors, his teachers, and Retired Emperor Ichijō (980 – 1011); for “the tranquility of the emperor [Konoe]”; and for “the protection of the state.”77 The other five monks involved in the Gokurakuji burial, however, make no mention of Miroku or the chronology of mappō but ask only for rebirth in Amida’s Pure Land for themselves, their teachers, and their parents. Indeed, the salvation of one’s parents was not an uncommon motivation. A burial at Shiōjisan was made in 1116 expressly “for the benefit of my mother”78 and another at Kurodani in Echizen Province in 1157 “so that my father and mother may attain rebirth
262 | d. max moerman
in the Pure Land.”79 Thus even in the high age of mappō consciousness, people buried sūtras to save more than just the teachings. Such multiple intentions, moreover, characterized sūtra burials from the very beginning. The sūtra mound constructed by Fujiwara Michinaga in 1007 at Kinpusen in Yamato Province is usually considered the first documented example of the practice. Kinpusen or “Golden Peak” (also known as Kane no Mitake and Mitake) was a goal for aristocratic pilgrims ever since the abdicated sovereign Uda traveled there in 900. Following Michinaga’s sūtra burial of 1007, it became the site of similar pilgrimages and sūtra burials performed by abdicated sovereigns and prominent aristocrats alike.80 Kinpusen had been a major center for mountain asceticism from the early Heian period and had long been associated with Miroku and his realm.81 The Daigoji monk Sonshi (832 – 909) identified the deity of Kinpusen, Kongō Zaō Gongen, as a manifestation of Miroku in his Reii sōjōe ingi ki of 900,82 and in his Ōmine dōjō shōgon jizai gi of 909 described Kinpusen as the inner realm of the Tosotsu heaven.83 When Kongō Zaō Gongen appears in the Sanbōe of 984, he is guarding the gold of Kinpusen to be used in the world overseen by Miroku.84 The Tale of Genji, a text roughly contemporaneous with Michinaga’s sūtra burial, also identifies Kinpusen with Miroku’s realm. In the “Yūgao” chapter, an old man preparing for a pilgrimage to Kinpusen is overheard praising Miroku (namo tōrai dōshi) while performing repeated full-length prostrations “touching his forehead to the ground.”85 The pilgrimage of 1007 was in fact Michinaga’s second attempt. In 998 he began transcribing sūtras and entered a long period of purifications but was unable to go to Kinpusen, as an epidemic was raging in the capital. His third attempt in 1011 was terminated in much the same way when the death of a dog created a state of impurity forcing him to end his preparations prematurely. The successful pilgrimage of 1007, as one might expect from Michinaga’s other cultural activities, was no simple production. Three years earlier, and one week after completing his transcriptions of the sūtras, Michinaga received ten ritual implements cast in silver used by the mountain ascetics of Kinpusen from the abdicated sovereign Kazan (968 – 1008).86 Michinaga had an elaborate sūtra tube cast and inscribed with a lengthy ganmon and twelve Sanskrit characters praising the Lotus Sūtra (figure 10.4). He then began a seventy-three-day period of purification.87 He left the capital at two in the morning on the second day of the eighth month with sixteen aristocratic attendants and arrived at Kinpusen eight days later. After bathing and purifying himself on the morning of the eleventh he climbed the mountain, stopping to make offerings of silver and silk at the Komori or Mikumari fertility shrines along the way. Once at the central sacred area, Michinaga presented lamps and parasols, one hundred copies of the Lotus Sūtra, one hundred copies of the Benevolent Kings Sūtra, one hundred fascicles of the Heart Sūtra, and eight fascicles of the Essential Meaning of the Heart Sūtra (Daihannya haramitakyō hannya rishubun, T 1695) to “the thirty-eight gods” of the fertility shrine. These dedications were performed for the benefit of the sovereigns Reizei (950 – 1011) (whose consort was Michinaga’s sister Chōshi) and Ichijō (whose consorts included Michinaga’s eldest daughter Shōshi and niece
The Archeology of Anxiety | 263
Figure 10.4. Bronze plate incised with text of the Lotus Sūtra, 1141. Excavated from Chōanji sūtra burial, Oita Prefecture. Courtesy of John C. Weber. Photograph by John Bigelow Taylor.
Teishi), his nineteen-year-old daughter Shōshi, and the crown prince, the future Emperor Sanjō (son of Reizei and Chōshi).88 Michinaga then dedicated a set of eight scriptures in fifteen rolls that he had copied out himself in gold ink. These are listed in the ganmon as including “one copy of the Lotus Sūtra in eight rolls together with the Sūtra of Innumerable Meanings and the Sūtra of Meditation on the Bodhisattva of Universal Virtue, and one copy each of the Essential Meaning of the Heart Sūtra, the Amida Sūtra, the three Miroku sūtras, and the Heart Sūtra.”89 Michinaga described these actions as “burying the relics of the dharma body...in anticipation of the dawn of Miroku’s age.”90 Michinaga worshiped both Amida and Miroku. As stated in his ganmon, “The
264 | d. max moerman
Amida Sūtra promises that one who calls on Amida on one’s deathbed will be reborn in his Gokuraku Pure Land, [and] the Miroku sūtras allow one to avoid an inauspicious rebirth and to be received at Jison’s advent.”91 Michinaga asked to be reborn in Amida’s Pure Land, but only until Miroku’s advent: “When Jison becomes a buddha, may I journey from the Gokuraku realm to the place of Miroku Buddha, listen to his lectures on the Lotus Sūtra, and attain buddhahood.”92 Michinaga prayed that at that future time his “buried sūtras would spontaneously well up out of the earth”93 — like the jeweled stupa and the myriad bodhisattvas in chapters eleven and fifteen of the Lotus Sūtra — to be used by Miroku in his inaugural sermons. The multiple motives for Michinaga’s burial exceed even the explicitly stated soteriological equivalent of having one’s cake and eating it too. The glory and preservation of Fujiwara Michinaga was after all a family affair; it relied on his daughter’s production of an imperial heir. By the year 1007 Shōshi had been Ichijō’s consort for eight years but had yet to conceive a child. If Shōshi were to produce a son he would become crown prince, and Michinaga would become grandfather of an emperor. Unsurprisingly the kami of Kinpusen’s Mikumari/ Komori shrines were worshiped as child-granting and child-protecting deities. Kongō Zaō Gongen, the deity of Kinpusen and manifestation of Miroku to whom Michinaga addressed his prayers, was also in the Heian period the object of a cult for the prosperity of descendants. Michinaga seems to have been successful in attaining at least one of the goals of his sūtra burial because Shōshi became pregnant the following year and, according to the Eiga monogatari, Michinaga “thought that this must be the miraculous result of his pilgrimage to the Sacred Peak.”94 As her pregnancy progressed, Michinaga “addressed a stream of prayers and vows to the Sacred Peak, begging for his daughter’s safe delivery.”95 Michinaga may also have imagined himself to be an incarnation of Miroku, who was closely associated in Japan — as in Korea, China, and India — with the figure of the universal Buddhist sovereign (Skt. cakravartin). Early sūtra burial donors, such as Michinaga and those that followed for close to two centuries, expressed a wide range of desires. They prayed for the salvation of the Dharma, themselves, and their family members. Michinaga dedicated the merit from some of his pious exertions on Kinpusen to his brother-in-law Reizei, his son-in-law Ichijō, his daughter Shōshi, and his nephew the crown prince. Thus although the sites, scriptures, and at times the ganmon themselves indicate an anxiety over the death of the Dharma, the death of the individual received an equal if not greater degree of attention. Fujiwara no Moromichi (1062 – 1099), for example, the great-grandson of Michinaga, followed his forebear’s practice of burying sūtras on Kinpusen. In 1088 Moromichi dedicated a large number of memorial transcriptions to an equally large number of family members.96 In the colophon of his gold ink copy of the Lotus Sūtra, Moromichi, who at the time was suffering from an earache, reveals some of his motivations: In copying this sūtra during the period of ritual purification for my pilgrimage to Kinpusen, I pray for the purification of my inner ear, one of the Six Roots,
The Archeology of Anxiety | 265
and thinking of the importance of the daughters of this house, hope that the merit of [copying] the One Vehicle of the Lotus will provide them with the karmic bond to be present at the three sermons beneath the Dragon Flower Tree.97 Another, longer ganmon included in the sūtra mound speaks also to a concern with the future glory of Miroku and of Moromichi’s lineage as well. The final section begins with Moromichi’s statement I have copied out the Threefold Lotus Sūtra, the Heart Sūtra, and the Diamond Life-span Sūtra [Kongō jumyō darani nenju kyō, T 1134b] by hand in gold letters and buried them at Kinpusen in a bronze vessel in order to advance the noble teachings of the One Vehicle of Shaka and to establish the karmic bond to be present at Jison’s three assemblies. With faith that these offerings will surely enjoy the longevity of metal and stone, I present them to Kongō Zao [Gongen], with reverence for his miraculous powers, and to the [fertility] deities of the thirty-eight sites. Yet it concludes with a prayer for those born into this hereditary house to quickly rise to the third rank, for the past karma of its deceased fathers and grandfathers, and for the prosperity of its descendants.98 To these ends Moromichi offered one copy of the Diamond Life-span Sūtra for the reigning sovereign Horikawa and his consort; one copy each of the Benevolent Kings Sūtra for the longevity and prosperity of his brother-in-law (the retired sovereign Shirakawa), his sister Fujiwara no Kenshi, and their four sons; ten copies of the Lotus Sūtra, five copies of the Benevolent Kings Sūtra, and one hundred rolls of the Diamond Life-span Sūtra for the longevity and prosperity of his father, Fujiwara no Morozane; five copies of the Lotus Sūtra, five copies of the Benevolent Kings Sūtra, and one hundred rolls of the Diamond Life-span Sūtra for the health and longevity of his mother, Fujiwara no Reiko; three copies of the Benevolent Kings Sūtra and one hundred rolls of the Diamond Life-span Sūtra for his son Fujiwara no Tadazane; and for his wife, Fujiwara no Hiroko, one roll each of the Kannon Sūtra, the Essential Meaning of the Heart Sūtra, the Heart Sūtra, the Diamond Life-span Sūtra, the Sūtra of the Eight Secret Dhāran. īs (Hachi myō fumitsu darani kyō, T 1365), and the Sūtra of the Eight Spells of Heaven and Earth (Tenji hachiyō shinju kyō, T 2897) together with five copies each of the Lotus Sūtra, the Benevolent Kings Sūtra, and the Diamond Life-span Sūtra. These offerings were all in addition to his personal gold ink transcriptions of the Lotus Sūtra, the Heart Sūtra, and the Diamond Life-span Sūtra. The disproportionate range of sūtras that Moromichi dedicated to his wife suggests that the principal reason for this sūtra burial was, like that of his grandfather Michinaga, to pray for the birth of descendants. Hiroko had given birth to one son, Tadazane, in 1078, and another consort had produced a second son two years later. But Moromichi was still without a daughter to marry into the imperial line, which was essential
266 | d. max moerman
to the Fujiwara strategy for maintaining their political power. Moromichi explained as much in his ganmon: “I am a young man in the prime of my life and yet I have not been blessed with many children. This I bemoan. My prayer is that I might have another.”99 The efficacy of Moromichi’s prayer, unlike that of his grandfather, was not immediately felt. He had to wait another ten years for the birth of his next child, and, while it was the daughter he sought, she was to die young at the age of only ten.
Beyond Centers and Peripheries As the examples of Michinaga, Moromichi, and Zenne suggest, the motivations for Heian sūtra burials were, like their contents and locations, so various that no single explanation can be meaningfully applied to all. The stated intentions of the donors exceeded a concern with the salvation of the Dharma to include the salvation of oneself and of one’s family members both living and dead. This range of religious desire is reflected in the scriptures chosen and the goals to which they were directed: rebirth in Miroku’s Tosotsu heaven or Amida’s Gokuraku Pure Land. Gokuraku rebirth was seen not as a means of final escape but as an intermediary stage in a larger eschatological plan: a place for the fortunate to wait before returning to earth to attend Miroku’s sermons ushering in the next age. Other donors who asked for rebirth in the Tosotsu heaven understood their goal also as a temporary station from whence to descend with Miroku in the far distant future. There remains, however, a fundamental tension between these two motives, a difference in the way they approach history. The preservationist aspect of sūtra mounds, saving the Dharma in its material forms for the Future Buddha, represents an act of historical responsibility: an investment in the future. The advent of Miroku’s golden age cannot be accelerated; its eventual appearance after a long period of decline can only be prepared for. The other intention of sūtra burials, saving oneself and one’s family through immediate future rebirth in either Amida’s or Miroku’s paradise, follows a different model entirely. Although retaining a this-worldly emphasis (the petitioner would continue to amass spiritual capital to assure his future reward) its goal has become “severely dehistoricised.”100 Buddhism’s cosmological timetable, the grand historical model of which mappō is part, is circumvented. The Pure Land path of personal salvation with its eschatology of the immanent seems to obviate the need for institutional preservation. This divergence is borne out in the sūtra mounds themselves. A preoccupation with preserving the sūtras throughout the age of mappō in anticipation of Miroku’s advent was largely limited to the Heian period.101 Sūtra burials after the Heian period rarely mention Miroku’s age and are more directly related to the fate of the individual after death. As their ritual function changed from preservation to memorialization, devotion to Amida came to replace the cult of Miroku. In Japan, as in China, Miroku’s paradise cult was absorbed and superseded by that of Amida, and the transformation of sūtra burials may have been part of
The Archeology of Anxiety | 267
this larger religious shift.102 Rather than the Miroku iconography of early sūtra mounds, later burials exhibit far more pronounced Amida imagery. The engraved mirrors and hanging bas-reliefs (kakebotoke) that often accompanied the sūtras were decorated with raigo — scenes of Amida descending to welcome the dying into his Pure Land — rather than portraits of the future Buddha. Yet as we have seen, sūtra burials from their earliest examples were concerned with the postmortem salvation of both the religion and the religionist. The Pure Land faith of the Heian period was not limited to a single temporal orientation. Concern with a future rebirth, nostalgia for a past golden age, and visions of a paradise in the present world were in no way mutually exclusive. Such multiple intentions may explain the combined presence of the Miroku and Amida cults as well as the ambiguous place of mappō in the sūtra burials. For although presented, both implicitly and explicitly, as the ostensible reason for the practice, mappō appears on closer examination to have functioned more as a rhetorical center on which other personal, familial, and political anxieties converged. If the practice of sūtra burial reveals anything about the role of mappō thought in the Heian period, it is the range of concerns that are contained within this discourse. As an umbrella term, mappō is able to embrace a variety of religious desires while at the same time charging them with a heightened sense of historical urgency. The practice of sūtra burials raises questions not only about the soteriological centrality of mappō in Heian religious culture but also about the location of that culture’s geographical center. If sūtra mounds indicate a preoccupation with the Final Age, however widely we may define this concept, it was a preoccupation felt beyond the boundaries of the capital. The large number and early appearance of sūtra mounds in northern Kyushu suggest that this distant site may have figured more centrally to the history of Heian religion than has often been recognized. Preparation for the future age of Miroku was a major concern in the region; and the practices, individuals, and institutions involved in these religious productions reveal the intricate network of relations between the capital of Heian-kyō and the “capital of the western periphery.” In the Final Age, history and salvation presented a formidable set of interrelated problems. To many monks and aristocrats of the Heian period, the practice of sūtra burial provided a solution of sorts. It offered the ritual strategies and material means whereby the end-time could be prepared for and paradise secured. For scholars of Heian Japan, however, the evidence of sūtra mounds can address another set of problems and provide another kind of buried treasure. Sūtra mounds, like so many time capsules, constitute a unique kind of archive for a geography of the Heian imagination. They identify the desires, the individuals, the cults, and perhaps most significantly the sites central to the practice of Heian religion. As such they may offer a map to a new sort of history, a spatial history that might begin to explore the vast and less-charted landscape that lies between and beyond the dichotomy of center and periphery.
268 | d. max moerman
Notes 1. Taira Masayuki, “Mappō matsudaikan no rekishiteki igi,” 110 – 154. 2. See Etienne Lamotte, History of Indian Buddhism, 191 – 202; David Chappell, “Early Forebodings of the Death of Buddhism,” 122 – 154; Jan Nattier, Once upon a Future Time: Studies in a Buddhist Prophecy of Decline. 3. The first description of the tripartite system appears in the Li shiyuan wen (in Taishō shinshū daizōkyō [hereafter T], 46.786c), attributed to Nanyue Huisi (515 – 577), teacher of the Tiantai patriarch Zhiyi (538 – 597). See Nattier, Once upon a Future Time, 110 – 111. The chronology also played a crucial role in the Pure Land thought of Daochuo. Such a provenance may help to explain the importance of mappō thought in the Tendai and Amidist traditions. 4. Although Japanese scholars have long assumed that mappō was a belief and a term (saddharma-vipralopa) found in Indian Buddhism, Nattier presents a compelling argument for its Chinese origins. See Nattier, Once upon a Future Time, 91 – 118. 5. Taira, “Mappō matsudaikan,” 144. 6. For a survey of mappō thought in the Heian period, see Michele Marra, “The Development of Mappō Thought in Japan (I),” 25 – 53. 7. Fusō ryakki, Eishō 7 (1052) 1/26. 8. Marra, “Development of Mappō Thought,” 39 – 40. 9. Sanbōe, in SNKBT, 31:4; Edward Kamens, The Three Jewels, 92. 10. Ishida Mizumaro, ed., Genshin, Ojōyōshū, 10. 11. In the early Kamakura period, the Tendai prelate Jien (1155 – 1225) argued for just such an understanding and thus the need for a political response to mappō. Jien also buried numerous copies of the Lotus Sūtra at the Iwashimizu Hachiman, Kamo, Kasuga, and Hiyoshi (Hiei) shrines (Gyokuyō, Juei 1 [1182] 9/14). 12. See, for example, Shunki, Eijō 7/8/28; and Chūyūki, Eichō 1 (1096) 10/28. 13. For an example of such blood writing, see Nara kokuritsu hakubutsukan, ed., Kyōzuka ihō, pl. 60. 14. Genji monogatari, in SNKBT, 22:163 – 164; Edward G. Seidensticker, trans., The Tale of Genji, 713. 15. Eiga monogatari, in SNKBZ, 32:349; William H. McCullough and Helen Craig McCullough, trans., A Tale of Flowering Fortunes, 2:593. 16. See Gregory Schopen, “The Phrase ‘sa prthivipradesas caityabhūto bhavet’ in the Vajracchedikā,” 147 – 181. 17. T 9.262.31b; Leon Hurvitz, trans., Scripture of the Lotus Blossom of the Fine Dharma, 178. 18. J.Edward Kidder, Early Buddhist Japan, 140. This is not to suggest that the cult of relics was unimportant in Heian Japan. On the significance of the cult, see Brian Ruppert, Jewel in the Ashes. 19. T 9.262.30b. Hurvitz, Scripture, 174. 20. T 9.262.31a. Hurvitz, Scripture, 176. 21. T 9.262.31a. Hurvitz, Scripture, 177. 22. T 9.262.54c. Hurvitz, Scripture, 300. 23. T 9.262.61c. Hurvitz, Scripture, 335. 24. On the important, and much debated, image at Hōryūji, see Christine M.E. Guth, “The Pensive Priest of Chūgūji,” 191 – 213. 25. Janet Goodwin, “The Worship of Miroku in Japan,” 41 – 60. The colossal eighth-century image of Miroku at Kasagidera, carved directly onto the cliff face, is discussed in Karen L. Brock, “Awaiting Maitreya at Kasagi,” 214 – 247.
The Archeology of Anxiety | 269 26. Cited in Goodwin, “Worship of Miroku,” 67. 27. Ibid., 69. Although the Lotus Sūtra addressed the promise of Tosotsu rebirth to men, Miroku’s heaven was a goal for both female and male devotees. 28. Quoted in Akira Matsumoto, “Kōbō Daishi no nyūjō setsuwa seiritsu no haikei,” 309. 29. Allan Andrews, The Teachings Essential for Rebirth, 51. 30. On the Miroku faith of one leading Heian courtier, Fujiwara no Munetada, see Hiraoka Jōkai, Nihon miroku jōdo shisō tenkai shi no kenkyū, 550 – 553. 31. These are the jōza sanmai (constant sitting), jōgyō sanmai (constant walking), hangyō hanza sanmai (half-sitting, half walking), and the higyō hiza sanmai (neither walking nor sitting). 32. Genkō shakusho, 3:62. On Ennin’s transcription and the copying procedures of later Heian examples, see Seki Hideo, Kyōzuka to sonno ibutsu, 29 – 30; Willa J. Tanabe, Paintings of the Lotus Sūtra, 44 – 49; Kawada Sadamu, “Decorative Art,” 169; and Miyake Toshiyuki, “Sūtra Mounds,” 173 – 174. The ten kinds of offerings (to a Buddha) listed in the Lotus Sūtra are flowers (ke), incense (kō), ornaments (yōraku), powdered incense (makkō), unguent (zukō), burning of incense (shōkō), canopies and banners (sōgai dōban), clothing (ebuku), dancing and music (gigaku), and joining one’s hands in worship (gasshō). 33. Eigaku yōki, 549b. 34. Kyōzuka ibun (hereafter KI), 120a, no. 241. 35. KI, 157a, no. 327. 36. See the chronological table in Seki Hideo, Heian jidai no maikyō to shakyō, 709 – 739. 37. These extrascriptural materials are usually interpreted as representing the donor’s concern with the protection of the sūtras. See, for example, Sekine Daisen, Mainōkyō no kenkyū, 281. Chijiwa Minoru, however, has suggested that they may also represent a form of offering directed more toward the kami than the buddhas. See his “Hachiman shinkō to kyōzuka no hassei,” 444. 38. On the sūtra burials at Mount Inari, see Fushimi Inari Taisha Shamusho, Inarisan kyōzuka. For those at Kurama, see Hosaka Saburo, Kuramadera kyōzuka ibutsu, and Naniwada Tōru, Kuramadera kyōzuka ihō oboegaki. On sūtra burials in Fukushima, Tochigi, Yamanashi, Aichi, Mie, Gifu, Yamaguchi, and Ehime prefectures, see Miyake Toshiyuki, Kyōzuka ronkō, 69 – 93, 117 – 144, and 178 – 206. 39. For the chronology and locations of these sūtra mounds, see Seki, Kyōzuka to sono ibutsu, 37 – 53. 40. Chijiwa, “Hachiman,” 426. For example, 173 sūtra burials (of known location) were performed in the hundred years between 1064 and 1163. Of these, 104 took place in Kyushu (Seki, Heian jidai no maikyō to shakyō, 710 – 724). 41. Bruce Batten, “State and Frontier in Early Japan,” 248 – 249. 42. Nara kokuritsu hakubutsukan, ed., Kyūshū chihō ni maikyō sareta yakimono, 4. As the gateway to Japan from the continent, Kyushu was the first area to receive and incorporate Korean and Chinese forms of knowledge and practice. Yet the influence of Chinese or Korean Buddhist practice on Japanese sūtra burials is unclear. No sūtra mounds have been discovered in China, and the one example on the Korean Peninsula, a fragment of a Lotus Sūtra transcribed on paper and the remains of what might be a sūtra receptacle, remains inconclusive. Miyake Toshiyuki “Sūtra Mounds,” 172. For further discussion of the cultural and economic relations between Japan and the continent in the Heian period, see the essays in this volume by Bruce Batten and Robert Borgen. 43. Nara kokuritsu hakubutsukan, Kyōzuka ihō, 338. 44. Oda Fujio, “Kyūshū no kyōzuka,” 133. See also Chijiwa, “Hachiman,” 433. 45. See Ross Bender, “The Hachiman Cult and the Dōkyō Incident,” 125 – 153.
270 | d. max moerman 46. Shoku Nihongi, Tenpyō 9 (737) 4/1 and Tenpyō 13 (741) int. 3/24. Ōita Kenritsu, Hachiman daibosatsu no sekai, 36. Takahashi Mitsugu, “Kojiki, Nihon shoki, Shoku Nihongi ni mieru Usa Hachiman,” 185. 47. Allan G. Grapard, “Lotus in the Mountain, Mountain in the Lotus,” 25. 48. Bender, “Hachiman Cult,” 137 and 153. On the oracles of Hachiman, see Allan G. Grapard, “The Source of Oracular Speech,” 77 – 94. 49. Cited in Chijiwa, “Hachiman,” 433. See Shigematsu Akihisa, ed., Hachiman Usagū gotakusenshū. For a discussion and analysis of this text, see Grapard, “Lotus in the Mountain, Mountain in the Lotus.” 50. Chijiwa, “Hachiman,” 426 – 428. 51. Heian ibun (hereafter HI), 12:70 – 71, no. 57. 52. Nihon Montoku tennō jitsuroku, Ninju 2 (852) 2/8; Batten, “State and Frontier,” 139. 53. Oda, “Kyūshū,” 133. Nara kokuritsu hakubutsukan, Kyūshū chihō ni maikyō sareta yakimono, 9. For examples, see KI, 49b, no. 84; and HI, 12:189, no. 195. 54. It stands 62.7 cm. high while few other such images exceed 30 cm. Sekine Daisen, Mainōkyō no kenkyū, 274. 55. KI, 21a – b, no. 33, and 27b – 28a, no. 42. 56. KI, 60, no. 109. 57. Oda, “Kyūshū,” 136. 58. Gikai was appointed head administrator (kengyō) of Iwashimizu Hachiman in 939 and became the fourteenth Tendai zasu in 940. See Chijiwa, “Hachiman,” 439. 59. See Allan G. Grapard, “Geotyping Sacred Space,” 215 – 249. Indeed, the very notion of religious centers and peripheries might need to be reconceived in the context of the Japanese islands when one considers that the early Buddhist sites of northern Kyushu had more in common with those on the Korean Peninsula than those of the Kinai. 60. Go-Shirakawa in no chō kudashibumi an, Yōwa 1 (1181) 12/8, in Imakumano kiroku. 61. KI, 32, no. 52. 62. Grapard, “Geotyping,” 225. 63. Hikosan ruki, 463. Cited in Grapard, “Geotyping,” 225. 64. The Hikosan ruki mentions a third copper plate Lotus Sūtra engraved at Hikosan in 1145. Seki, Heian jidai no maikyō to shakyō, 720. 65. Ōita Kenritsu, Hachiman, 36. 66. Shigenaga was also, according to the Hikosan ruki, responsible for carving the Hikosan plates. 67. Nara kokuritsu hakubutsukan, Kyōzuka ihō, 345 – 347, 478, and pl. 32.1 – 32.4 68. See, for example, Ishida Mosaku, Kōkogaku kōza, 30 – 31; Yajima Kyōsuke, Kyōzuka, 2 – 3; Gotō Shuichi, Nihon rekishi kōkogaku, 702; Kurata Osamu, “Maikyō,” 14 – 21; Hosaka Saburō, “Sotoba to kyōzuka,” 73; Miyake Toshiyuki, “Kyōzuka zōei ni tsuite,” 47. 69. HI 12:127, no. 123. 70. HI 12:131, no. 130. 71. HI 12:357 and 394, nos. 407 and 449. 72. HI 12:161 – 162, no. 163; KI, 83 – 84, no. 168. 73. HI 12:265 – 278; Taira, “Mappō matsudaikan no rekishiteki igi,” 115 – 118. 74. HI 12:269. 75. HI 12:268. 76. HI 12:274. 77. HI 12:270. 78. KI, 44b, no. 74. 79. KI, 117b, no. 233.
The Archeology of Anxiety | 271 80. For example, Fujiwara no Yorimichi traveled there in 1049, the retired sovereign Shirakawa and Fujiwara no Moromichi both in the years 1088 and 1092, and Minamoto no Masakane in 1103 and 1108. 81. Miyake Hitoshi, Ōmine shugendō no kenkyū, 15. 82. Shugendō shōso, 1:80. On this text, see Miyake Hitoshi, ed., Shugendō shōso kaidai, 139 – 140. 83. Shugendō shōso, 1:62; Miyake Hitoshi, Shugendō shōso kaidai, 136 – 137. 84. Sanbōe, in SNKBT, 31:195; Kamens, Three Jewels, 328. 85. Genji monogatari, 14:141; Seidensticker, Tale of Genji, 68. 86. Midō kanpaku ki (hereafter MKK), Kankō 1 (1004) 5/21. 87. MKK, Kankō 4 (1007) 5/17. Although Michinaga’s ganmon, inscribed on the exterior surface of the sūtra tube, refers to one hundred days of purification, the standard period lasted around seventy. Ritual preparations for pilgrimages to Kinpusen could last for twenty-one, fifty, or one hundred days. Even the shortest of these, however, was longer than those for Kumano or Kōya, which rarely lasted more than a week. Michinaga began the rites on the seventeenth day of the fifth intercalary month and ended them on the first day of the eighth month. Michinaga’s ganmon is reproduced in KI, 6b – 8a, no. 10, and in HI 12:89 – 90, no. 86. 88. MKK, Kankō 4/8/2 – 11. 89. KI, 6b. 90. KI, 7a. 91. KI, 7a. 92. KI, 7a. 93. KI, 7a. 94. Eiga monogatari, in SNKBZ, 31:388; McCullough and McCullough, trans., Flowering Fortunes, 1:265. 95. SNKBZ, 31:391; McCullough and McCullough, Flowering Fortunes, 1:267. 96. Go-Nijō Moromichi ki, Kanji 2 (1088) 7/1. This was the first of two sūtra burials on Kinpusen. Moromichi journeyed there again two years later. 97. KI, 20b, no. 30. 98. KI, 19b – 20a, no. 29. 99. KI, 19a. 100. Jan Nattier has analyzed these two models as the “here/later” and the “there/later” in “The Meanings of the Maitreya Myth,” 23 – 47. 101. Kurata Osamu, “Maikyō,” 250 – 253. 102. Joseph Kitagawa, “The Career of Maitreya, with Special Reference to Japan,” 246.
part iv
Negotiating Domestic Peripheries
11
d William Wayne Farris
Famine, Climate, and Farming in Japan, 670 – 1100
C
rop failure and famine have a long history in Japan. Yet the topic has received virtually no attention in English-language research on the ancient period, meaning here 670 – 1100, nor have Japanese historians specializing in those years systematically analyzed it. This chapter will address three basic and seminal questions about food shortages in that era. First, how frequent and severe were they? The story of these crises in the early modern or Tokugawa period (1600 – 1868) is well-known, and many believe that they had widespread demographic, social, and political effects.1 Can the same be said for the ancient period? Second, what were the causes of crop failure? Consistent with this query, did ancient farming techniques enhance or reduce the chances of famine? Third, what were the responses to starvation, both governmental and social? Did they vary by region, most particularly between the capital (or core) and the hinterlands (periphery)? Were these responses effective?
An Overview of Frequency and Causes Except for intermittent coverage usually tied to elite political events, Japanese historians focusing on the ancient period have generally neglected the problem of food supply. Fortunately, the nature and impact of famine have become a hotly debated topic among scholars studying the early modern epoch, and one of them, Saitō Osamu, has recently conducted a fairly thorough study of starvation during the entire premodern era (see table 11.1 and table 11.3). For the period under consideration, two conclusions seem readily apparent: first, far more hungry years were recorded for the late seventh, eighth, and ninth centuries than the tenth and eleventh. During the heyday of the Chinese-style state, famine appears to have stalked the land approximately once every three years, while in the tenth and eleventh centuries food shortages were much more rarely noted. Second, drought was a more usual cause than cold, wet weather. Of the 125 famine years listed in table 11.3, forty-six, or 36.8 percent, resulted from drought, with thirty-three (26.4 percent) being due to cold, wet weather, and the rest unknown. More important, of the seventy-one years for which widespread famines were chronicled, over half (thirty-eight) arose from drought, two and one-half the number (fifteen) for cold, rainy summers. How are we to explain each of these patterns? First, there is the dramatic decline in the reported incidence of severe food shortages after 900. A careful observer may draw one of two conclusions: that the extremely low numbers for
276 | william wayne farris
Table 11.1. The incidence of famine in Japan by century, 600–1900 Century 7th 8th 9th 10th 11th 12th 13th 14th 15th 16th 17th 18th 19th
Incidence of famine (W)
(L)
Weighted total
6 28 25 3 6 8 7 10 14 8 11 9 4
2 16 24 5 5 3 3 2 12 22 3 3 4
7 36 37 5.5 8.5 9.5 8.5 11 20 19 12.5 10.5 6
Sources: Saitō, “Frequency of Famines,” 26. (W) stands for widespread, (L) for local, in Saitō’s view. The weighted total was calculated by assigning a value of one to a general famine, one-half to a local outbreak. Because totals for the tenth and eleventh centuries were modified using Sasaki’s database (Sasaki Junnosuke, ed., Nihon chūsei kōki kinsei shoki ni okeru kikin to sensō no kenkyū, 9–14) as well as Shōyuki, vols. 2, 4, and 7, my figures for those centuries differ somewhat from those of Saitō. See table 11.3.
the ninth and tenth centuries represent reality or that many hungry years went unreported. If we adhere to the former hypothesis, then the tenth and even eleventh centuries would appear to have been halcyon days of peasant plenty, during which time agriculture could have been improving rapidly throughout the islands. Although he does not deal with famine, Dana Morris adopts this position in his essay in The Cambridge History of Japan. In his view, wider diffusion of the plow allowed farmers to turn soil more deeply, clear grain stalks for roughage, and convert remaining “wasteland” to tillage. In addition, rice was more often harvested at ground level, and more peasants came to possess livestock that were stabled year-round. Manure became available for fertilizer, increasing yields. Abandoned or “fallowed” fields virtually disappeared, in what Morris has termed a dramatic transformation of Japanese agriculture taking place between 800 and 1100.2 Unfortunately, only one aspect of Morris’ interpretation finds support in the historical record. Kōno Michiaki has recently used written, artistic, and material sources to suggest that the plow did indeed become more widespread during the mid-Heian period (900 – 1100), about the same time that Morris proposed.3
Famine, Climate, and Farming in Japan | 277
However, Kōno limits ownership of livestock and the new plow, which could cut the soil only about three centimeters more deeply than a peasant’s hoe, to wealthy peasants known as tato or rikiden no yakara (independent cultivators).4 In addition, Morris’ ideas about the stabling of livestock and the use of manure are highly conjectural. In fact, manure was rarely used as fertilizer even in the Kamakura era.5 Moreover, untended fields comprised between 20 and 30 percent of most estates as late as the thirteenth century.6 Finally, in recent years as archaeological excavations of lands cultivated during the Heian period have multiplied, scholars have made an interesting discovery.7 Beginning in the late tenth century and continuing into the second quarter of the eleventh, fields became barren, and estates and provincial arable disappeared across both eastern and western Japan. No one has yet made a convincing argument regarding the causes of this dislocation, although archaeologist Uno Takao attributes it to the decline of one local elite and the rise of another.8 Coincidentally, the field pattern known as the jōri sei underwent substantial reorganization during the same decades.9 It is difficult to see how Morris’ purported improvements in agricultural technology, expansion of arable land, or rise in yields could have taken place at the same time that wasteland was increasing dramatically. Another possibility may have also left peasants with more food: tax collectors may have taken less, leaving more of the surplus for cultivators. This also seems unlikely, however, as the tenth and eleventh centuries marked the height of the custodial governors (zuryō), known for their rapacity. We have only to recall the famous Protest of the District Magistrates and People of Owari, dated to 988, treated in more detail by Charlotte von Verschuer in this volume. There Governor Fujiwara no Motonaga is berated for collecting all manner of “excessive” taxes; he even falsified land records to claim taxes for unproductive (or nonexistent) fields.10 In their analysis of the tax system of the period 900 to 1100, Nishiyachi Seibi and Sakamoto Shōzō have argued that the basic tax rate remained about as high as in the eighth and ninth centuries.11 If the reduction in lean years during the mid-Heian period had little to do with agrarian improvements or a lightened tax burden, then the only remaining conclusion is that the decreased volume and altered nature of primary source materials resulted in a significant understatement of the frequency of famine during 900 – 1100.12 Certainly the data for the centuries both before and following those two hundred years immediately raise strong suspicions. Is it reasonable, for instance, to believe that agriculture could have improved so dramatically between 889 and 916 that the incidence of famine would have dropped from more than once in three years in the ninth century to about once in twenty in the tenth? Moreover, because Saitō relied on prewar sources, his table significantly understates the frequency of hunger in the early medieval age (1100 – 1300). When I updated Saitō’s list of major famine years for the twelfth and thirteenth centuries according to a database recently constructed by Sasaki Junnosuke, I uncovered five more years for the twelfth century and six more for the thirteenth.13 Because both the well-documented periods immediately preceding and following the
278 | william wayne farris
tenth and eleventh centuries contain many more years of crop failure, it would stand to reason that the low figures for the mid-Heian period are due to problems with the sources. Heian historians are well aware of the perils of ninth- and tenth-century evidence. Most famines recorded for the eighth and ninth centuries were described in the famed Six National Histories (Rikkoku shi), which end in 887. To be sure, there are other, more laconic annals, but they do not match the Rikkoku shi in completeness or detail. Turning to documents, scholars know that the eighth century has the huge Shōsōin cache of about 10,000 fragments (25 volumes in Dai Nihon komonjo). The tenth century, on the other hand, known as the most poorly documented century in Japanese history, comprises only about 268, or 5 percent, of the records in Heian ibun, a generally inclusive compilation encompassing about 5,350 items. The eleventh century fares better, with 1,095 (20 percent), still a disproportionately small number for the entire 391 years of the Heian era. When we remember that there are virtually no aristocratic diaries extant until the late tenth century, and that these sources rarely address problems outside the capital, we realize that the sudden dearth of data on famines after 900 probably arises from changes in the sources. It should come as no surprise that scholars have concentrated on elite politics, literature, and art for the epoch 900 – 1100, because the record for other phenomena is so poor. This condition — the lack of documents, especially for the provinces — may well be due to random factors. For long bygone eras, the written and material sources handed down to posterity are preserved primarily by accident. At the same time, however, since the age from 900 to 1050 represents the first stage of Sakamoto Shōzō’s dynastic state (ōchō kokka), when the central government had little interest in the periphery beyond what taxes could be extracted, the pattern of survival makes a little more sense.14 By the tenth century, the court had largely abandoned close supervision of provincial affairs, turning them over instead to the custodial governors as tax farmers. Gone were the days of census taking, land allocation, and head taxes, all of which institutions presupposed a thorough control of the populace and productive land. The district magistrates (gunji), who had ably assisted the capital aristocracy in the seventh and eighth centuries, began to die off or lose their power, and the post-900 capital elites fobbed off the costs of local government on wealthy or powerful provincial families.15 The second major conclusion to be drawn from table 11.3 is the centrality of drought as a cause. This phenomenon seems to have been especially dramatic for the eighth and ninth centuries, when 60 to 70 percent of general famines for which an origin can be surmised resulted from the lack of rainfall. Furthermore, almost all crop failures stemming from cold, wet summers were unusual in the eighth century and were concentrated in the 770s, while in the ninth century the 860s and 870s contained almost all such shortages. Drought certainly seems to have been a recurrent menace, but is there any way to determine just how dry these centuries were? It is necessary to consider several pieces of evidence. For one thing, it may be helpful to examine modern patterns of precipitation, as they may overlap with
Famine, Climate, and Farming in Japan | 279
the most drought-ridden provinces. Japan is, of course, a wet country, but rainfall varies considerably according to region. Generally, experts maintain that the eastern Tōhoku, Nagano, and eastern Inland Sea littoral are the drier areas, with Kyushu (especially the southern section), southern Shikoku, south-central Honshu, and the Japan Sea coast being the wettest. This information, however, has at least two drawbacks: it is uncertain how applicable modern weather data are to the distant past, and not all precipitation occurs during the growing season. In terms of rainfall from April through July, for example, Hiroshima and Fukuoka are nearly the same, and Nagoya and Fukui are not far behind. Both Osaka and Tokyo are more moist than Akita.16 When we map the geographical extent of ancient droughts against this information, there seems to be no easily discernible pattern. Areas where we would expect large amounts of rainfall during the growing season, such as northern Kyushu, the provinces along the Inland Sea, the northern coast of the Japan Sea, or Sagami and Musashi in the Kanto, appear at least as frequently, if not more frequently, than dry Mutsu or the Inland Sea littoral. The typical drought report lists diverse provinces, such as Musashi, Echizen, and the Inland Sea littoral (715), or Shimōsa, Echizen, Ōmi, Ise, Iga, and Owari (835). The heterogeneity of these regions defies simple categorization and leads scholars to believe that something other than archipelago-specific patterns of precipitation were at work in ancient times.17 Another method is to count terms referring to climatic events in Japanese written sources (table 11.2). The eighth century comes across unambiguously as a time of drought, with the highest occurrence of words for dryness (43 percent of all terms) among the centuries for which data were tabulated. It certainly matches the trend observed in table 11.1, and leads historians to infer that the 700s may have been the driest century on record for the islands. The eleventh century also seems to have been desiccated (33 percent), although not so much as the eighth. The thirteenth century, listed in table 11.3 for comparison’s sake, is normally considered one of the wettest centuries on record in Japan, and yet the percentage of terms for drought (26) is fairly high in comparison to the eleventh or even the eighth centuries. The ninth century, with only 7 percent of weather terms referring to drought, leaps out as exceptional. This result flatly contradicts the impression gained from table 11.2, where drought appears to have been a major problem. We might attribute this information in some unknown way to the fragmentary nature of the sources, and it is true that most of the years omitted occurred early in the century, when drought was at its height, yet this explanation seems unsatisfactory. Another possible factor is the more verbose nature of the record, especially the Nihon sandai jitsuroku, covering the years from 858 to 887, when cold, wet weather was more prominent. Moreover, several references to floods concern the Kamo River in Kyoto, a constant nuisance that would have caught the notice of the court. Still, none of these answers seems to account for such a low drought percentage for a presumably dry century. A further problem is that no reliable information is available for the tenth century.
280 | william wayne farris
Table 11.2. Climatological trends for selected centuries Cold Heat Dryness Rain The eighth century: 697–791 59 10 24.60% 4.20%
67 28%
Number of references
Drought index
103 239 43.10%
43
The late eighth and ninth centuries: 796–797; 799; 804–806; 808–815; 833–887* 76 121 43 635 395 62.20% 12% 19.10% 6.80%
7
The eleventh century: 1000–1099 35 52 24 31.30% 14.50% 21.10%
55 33.10%
166
33
The thirteenth century: 1200–1299 85 121 377 47.70% 10.70% 15.30%
207 26.20%
790
26
Note: *Data are fragmentary for 796, 797, 806, 808, 809, 810, 813, and 814. Sources: For the eighth and ninth centuries, I used the Rikkoku shi sakuin for the Shoku Nihongi, Nihon kōki, Shoku Nihon kōki, Nihon Montoku tennō jitsuroku, and Nihon sandai jitsuroku. For the eleventh and thirteenth centuries, I have relied exclusively upon Sasaki Junnosuke, ed., Nihon chūsei kōki kinsei shoki ni okeru kikin to sensō no kenkyū. In weighting these various entries, I counted snow, frost, and hail as one point each for moisture and cold; drought counted as one for both hot and dry; prayers for rain were entered as one point for dryness, and prayers to stop rain as one point against. Flood, great rain, violent rain, long rain, and even wind and rain were all considered as one point for dampness. To figure the percentages, the number of references to a given climatic trend was divided by the total number of events reported. Thus the drought index for the eighth century would be 103/239=43.1.
One statistic gleaned from the Six National Histories, however, suggests that the low drought percentage computed in table 11.2 may be grossly misleading. If we limit the search to the frequency of prayers either to induce or halt rain during the ninth century, the former is far in the majority, seventy-eight to thirty-two. Based simply upon this type of data, the drought percentage would be a staggering 71. Yet this is merely one way to redefine the problem, and may be considered somewhat subjective since it is limited to only one kind of evidence. Yet another way to reconcile the prevalence of drought years in table 11.3 with the low drought percentage recorded for the ninth century in table 11.2 is that rainfall was indeed plentiful but did not remain in the soil for long. If, as Conrad Totman has suggested, there was an “ancient predation” in which forest cover was
Table 11.3. Famine years, extent and causes, 670–1100 Year
W/L
676 677 682 697 701 702 703 704 705 706 707 708 710 711 713 715 719 720 723 726 733 737 746 748 749 750 760 762 763 764 765 766 767 768 769 770 771 772 773 774 775
L W L W W W W W W W W L L L L W W W L L W W W W W L L W W W W W W L L W L W W W W
Causes Drought Drought Cold, rainy summer Unknown Drought Unknown Drought Drought Drought Drought Cold, rainy summer Cold, rainy summer Cold, rainy summer Unknown Unknown Unknown Drought Drought Unknown Unknown Drought Drought Drought Unknown Drought Unknown Unknown Drought Drought Drought Drought Unknown Drought Unknown Unknown Cold, rainy summer Cold, rainy summer Cold, rainy summer Cold, rainy summer Unknown Cold, rainy summer
Table 11.3. (Continued) Year
W/L
777 779 780 781 782 785 789 790 791 796 797 799 804 805 808 809 817 818 819 820 822 823 824 826 830 831 832 834 835 836 837 838 839 840 841 843 844 845 846 847 848 849
L L L L L W W W W L W W L W W W W W W W L L W L L L W W W W W W W W L W L L L L W L
Causes Cold, rainy summer Unknown Unknown Unknown Unknown Unknown Drought Drought Unknown Cold, rainy summer Unknown Drought Drought Unknown Unknown Drought Drought Drought Drought Drought Drought Drought Drought Unknown Drought Unknown Drought Drought Drought Drought Cold, rainy summer Unknown Drought Drought Unknown Unknown Unknown Unknown Unknown Cold, rainy summer Unknown Cold, rainy summer
Table 11.3. (Continued) Year
W/L
852 854 855 859 861 862 863 866 867 868 870 873 876 878 879 880 881 883 884 885 886 887 889 916 917 931 942 957 989 996 997 1008 1017 1018 1021 1022 1025 1029 1030 1059 1082 1088
W W L L L L W W W L W W W W L L L L L L L L L L W L L W L W L L W L W L W W L L W W
Causes Unknown Unknown Drought Cold, rainy summer Unknown Cold, rainy summer Cold, rainy summer Cold, rainy summer Cold, rainy summer Unknown Cold, rainy summer Cold, rainy summer Cold, rainy summer Drought Unknown Drought plus rebellion Unknown Cold, rainy summer Unknown Cold, rainy summer Cold, rainy summer Cold, rainy summer Cold, rainy summer Unknown Drought Unknown Unknown Drought Unknown Unknown Unknown Cold, rainy summer Cold, rainy summer Cold, rainy summer Drought Cold, rainy summer Drought Cold, rainy summer Drought Unknown Drought Drought
284 | william wayne farris
Table 11.3. (Continued) Note: W=widespread; L=local. Sources: Most of these data were collected from Ogashima Minoru, ed., Nihon saii shi, 1–20; and Nishimura Makoto and Yoshikawa Ichirō, eds., Nihon kyōkō shi kō, 3–93. For the tenth and eleventh centuries, I updated Saitō’s work using Fujiwara no Sanesuke, Shōyūki, vols. 2, 4, and 7; and Sasaki Junnosuke, ed., Nihon chūsei kōki kinsei shoki ni okeru kikin to sensō no kenkyū, 9–14. I wish to express my gratitude to Saitō Osamu for providing his judgment about the extent and cause of each famine in a personal communication, but the responsibility for the information in this table is mine. Note that in comparing table 11.1 and table 11.3, I have omitted any reference to famine before the appearance of reliable records in the late seventh century.
stripped from the Kinai and environs, then more rain may have simply eroded away the remnants of topsoil, leading to chronic crop failure and famine. This solution seems most appropriate for regions where trees had been harvested to build capitals such as Naniwa and Kyoto, and fits recent archaeological evidence of environmental degradation in and around the Kinai.18 Be that as it may, it is important to return to the big picture and integrate what Japanese scholars have written about climate in ancient times. Instead of looking at precipitation, their efforts have focused almost exclusively upon the question of temperature. In the 1960s and 1970s, Yamamoto Takeo, Japan’s best-known climatologist, conducted a study using different sources to measure change over the centuries.19 By examining the dates when cherry blossoms appeared as listed in aristocratic diaries, Yamamoto came to the conclusion that on average, flowers bloomed in February or early March in the ninth and tenth centuries, which Yamamoto argued was equivalent to a 2.6 degree centigrade warming trend, as compared to the 1970s. Recently, archaeologists have entered the fray. In particular, Sakaguchi Yutaka has traced the extent of creeping pine (haimatsu; Pinus pumila) in mountains along the border of Gunma and Fukushima prefectures by examining pollen bogs.20 Since the creeping pine thrives in cold weather, finding little of its pollen could mean that the weather was warmer, causing the plant to retreat farther up the mountains. In his study, Sakaguchi gleaned statistics for 7,600 years; not surprisingly, he found unusually high temperatures for almost all the eighth and ninth centuries, and portions of the tenth and eleventh. Taken together, the written and archaeological data lead to the strong impression that the climate of the era from 670 to 1100 in Japan was considerably hotter than in the preceding or following eras. This conclusion — that the late seventh through eleventh centuries were unusually warm — has become the conventional wisdom among Japanese historians.21 In this view, however, the combination is not hot and dry, as the evidence from table 11.2 and table 11.3 implies, but warm and wet. The mid- to late Heian period becomes a climatological “optimum,” when warm temperatures combined with
Famine, Climate, and Farming in Japan | 285
plentiful rainfall to spur an “age of widespread land clearance,” a term borrowed from medieval European history. This interpretation would also enhance the plausibility of the technological and agrarian improvements posited by Morris. It seems unwise to disregard the evidence presented in table 11.2 and table 11.3, however, and so I searched for ways to resolve the wet or dry problem. I stated above that the regions most afflicted with drought did not fit modern archipelago-specific patterns of precipitation; we must face the possibility that we are dealing with Northeast Asian or global (El Niño) trends. Historical and archaeological sources from China may suggest an answer. According to China’s leading expert on climatology, Liu Zhaomin, written records indicate that China experienced hot, drought-ridden summers from 600 to 985. To match Liu’s work to the relevant period for Japan, from 670 to 1100, the term drought is mentioned during the summer for 218 years, more than half the time; records of excessive or frequent moisture appeared only about 60 percent as often. Overall, only eightythree years had snow or were noted as unusually cold, while thirty-one (most in the eleventh century) had no snow or ice at all. Snowy or cold summers occurred only ten times.22 Archaeologists support Liu’s contentions. According to Zhang Jiacheng and Thomas Crowley, “peak warm events” took place in China in the eleventh century, with temperature fluctuations at 1.0 – 1.5 degrees centigrade above preceding centuries. Zhang and Crowley’s most important contribution, however, was to measure diverse data from China and around the world. Scientific investigations show that between 800 and 1100, dust was at a minimum in the Chinese atmosphere, allowing more sunlight and greater heat to reach the surface. In Iceland over the same period, ice floes all but disappeared, while in the Western Sierra mountain range, tree-ring growth was at its narrowest. These bits of data hint that Northeast Asia, and probably the world, was passing through an era of unusually hot, dry weather.23 This hypothesis is not new to European medievalists. Palaeoclimatologists in Western Europe almost universally accept — through examination of pollen, glaciers, and other archaeological and written materials — that the continent entered a period of unusually hot and dry weather beginning about 750 or 800 C.E. and lasting until 1150 or 1200. Experts estimate that on average the temperature was more than one degree centigrade hotter than during immediately preceding centuries. In northern Europe, the late eighth through the late twelfth centuries were characterized by mild winters and dry summers. For these Europeans, the weather was a godsend, since it was essential in allowing peasants to exploit the heavy clay soils found widely in France, Germany, the British Isles, and elsewhere in northern Europe.24 At least one other historical fact may be related to this hypothesis of a hot and dry climate lasting from about 700 to 1100.25 In Central America, the famed Maya civilization, with its beautiful stone pyramids and the most accurate calendar in the world, was coming to an end about 900. Historians and archaeologists have long wondered why a civilization with such an advanced agricultural base would suddenly disappear. Great cities were all but abandoned. Several theories
286 | william wayne farris
have been advanced, including intensifying warfare, overpopulation, and unsuccessful attempts to boost agrarian productivity that ultimately exhausted the soil. Recently, however, some scholars have concluded that a major drought in the ninth century may have contributed to the collapse of classical Maya civilization. Such a climatic trend dovetails nicely with what scholars know from elsewhere in the world.26 Viewed from a global perspective, climatic patterns in ancient Japan seem more likely to have been hot and dry than warm and wet. The more salubrious weather of northern Europe, the summer droughts of China and Japan, and the collapse of agriculture in Central America could well have been part of a worldwide trend, the causes of which are yet to be determined.
Farming Technology and Ancient Famines: A Chronology of Irrigation Over 85 percent of the Japanese archipelago is mountainous, and there are few large, flat plains, notably the massive Kanto (5,000 square miles) and much smaller Kinai and Nōbi regions. Today people normally think of wet-rice farming as taking place most readily in places like the rich Shōnai or Echigo coastal plains, where soil fertility is the highest, and therefore it is easy to assume that conditions have always favored such sites. As this notion has come under increasing scrutiny, many scholars have concluded that the common notion that farmers first developed Japan’s plains and then worked their way up the valleys as they ran out of fertile soil actually reverses the historical process.27 Rather, as Saitō Osamu, Kinda Akihiro, and numerous other scholars have observed, peasants first concentrated their efforts at wet-rice cultivation in the small flat stretches of Japan’s innumerable mountain valleys. To be sure, these parcels were tiny and isolated and did not possess anywhere near the fertility of large plains or river deltas, but other points were strongly in favor of starting there.28 First, in mountain valleys, unlike in plains or deltas, peasant dwellings could be situated well above the water table, making for drier, healthier living conditions in the pit dwellings preferred by almost all ancient cultivators. A tourist has only to visit Toro in Shizuoka, where paddies and homes were constructed below the water table, to become aware of the unhealthiness of dank living conditions. Second, residents could more easily supplement a diet of brown rice with game animals and wild berries, roots, and nuts from nearby hills. They might also do some slash-and-burn cropping. In ancient times, when wet-rice agriculture likely did not dominate the countryside, considerations such as these weighed heavily in selection of peasant settlements. Third and perhaps most important, mountain paddies were much easier to irrigate, with labor and technological inputs being minimal. In some cases, these valley paddies relied exclusively on natural irrigation, that is, the simple runoff cascading down any well-forested mountain after a spring or summer rain. Others required some leveling, ditching, and perhaps even simple pond building, but the spring and summer runoff was important in watering these few small paddies, too. Elaborate irrigation works, which required
Famine, Climate, and Farming in Japan | 287
huge labor inputs unavailable in the Nara and Heian periods, were unnecessary for such sites.29 Japan developed, as Saitō has argued, from rice cultivation in “mountain valleys to alluvial plains.”30 Saitō is an expert on the early modern period, but studies of farming in the Nara and Heian periods also support this characterization. To cite one seminal article, in 1978 geographer Kinda Akihiro analyzed wet-rice agriculture in the Yamato basin during the Heian period.31 He found two types of irrigation ponds — valley ponds (tani ike) and saucer ponds (sara ike) — among the fifteen thousand that then watered the region. Relatively small valley ponds were constructed by simply damming the upper end of a valley to catch the runoff, and remained full the whole year. For example, one valley pond near Akishino was bounded on three sides by hilly ridges and on the fourth by a small road.32 The saucer ponds, which were much larger, were placed on slightly elevated, flat portions of the basin, where they filled with rainfall during the monsoon season, when water was most abundant. Because their precious liquid could be dispensed during the summer growing season, they were essential for irrigating level areas in the basin. Referring to old maps and documents, Kinda found that only one saucer pond predated the Kamakura era (1185 – 1333). Following this reasoning, he showed that the flat center of the basin lacked watering facilities for the ancient period; most paddies around Nara, situated near the capital and certainly one of Japan’s most agriculturally advanced regions, were in and near mountain valleys, where peasants could take advantage of natural water flow or easily construct a valley pond. This chronology of agrarian development — from mountain valleys to alluvial plains — has two important implications for famines of the early age. In his study of food crises over the premodern era, Saitō has shown that the percentage of crop failures caused by drought diminished over time. Using Saitō’s figures, we note that while about 40 percent of all famines were caused by drought in the seventh through eleventh centuries, the percentage had dropped noticeably by the late medieval (29) and early Edo periods (24). What is more, in the eighteenth and nineteenth centuries, no harvest failures occurred as a result of drought; long, cold summer rains were the sole cause.33 To be sure, the Nara and Heian eras seem to have been especially hot, dry times, but when we think about the causes of famine in Heian, it is important to realize that one reason famines happened so frequently was that farmers of the day had not yet taken advantage of the elaborate irrigation facilities (such as saucer ponds or waterwheels) that could insulate their paddy fields from inclement weather. In this sense, the story of harvest failure in the early period is also an important key to understanding how agriculture developed over the entire premodern period. Second, because most paddies relied upon natural watering, adequate forest cover was essential to the task. If timber was stripped from most mountains in the Kinai, rains would not come as often. The moisture that did fall would promptly run off, and paddies could be buried in debris or die of thirst because the mountains held no moisture in reserve.34 Given the current evidence of ancient ecological degradation in the Kinai, we should recognize that even in the
288 | william wayne farris
Nara and Heian periods, famine was a product of climatic and human variables.35 In other words, the very nature of ancient wet-rice agriculture contributed to the high incidence of food shortages.
Farming Technology and Ancient Famines: Planting Dry Fields and Spring Hungers Eleven times between 670 and 840 the central government issued commands for farmers to plant dry crops, such as wheat, millet, barley, soybeans, and buckwheat. A typical regulation (kyaku) reads as follows: The Council of State orders: That barley and wheat ought to be planted. Concerning the above, in a regulation dated the fifteenth day of the ninth month of the second year of Tenpyō Jingo [766], the Council of State said: “...These grains are the best for sustaining a person when all food is gone and for succoring him during poverty. Have the provinces of the realm (tenka) encourage commoners to plant barley and wheat. Designate one diligent official of the district or province and have him specialize in this matter. Give the name of the specialist to the messengers who assemble at court (chōshū shi) and have them report.”...It is now heard that the foolish people ignore this advice and give it no heed. When they come to the end of their grain, they vainly suffer from famine; or, even if they do plant [these dry crops], they have already missed the proper season. They emptily expend their labor but have no result. This derives from the district and provincial officials’ disrespect for regulations....Who could call them good officials? In the Monthly Injunctions [Yueling, found in the Liji, or Record of Rites], it says, “[I]n the middle month of autumn, encourage the planting of wheat and barley. Do not let people miss the planting season. If there is such a case, there is no doubt that it is a crime.” Hereafter, beginning in the eighth month you [local officials] ought to make an effort to have [these grains] planted and not miss the season.... Kōnin 11 [820]/7/936 This order raises several questions. Did this law have real meaning, or was it a routine reiteration of nostrums from China? Did the court really know more about cultivation than the peasants who were doing the work? In an agricultural system in which land was frequently left untended because of infertility, how likely would a second crop have been to mature? Were these unirrigated grains to have been sown in special fields or in the drought-stricken, parched paddies? The historical context helps to answer some of these queries.37 As may be observed from table 11.3, the year 820 was one of widespread famine induced by drought, coming at the end of a long spate of such farming disasters. Moreover, when the eleven edicts preserved for 693, 715, 719, 722, 723, 766, 819, 820, 839, and 840 are plotted on a graph charting the occurrence of terms for drought, it seems clear that the court was apt to issue the laws when it was especially hot and dry and the wet-rice crop may have suffered (figure 11.1). In seven of the eleven cases,
Famine, Climate, and Farming in Japan | 289
famines — usually widespread — were recorded. The court’s policy of last resort during the eighth and ninth centuries — whether realistic or not — seems to have been to encourage an ad hoc double-cropping of dry crops, in hopes of keeping its tax producers alive. There is at least one point that figure 11.1 does not explain. While it is true that the orders almost always appeared during severe drought, not every such peak is so represented. In particular, the early and latter two-thirds of the eighth century and the last sixty years of the ninth century contained many dry spells when no such proclamations were drafted. Of course, we can never be certain that mere repetition of terms in the Six National Histories comprises an accurate representation of the weather, but still scholars are left wondering why the court did not promulgate these mandates more often. Perhaps those years represented on the graph were especially harsh ones. Or it might have been assumed that the first commands were still in effect. Given the fragmentary nature of the written record for the ninth century, some sources may even have been lost. Ultimately, we return to the basic question of how well the government elites understood the contemporary agrarian system, a query that cannot be answered short of knowing what was in the minds of lawgivers. In any case, the court’s insistence upon the role of dry grains as emergency food supplies when famine was common seems apparent. The various commands to raise wheat, barley, or other unirrigated grains during famines contain another critical insight about food shortages and the ancient farming system. Of the eleven orders, ten were promulgated during the late summer or fall: to be specific, the dates of implementation were 693/3, 715/11, 719/11, 722/6 and 9, 723/10, 766/10, 819/7, 820/8, 839/9, and 840/6. And because the 693 edict did not explicitly mention the kinds of dry crops discussed so far, I am tempted to set it aside as an exception. If that is so, then the court, observing recent failures of the harvest and fearing the worst for its tax producers, was encouraging the planting of wheat, soybeans, and so on in the fall to be harvested the next spring, when food supplies would have been at their lowest and the danger of mass starvation the greatest. Furthermore, Tamura Noriyoshi has shown through the use of death registers (kako chō) extant for the medieval period that cultivators were most likely to have died in the spring or early summer (the first through the sixth months). This trend was true for both men and women, and insofar as can be determined, also for children. Tamura posits a “seasonality of death,” with medieval peasants maintaining a hand-to-mouth existence that kept them in a state of chronic malnutrition.38 The persistence of famine during the eighth and ninth centuries, along with the repeated commands to sow unirrigated crops in the autumn, seems consistent with a similar condition of “spring hungers” for peasants of the ancient period. Fortunately, there is direct evidence from the mid-eighth century to corroborate such an interpretation (figure 11.2).39 As may be noted from table 11.3, the year 762 was the first in a string of major famines wreaking havoc throughout the 760s. As figure 11.2 shows, the price of brown rice (represented by the solid line) began
Figure 11.1. Drought and government orders to plant dry fields. Note: The horizontal axis represents time. The vertical axis lists the number of references to drought in a given year. A circle indicates the issuance of an edict to sow unirrigated fields. Sources: SZKT, Nihon shoki, Shoku Nihongi. Nihon kōki, Shoku Nihon kōki, Nihon Montoku tennō jitsuroku, and Nihon sandai jitsuroku. See Farris, “Demographic Change and Agricultural Development in Early Japan, 645 – 900.”
Figure 11.2. Fluctuations in the price of brown rice in the famine year of 762. Note: The horizontal axis represents months in the years 761 – 763. The vertical axis indicates the price of brown rice in copper coins. Prices are for one shō (present-day four shō). When more than one price was available for a given time, an average was used. Between B and C, there were no written sources extant. For reference, an unskilled worker’s wages averaged ten coins per day during this time. Source: This figure is part of a more complex graph composed by Aoki Kazuo, Nihon bunka shi taikei. Nara jidai, vol. 3 135.
Famine, Climate, and Farming in Japan | 291
to spike early in the second month, and by the fifth month had reached its high for the year. After a period for which there are no records, and when I suspect that the grain was simply unavailable, the price dips in the early fall of 762, only to commence another upswing late in 762, which seems to have peaked early in 763, another year of widespread drought, crop failure, and famine. To be sure, the price of rice should not be automatically equated with a mortality curve. If the information contained in figure 11.2 is to be credited with some validity, however, then it seems likely that Tamura’s notion of a chronically undernourished peasantry suffering most during the spring hungers is just as applicable to the Nara and Heian periods as it is to the medieval age. And such an inference seems only reasonable, as we would not expect the ancient epoch — when annual double-cropping, champa rice, the waterwheel, animal fertilizer, et cetera were rare or unknown — to have been more prosperous than the late medieval age. To sum up, evidence presented so far shows that hunger was common for many people during the eighth and ninth centuries and, although records are paltry, most probably for the tenth and eleventh as well. Hot, dry summers were at least partially responsible for most of the major famines, a climatic trend verified independently through written and archaeological materials for Japan and seemingly linked to conditions in China, Western Europe, and the Americas. Ancient farming, with its emphasis upon mountain valleys naturally watered by rainfall, left fields frequently subject to drought and erosion. Orders to plant dry crops and evidence on the price of brown rice, the preferred grain in commoners’ diets, suggest that Tamura’s concept of spring hungers applied equally well to the ancient and medieval periods. Given these conditions, what were the responses of the political and other elites, who usually had plenty to eat?
A Chronology of Hunger: Effects and Countermeasures In this section I discuss the evidence for noteworthy famines, a task that may seem daunting but given the laconic nature of most entries can be handled in reasonable space.40 Viewed overall, court responses to crop failures reveal a progressive diminution of resources and lack of will. Through the period 670 to the early ninth century, the central government was aggressive in doling out relief to victims, granting tax remissions, and even donating loan rice, which was used for taxes at the local level. Religious ceremonies were also thought to be effective. By the early years of the 800s, however, the crisis grew so unmanageable that the court blamed its local officials, and then provincial functionaries accused the victims themselves, badgering them for drinking too much or not planting barley in the right season. By 900, the most the court could manage was to calm Kyoto urbanites by furnishing them with cash or grain. During the tenth and eleventh centuries, even noting famine outside the capital seems to have been too demanding a task for aristocratic diarists or court annalists. Throughout the period under consideration, the great mass of commoners, whose hardships eventually resulted in a crisis that led to impoverishment of the government and
292 | william wayne farris
retreat of the rulers, provided an important force for the changes in Japanese institutional and political structures. Information for the seventh century is particularly sparse, but seems to have set several trends for how the court would deal with starvation thereafter. In 676, only four years after Emperor Tenmu’s accession to the throne, local famine struck Shimotsuke and Yamato, and the court organized prayers to Shinto and Buddhist deities. This was to remain a tried-and-true response well into the medieval era. One other brief note strikes a chord with those who know about famine around the world: “The local peasants suffered a bad year and starved. They wanted to sell their children, but the court considered this immoral (futoku).”41 Mention of such behavior resonates with those who know about the Kangi disaster of 1229 – 1232, the African crises, or any of a thousand others of the twentieth century: starving families sell their relatives in hopes of receiving grain for themselves and saving the person sold into bondage by having him or her gain sustenance in return for services rendered to the buyer-patron. The court, however, was willing to have none of such “immoral” behavior, indicating that there were limits beyond which it would not go. In 697 famine ravaged Harima, Bizen, Bitchū, Suō, Awaji, Awa, Sanuki, and Iyo, respectively; Emperor Monmu (r. 697 – 707) resorted to a more acceptable, familiar countermeasure by not collecting the principal or interest due on rice loans.42 Turning to the eighth century, we observe that in addition to prayers (especially for rain) and tax relief, the court began in 704 to distribute (shinjutsu) grain to suffering peasants.43 The next year Monmu seemed especially overwrought and granted an amnesty as well.44 When in 706 an epidemic was also recorded, historians face the question of whether there was really an infectious disease at work or rather that the consecutive years of poor harvests had resulted in widespread starvation, one symptom of which is acute diarrhea. It is worth remarking that a correlation coefficient calculated for the incidence of epidemics and famines for the years 697 – 758 comes to .88, somewhat suggestive of a relationship between the two phenomena.45 This connection may be viewed in many ways: the plague may have weakened an already hungry populace, leading to the inability to labor in the fields and spreading the famine; or famine may have reduced the ability to resist infection. In any case, descriptions of famine commonly included references to pestilence. The court was fortunate to have only minor crop failures to respond to while construction was ongoing at Nara. The general famines of 715, 719 – 720, and 722 – 723 prompted the first commands to plant dry crops, but were otherwise not unusual. In 717, amidst these harsh times, the court issued an order that was to become a theme repeated more often in later times: local officials were reporting nonexistent harvest failures to avoid taxation.46 A rift was developing between capital and peripheral officialdom even at this early date. In 733, drought provoked a serious crop failure on the eve of the smallpox epidemic. The court of Emperor Shōmu (r. 724 – 749) responded with prayers, an amnesty, bans on rice wine and killing animals, and grain doles for the elderly,
Famine, Climate, and Farming in Japan | 293
infirm, and widows.47 The famine of 737 was intimately intertwined with the 735 – 737 epidemic. As the islands began to recover from the epidemic in 746, hot, dry weather led to another famine, but the responses were all familiar, except that the elite decided to bestow hemp upon themselves “to encourage industry.”48 Throughout this time, grain doles and prayers for rain were the most common reactions of the political elite. From 762 through 766, widespread famines struck every year, and it is no accident that these years also represented the most turbulent in Nara political history. Fujiwara no Nakamaro was ousted in favor of a relative, Empress Shōtoku, and her ally, Dōkyō. For his part, the sinophilic Nakamaro tried everything to prove his worthiness to lead: he forgave overdue loans, stopped corvée labor, remitted taxes, and of course, doled out grain and prayed. He even raised the rank of a low-level bureaucrat who gave free provisions to more than ten people, and then made a pledge to reward other generous persons similarly, a sure sign that the government was running low on its own resources to combat hunger.49 None of this did him any good, and he was overthrown in 764, as much for his failures in monetary and agricultural policy as for his poor relations with Shōtoku. The strong-willed woman, who died in 769, fared no better than her brother. Her response to events, however, was somewhat different. While adopting Nakamaro’s device of awarding higher court rank to “good” local officials who contributed grain, she also specifically mandated that the periphery (Kyushu) send its surplus rice to the capital to lower prices, which had skyrocketed there to one hundred coins per measure (to).50 Her emphasis upon the welfare of those dwelling in Nara placed the core above the hinterlands; it is the first notice in Japanese history that the city had special problems and therefore priority during lean times. The difficulties of urbanites and preoccupation with their welfare would continue as themes into the Tokugawa period. In 766, she also issued another command to plant dry crops. Hunger stalked the land throughout the reform era 770 – 780, too, and was usually caused by cold, damp weather. Court countermeasures included the normal policies of prayer, tax relief, and grain doles, but in 773 the central government attempted to relieve the burden from uncollected rice loans by granting rank to wealthy individuals who gave from their personal reserves.51 Also in the same year, the Council of State warned local officials against ignoring or taking advantage of the plight of commoners. There is little in the way of quantitative materials for any of these crises, but in Inaba in 779, three thousand were reported to be starving; in 780 a rich woman from Iyo contributed grain to 158 poor people.52 The next general famine in 785 resulted in complaints of excessive expenditures by local officials, and the conflict between the capital and provinces was widening. When another crisis struck in 789, the remedies were mostly the same, including opening the government storehouses of rice sheaves to lower prices and provide grain for victims. This famine marked the designation of a new category of loan rice, called succoring emergencies (kyūkyū), and it remained in force into the tenth century.53
294 | william wayne farris
Three widespread famines over the years 790 and 791 undoubtedly helped ruin Kanmu’s prospects for a successful move to Nagaoka. In 790, there were said to have been 88,000 people starving in Kyushu alone; if one accepts Sawada Goichi’s estimate for Kyushu’s population (664,400) for about 800, then over 13 percent of the population was in dire straits.54 Although we may entertain doubts about the accuracy of these figures, it is worth noting that this crisis witnessed the beginning of a series of policies to revive agriculture there.55 Outside of canceling court ceremonies, however, no other policies were instituted. Three years after Heavenly Sovereign Kanmu moved to Kyoto in 794, a widespread crop failure struck Iki, Kai, Shimōsa, Musashi, and Tosa, and in 799 a major famine was so severe that the government legalized private rice loans to stem the tide of suffering. In Sanuki alone, for example, 2,000 koku were loaned at no interest to households on the verge of dispersing. In 804, only Awaji, Yamato, and Settsu were victimized, but it required 93,900 sheaves of rice to aid the impoverished of tiny Awaji. In 806, an edict decried the growing disparity between rich and poor, undoubtedly partly a result of the legalization of private rice loans. The same law tried to force a borrower’s family or neighbors to repay overdue loans. And so the first decade at Kyoto featured similar crises and few new ideas for how to deal with them.56 Despite the paucity of sources from the ninth century, hunger appears as a constant concern. In 808, Fujiwara no Sonondo, a court observer dispatched to the Inland Sea littoral, reported that five provinces there had not been able to pay taxes for twenty-four years; the governors were either dead or had left office, while the people faced disease or starvation. The next year was no better, as the court slashed salaries and prayed for rain. In 812, probably as a result of earlier shortfalls, relief was handed out to residents of Kyoto, and local storehouses were emptied to lower the price there. Unlike their situation in Shōtoku’s era, however, most provinces were too financially straitened to aid the urbanites. More grain relief was meted out in 814.57 By 819, the court admitted that it had no more reserves to meet the shortfall arising from decades of bad harvests, and it is no accident that this year saw a repetition of the command for peasants to plant dry crops. The conflict between local and capital elites also escalated, as the court once again made accusations of prevalent fraud in the reporting of damaged fields. The next year was even worse, and all the government could do was reiterate its call for cultivators to plant barley and wheat (translated above). The 820s, 830s, and 840s were one long drought, with one crop failure following another, and famine a permanent presence. By this time, the court had more or less run out of policy alternatives. We see in 823 special privileges allotted to capital residents again: they were given cash, and attempts were made to drive down the cost of grain there. In the major crisis of 824, the court merely had sūtras intoned and prayed for rain. Starvation in the years 832, 834, 835, 836, 837, and 838 afflicted the entire realm: the government was so desperate that it listed the names of “good officials” who used portions of their salaries to aid the hungry and set limits on the amount of sheaf rice doled out to men and women, with
Famine, Climate, and Farming in Japan | 295
three going to adults, two to youths, and one to children. Government granaries were empty, and even residents of the capital received only intonations of sūtras to lower the price of food.58 During 839 – 840, when general famine struck again, the court could only implore its local officials to order the peasants to plant dry crops. Prayers and other appeals for rain fill the pages of court annals. As an afterthought, the government tried to entice cultivators into building more irrigation works, but no further mention is made of that novel idea. In 841, some 26,618 peasants in Dewa were given tax remission “on account of crop failure,” while a Sagami official loaned the poor grain and hemp cloth to help them pay their revenues. In 847, a commoner won court rank for paying the taxes of the poor in Wakasa.59 The year 848 witnessed another general crisis because of too much rain; in 849, starvation victims in the capital were granted grain, and the consort of Retired Emperor Saga dispensed 500,000 coins to the poor and then toured Kyoto to observe the effects of her largesse.60 Thus the horrific early decades of the ninth century came to an end, but there is one further monument to the enduring legacy of starvation during this time. In 2000, archaeologists digging in erstwhile Kaga Province came across a remarkable find — a wooden placard describing village life during these harsh times in detailed and colorful language: The district [of Kaga] orders: The local officials (gō ekichō narabini sho tone) of Fukami village (mura) to implement these...articles. One: Farmers (denpu) should go down to the fields every morning at the hour of the Tiger [about four a.m.] and return at night at the hour of the Dog [about eight p.m.]. Two: Farmers should be banned from imbibing rice wine and eating fish as they please. Three: Peasants should be prevented from refusing to work on ditches and dikes. Four: Before the thirtieth of the fifth month it ought to be reported that fields have been completely planted. Five: Caves and cliffs should be searched throughout local villages for suspicious characters. Six: Peasants should be banned from not feeding silkworms in Kuwahara (or the mulberry fields). Seven: Villages ought to be proscribed from allowing drunkenness and gambling (giitsu). Eight: Agriculture ought to be seriously worked at. The above-mentioned village heads should report the names of peasants. In looking over relevant materials, it says in the provincial order of last...month/28: “Encouraging agriculture is a part of the laws. But the peasants run away as they please and do not cultivate. The main things they do are drink rice wine, eat fish, and injure others in disturbances. Planting is always late; rather claiming that [they are unable], they are simply not exhausted at
296 | william wayne farris
all. Also, this brings about famine. If these district officials do not administer at this [garbled text] period, then how could it be any other way? The district ought to be aware of these matters and pass them along by word of mouth. They should investigate and make decisions.” In accordance with the sense of this order, give commands to landholders (denryō) that each in every village report and circulate this. If there are lazy persons, then capture them and send them to the district office....Set this order up on placards along the road and strictly carry out the bans. If landholders (denryō; tone) despise this and hide out, then punish those persons. Do not be lenient. When the order arrives, carry it out. [Signed by seven officials in the district office] Kashō [illegible]/[illegible]/7 (848 – 851) Received: [illegible]/15 denryō Hasetsukabe Yoshimaro61 Historians could hardly hope for a more specific description of how local officials at the district level tried to encourage and cajole farmers into tilling their fields. The order envisions sixteen-hour workdays, with no treats like fish or wine. Irrigation facilities were to be maintained, and deadlines for planting were set; sericulture was to be diligently pursued. Villagers were to watch out for “suspicious characters” and by all means refrain from drunkenness and gambling. The main text presents a gloomy picture of peasant discipline, with farmers’ absconding as they pleased and stirring up trouble. For these reasons, planting was always late, and the numerous famines described above were the result. Lastly, the order even establishes a system for reporting malcontents and backsliders. The gist of the command is forthright: farmers were not working hard enough; they were drunken, avaricious, and violent, and the result was famine. The officials of Kaga district were following the government’s ninth-century policy of leading the farming populace out of this sad state of affairs and trying to get rural Japan into shape. We should also note that at this time, as both before and after the ninth century, district functionaries sought to single out the leading landholders and blame or make them responsible.62 Only two points about this interesting find are unclear. First, the opening summary of the order states that there should be ten articles, and yet there are only eight. Second, at a few points in the text the writing is garbled and impossible to read. Under these dire conditions, the weather suddenly switched from too dry to too wet, and more famines resulted. In 854, general famine is suggested by entries that state that one-third of the taxable adults of Owari received remissions, that it required 35,000 sheaves to succor the starving in Iwami, and that there were nine thousand victims in Hizen and Bungo (about 5 percent of the population, according to Sawada). Soldiers fled in Mutsu. A local shortfall hit Dewa in 855, and nineteen thousand individuals living there, or 32.7 percent of the total populace, were relieved from taxes.63 The 860s and 870s then witnessed the brunt of the cold, damp weather. The years 863, 866, and 867 all were years of widespread famine, with conditions particularly harsh in Kyoto. The court reported that the price of polished rice
Famine, Climate, and Farming in Japan | 297
had jumped 35 percent and that of brown rice 40 percent. Officials of the Eastern and Western markets purchased and distributed what grain they could afford to buy. In Owari, Awa, and Bizen the remainders of loan rice were divided among the starving. Kami and buddhas were implored time and again to bring relief, but peasants continued to flee the land, and the court blamed district magistrates. In 867, robbery and arson erupted amidst the food crisis, and doles of money, salt, and rice seemed to do no good. By the end of the year, the government had resorted to the Chinese convention of the “ever-normal granary” (jōhei sho), and when the price of grain was advertised at eight coins (it had been at least thirty), “people assembled like clouds.”64 The year 869 saw unusual amounts of rainfall, with the result that a major famine stalked the land in 870. The court tried the usual remedies of grain relief and prayer. When the weather was unnaturally wet and then dry out of season in 871 – 872, crop damage led to another general famine in 873; 7,113 people in the direst of straits received tax relief in Bingo. People and livestock drowned in the harsh rains of 874, and it is little wonder that 876 was a time of serious famine as well. In 878, drought reappeared to claim many victims; there were strange stories of commoners cutting down trees on Jingū’s tomb, resulting in a curse (tatari). The court seemed preoccupied with religious responses. Later that year, some grain relief was sent from Harima to Izumi, and the ever-normal granary was noted as operating in the capital. In Hōki, no rice loans had been possible from 860 to 877, such was the crisis in farming.65 Local famines then continued until the end of the ninth century. In 880 in Dewa, difficulties with the resident “barbarians” (emishi) provided a possible artificial cause for the shortage.66 In 883, Musashi, Izumo, and Nagato all tried handing out loan rice to tide over starving residents. Other than these two years, however, in the 880s, the government’s attention came to focus on the capital more than ever — a fitting prelude to the tenth century, when the provinces were valued merely for their ability to produce revenues. Written sources in the tenth century more often than not merely mention starvation, with little in the way of concrete policies to deal with the crises. In 908, provinces “declined,” but the elite did nothing except pray and cancel ceremonies. Only the capital received any grain relief. Harvest failure was noted in 913, but courtiers’ only concern was how much tax product it would cost them. In 917, one of three widespread famines known for the tenth century struck: the court described numerous robber gangs but otherwise seemed most concerned for its own welfare. (One measure was to ensure that Retired Sovereign Yōzei have enough water to make wine!) Untoward weather conditions were common in the 920s, but again the court was restricted to worrying about collecting enough in grain taxes. The capital utilized its granaries when starvation struck in 931, but officials also called out the police to crack down on robbers.67 In 940, there is a tantalizingly brief notice of general hunger, and since this was the year Taira no Masakado was killed, we may wonder if the war had something to do with the shortages. In 942, Kyoto was once again without sufficient grain, and officials responded with alacrity by distributing one hundred strings of cash
298 | william wayne farris
to the lowest-ranking officials for further handouts to capital inhabitants.68 One reason for this munificence must have been the numerous reports of brigands there. The age of Tenryaku (947 – 957) had its share of bad weather; courtiers expressed concern about how much grain could be collected as tax, as had become their wont. In 957, the second far-flung famine of the 900s arose; this time the crisis was serious enough to merit tax remission and grain doles. Ceremonies were cancelled and prayers intoned. Relief was provided only within the capital. One chronicle recorded that “a crazy woman tore off the head of a corpse and ate it in front of the Taiken Gate.”69 For nearly forty years thereafter, there are entries about bad weather, but nothing about famine. The last three periods of widespread hunger — 989, 996, and 997 — were given merely the briefest notice. Although more food shortages were recorded for the eleventh century, the entries differ little from those of the tenth. In 1008, excessive rain led to famine around the capital; otherwise, very little in the way of specific information is available about the crop failures and subsequent hungry years. Causes were occasionally described, and the extent vaguely referred to “all under heaven.” Besides the references to praying and giving relief to capital residents, however, historians have little to work with to describe the famines of the eleventh century.
Conclusions In the introduction, I posed three questions about food shortages in ancient Japan: How frequent and severe were they? What were their causes and how did they relate to farming techniques? And what were the social and political responses to food crises? After a thorough examination of the evidence, I conclude that famine stalked the Japanese islands frequently between 670 and 900. Although the evidence becomes increasingly sketchy and focused on Kyoto after 900, it seems reasonable to posit that hunger continued to be the lot of many common people, perhaps striking about once every three years, as it did in medieval times. There can be little doubt that the ancient period from 670 through 1100 was a perilous time for those who subsisted by agriculture. At the same time, famine played a clearly secondary role to epidemic disease in boosting mortality and restraining population growth. While reliable figures are rare, starvation seems to have added somewhat to mortality, even as it just as surely depressed fecundity in the years that it raged. To an unknown extent it contributed to the overall demographic pattern of the age. Yet starvation never threatened the elite as a plague might. McNeill’s “microparasitism” was the major theme of the Nara and Heian periods, but famine was always there to carry away its share of victims.70 Drought had a hand in well over half the major food crises of the age. Cold, damp summers ran a distant second. Written and archaeological evidence from Japan bolsters this contention; climatological trends in China, Western Europe, and the Americas suggest that worldwide weather was probably hotter and drier than before or after, too. War rarely played any role; yet there was a fairly strong
Famine, Climate, and Farming in Japan | 299
link between pestilence and hunger, although several explanations for the correlation may be admitted. The prevalence of drought in the records may be only partly a function of climate. Elaborate, sophisticated irrigation works were difficult for ancient peasants to build and maintain; the core areas of alluvial plains were barely exploited at all by peasants of the Nara and Heian periods. Farmers were more likely to use mountain runoff or swamp waters than stream flow to dampen their paddies. Because of this tendency, wet-rice cultivators were especially vulnerable to a period of dryness that their medieval or Edo descendants might have withstood. One way for agriculturists to combat the effects of harvest failure was to plant a second crop of wheat, barley, soybeans, or other unirrigated grain early in the autumn. Realizing this, the court ordered its tax producers to follow the prescriptions of Chinese texts and do just that, especially during times of drought. Price records from 762 suggest that the late winter and spring could be a “season of death” for rural folk, as their supplies of last year’s grain were exhausted. Viewed over the period 670 to 1100, court responses to the repeated cycles of famine indicated a growing inability to deal effectively with crisis, at least at the central level. While lawgivers doled out grain, remitted taxes, and even made generous allowances of sheaf rice in the eighth century, increasingly in the ninth century capital elites berated local officials for supposed corruption and laziness. They tried to bribe wealthy individuals into handing out relief by giving them the empty honor of low court rank. By the end of the ninth century, that was about the only resource the courtiers had left. Of course, religious palliatives, such as the intonation of relevant sūtras or prayers for rain to native gods, were popular throughout these centuries. From the late eighth century, sources reveal an increasing concern for the welfare of the residents of Kyoto. Urbanites received grain doles, money, and other special perquisites not applicable to rural folk. There was also some attempt to make the core-periphery relationship even more openly parasitic, as Shōtoku and others had rice shipped from the countryside to the city. By the tenth and eleventh centuries, when famine records are few and brief, it becomes clear that the worldview of the Heian elites had retreated totally within the city, with wellplaced aristocrats relying on their tax farmers to garner whatever surplus was available. The focus on Kyoto and exclusive concern for its welfare was a trend that continued well into the medieval epoch. Food crises provide links to many facets of society. In the case of ancient Japan, they relate crucial information about demography, climatology, agronomy, economics, social structure, government finance, and religion. They help the historian to realize that the grandeur and elegance of life at the Heian court flowered amidst great hardship for the overwhelming majority of the populace; there was a fragile balance between the core and periphery. Yet unlike elites in modern nation-states, ancient political elites — at least those at the highest levels — seem not to have tried to use the crises to expand their sway over other inhabitants of the islands. Rather, the horror of frequent famines, along with epidemics, may
300 | william wayne farris
have been a major factor in convincing the capital aristocracy to leave the details of local administration to the custodial governors and emphasize its own refined worlds of art, literature, and religion.
Notes 1. See Ann Bowman Jannetta, “Famine Mortality in Nineteenth-Century Japan”; Conrad Totman, Early Modern Japan, 236 – 245; and Saitō Osamu, “The Frequency of Famines as Demographic Correctives in the Japanese Past,” which is an English translation of Saitō’s “Kikin to jinkō zōka sokudo,” with the latter article pertaining only to the Tokugawa era. 2. Dana Morris, “Land and Society,” 184 – 194. 3. Kōno Michiaki, Nihon nōkōgu shi no kisoteki kenkyū, 280 – 297. It is important to note that the implement Kōno finds evidence for does not resemble Morris’ “moldboard” plow. 4. Ibid.; Cornelius Kiley, “Provincial Administration and Land Tenure in Early Heian,” 279 – 282, 313 – 314, 329 – 337. 5. On the application of manures, see Furushima Toshio, Nihon nōgyō gijutsu shi, 146 – 147, 205 – 207. 6. Yamamoto Takashi, Shōen sei no tenkai to chiiki shakai, 104 – 118. Toda Yoshimi, Nihon ryōshu sei seiritsu shi no kenkyū, 55 – 58, 180 – 182, and 330 – 331, argues for a rate of 30 to 50 percent for the mid-Heian epoch. It should be added that, because Morris fails to take abandoned fields into account in his estimates of population for 740 (five to six million) and 900 (seven to eight million), they are of doubtful reliability. 7. See Uno Takao, Shōen no kōkogaku, 61 – 67. 8. Ibid., 67. 9. Kinda Akihiro, “Kokuzu jōri puran to shōen jōri puran,” 11 – 33. 10. The best analysis of the Owari document is Iyanaga Teizō, “Owari no kuni gebumi,” 130 – 158. 11. Sakamoto Shōzō, Nihon ōchō kōkka taisei ron, 181 – 188; idem, Nihon no rekishi, 140 – 142; Nishiyachi Seibi, “Kajishi to suiko,” 313 – 328. While Sakamoto and Nishiyachi leave open the possibility of a modest rate reduction, it would not have been enough to explain the sudden disappearance of references to famines. 12. Saitō, “The Frequency of Famines,” 7. 13. See Sasaki Junnosuke, ed., Nihon chūsei kōki kinsei shoki ni okeru kikin to sensō no kenkyū. In research based upon this and other primary sources, I have found that, for example, the twelfth century included twenty lean years, both general and local, and the thirteenth century contained at least nineteen general and regional famine years. 14. Sakamoto Shōzō, Nihon ōchō kōkka taisei ron, 1 – 181; for a simplified explanation, see idem, Nihon no rekishi. 15. Cf. Totman’s description of Tokugawa local government in Early Modern Japan, 49. The dearth of famine materials for the tenth and eleventh centuries may not simply be a product of institutional structure or randomness. If it is true, as I have argued in Population, Disease, and Land and elsewhere, that epidemic diseases killed large percentages of the populace from about 700 to 1100, then it is quite likely that there may have been fewer residents of the archipelago in 950 or 1000 than there had been two centuries earlier. There could have been much less pressure on food and grain production, possibly leading to a reduction in the frequency of hungry years. This reasoning, however, depends upon several imponderables, such as the relative sizes of the cultivator and elite classes, farming productivity, and tax rates, among others. This point is somewhat speculative.
Famine, Climate, and Farming in Japan | 301 16. Glenn Trewartha, Japan: A Geography, 50 – 51. I also used Totman, History of Japan, 16 – 19. For the record, Trewartha shows that on average between April and July, it rains 73.7” in Fukuoka, 76.2” in Hiroshima, 68.6” in Nagoya, 63.6” in Fukui, 59.9” in Tokyo, 59.6” in Osaka, and 54.1” in Akita. 17. A year-by-year examination of materials for the years between 676 and 889 when provincial notices of drought and rain survive in great abundance yields the following conclusions. In comparing these notices to contemporary rainfall patterns, thirty-three of sixty-one years showed no discernible pattern at all. That means that, for example, both wet and dry areas experienced drought or excessive dampness at the same time. In thirteen years, normally rainy regions were dry, as in 836. Seven years, such as 876, revealed wet locales receiving too much water. Five years including 837 showed dry areas getting too much precipitation. In only three years did desiccated areas receive no rain, as in 704. Given the fragmentary nature of the evidence, the safest conclusion seems to be that climatic reports for the relevant period are random with respect to current patterns of precipitation. See Nishimura Makoto and Yoshikawa Ichirō, eds., Nihon kyokō shi kō, 3 – 69. 18. Conrad Totman, The Green Archipelago, 9 – 34; William Wayne Farris, Sacred Texts and Buried Treasures, 190. It should be noted that of the sixty-one years described in note 17, nineteen recorded drought in the some part of the Kinai. This would not, however, explain the frequency of drought in other parts of the islands. 19. Yamamoto Takeo, “Rekishi no nagare ni sou Nihon to sono shūhen no kikō hensen”; idem, Kikō no kataru Nihon rekishi, 11 – 21. 20. Sakaguchi Yutaka, cited in Isogai Fujio, “Nihon chūsei shi kenkyū to kikō hendō ron.” Isogai has expanded and updated this essay in Chūsei no nōgyō to kikō, 242 – 277. 21. Toda Yoshimi, Shoki chūsei shakai shi no kenkyū, 4; Kito Hiroshi, Jinkō kara yomu Nihon no rekishi, 65; Isogai, “Nihon chūsei shi kenkyū to kikō hendō ron,” 30; Nagahara Keiji, Taikei Nihon no rekishi 6: Nairan to minshū no seiki, 314 – 316. 22. Liu Zhaomin, Zhongguo li shi shang qi hou zhi bian qian, 99 – 115, 200 – 304. 23. Zhang Jiacheng and Thomas Crowley, “Historical Climate Records in China and Reconstruction of Past Climates,” 833 – 849. It should be noted that unusually cold temperatures also inhibit tree-ring growth and may decrease ice floes. 24. This information is summarized from Robert Gottfried, The Black Death, 22 – 23. Gottfried was guilty of plagiarism in writing portions of this book, but his thesis and evidence were nonetheless widely acknowledged, according to the late David Herlihy. On the soils of northern Europe, see William McNeill, A History of the Human Community, 1:350 – 351. 25. Richardson Gill, The Great Maya Droughts. 26. Indeed, one might link the collapse of Tang China and Silla and the disorder of the next two centuries there to this climatic trend. 27. For the traditional argument based on archaeological data from Gunma Prefecture, see Noto Ken, “ ‘Satozumi’ shūraku no kenkyū — shūraku hensen kara mita nōkōchi no kakudai katei to sono haikei,” 3 – 29. Interestingly enough, Noto also concludes that expansion of the arable ceased during the Heian period. 28. Close behind mountain valleys were low-lying, swampy lands where water was plentiful and labor requirements also small. On these areas, see Kagose Yoshiaki, Tei shitchi; Ogawa Naoyuki, “Kantō chihō ni okeru tsumida no denshō,” 3:93 – 134, 4:45 – 126. Also see Farris, Population, Disease, and Land, 100 – 105. 29. To be sure, at Toro (Yayoi period), there were many irrigation ditches, but these may have been used to siphon off excessive water rather than irrigate. Moreover, some historians make the argument that the elaborate technology employed in building the giant mauso-
302 | william wayne farris lea of the Tomb period (250 – 650 C.E.) could have been transferred easily to rice farming. While this remains a possibility, I find Saitō’s and Kinda’s emphasis on mountain paddies more convincing. 30. Saitō Osamu, “Inasaku to hatten no hikaku shi,” 196 – 239; Farris, Population, Disease, and Land, 94 – 117. 31. Kinda Akihiro, “Heian ki no Yamato bonchi ni okeru jōri jiwari naibu no tochi riyō.” Kinda’s study and other works are described briefly in my Population, Disease, and Land, 96 – 97. 32. Kinda Akihiro, Jōri to sonraku no rekishi chiri gaku kenkyū, 267. 33. Saitō, “The Frequency of Famines,” 27. To be specific, the late medieval era includes the fifteenth and sixteenth centuries, and the early Tokugawa means the seventeenth. Moreover, climate during the late Tokugawa period was subject to a “little ice age,” providing an additional reason that no major droughts took place at that time. 34. The physics of this phenomenon works as follows. Richly vegetated mountains tend to be cooler and slow down winds, making for both more rainfall and retention of water. Bald mountains are warmer, and clouds pass right by, so the soil becomes parched. To support this, we could note that the eastern Inland Sea region, where vegetation has been stripped, receives less rainfall. 35. Toda Yoshimi, Nihon ryōshu sei seiritsu shi no kenkyū, 280 – 319, has even given examples of aristocratic prohibitions placed on the wanton logging of mountains in the ninth and tenth centuries. Also note Okita Masaaki, “Yamato seiken ka no shizen to ningen,” 55 – 56. 36. Ruijū sandai kyaku, Order of the Council of State, Kōnin 11 (820) 7/9, 328. 37. For the best recent analyses, see Isogai Fujio, “Kodai chūsei ni okeru zakkoku no kyūhōteki sakutsuke ni tsuite,” 111 – 121 (reprinted and expanded in his Chūsei no nōgyō to kikō, 18 – 56); and Kimura Shigemitsu, Hatake to Nihonjin, 47 – 71. Also note Farris, Population, Disease, and Land, 106 – 107, for a translation and analysis of an earlier (723) order. 38. The following discussion is drawn from Tamura Noriyoshi, Nihon chūsei sonraku keisei shi no kenkyū, 374 – 411. On the conversion of rice paddies to dry fields, see Farris, Population, Disease, and Land, 107. 39. Kishi Toshio, “Shakai keizai,” 143. 40. The information for each famine may be found in court annals such as the Rikkoku shi until the late ninth century, and then Nihon kiryaku, Fusō ryakki, or aristocratic diaries thereafter. Fortunately for students of famine, these data have been assembled almost completely in three compilations: Ogashima Minoru, ed., Nihon saii shi, 1 – 20; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 3 – 93; and Sasaki Junnosuke, ed., Nihon chūsei kōki kinsei shoki ni okeru kikin to sensō no kenkyū, 9 – 14. My narrative follows these sources in general; for quotations the specific record will be noted. 41. Nihon shoki, Tenmu 5 (676) 5/7. See also Nishimura and Yoshikawa, Nihon kyōkō shi kō, 3. 42. Nihon shoki, Tenmu 11 (682) 7/27 (453); Shoku Nihongi, Monmu 1 (697) int. 12/7; Nishi mura and Yoshikawa, Nihon kyōkō shi kō, 3 – 4. 43. Shoku Nihongi, Keiun 1 (704) 4/19; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 5. 44. Shoku Nihongi, Keiun 2 (705) 4/3; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 5 – 6. 45. To complete the story, the correlation coefficient for disease and famine between 806 and 887 drops to 79. 46. Ruijū sandai kyaku, Imperial Edict, Reiki 3 (717) 5/11, 429 – 430; Nishimura and Yoshimura, Nihon kyōkō shi kō, 9. 47. Shoku Nihongi, Tenpyō 4 (732) 5/22, Tenpyō 5 (733) 1/27, 2/7, 2/16, 3/16, and int. 3/2; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 12 – 13.
Famine, Climate, and Farming in Japan | 303 48. Shoku Nihongi, Tenpyō 19 (747) 2/21. See also Nishimura and Yoshikawa, Nihon kyōkō shi kō, 14. 49. Shoku Nihongi, Tenpyō Hōji 8 (764) 3/14; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 17 – 18. 50. Shoku Nihongi, Tenpyō Jingo 1 (765) second month; Nishimura and Yoshikawa, Nihon kyokō shi kō, 19. One to equals about .496 U.S. bushel, and one koku, 4.96 U.S. bushels. 51. Shoku Nihongi, Hōki 4 (773) 3/14; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 22. The existence of wealthier persons with grain reserves helps remind scholars that famine may also be the result of maldistribution of food. The role attributed to these rank holders is similar to that associated with the wealthy gentry in later Chinese sources. See Joanna Handlin Smith, “Chinese Philanthropy as Seen through a Case of Famine Relief in the 1640s.” 52. Shoku Nihongi, Hōki 10 (779) 8/2 for the Inaba peasants; Hōki 11 (780) 7/22, for the rich woman; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 25. 53. See, for example, Engi shuzei shiki, Shokoku hontō no jō, in SZKT, 26:643 – 652. 54. For the figure of 44,000, see Shoku Nihongi, Enryaku 9 (790) 8/1; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 27 – 28. See also Sawada Goichi, Nara chō jidai minsei keizai no sūteki kenkyū, 189. Later, on p. 250, Sawada uses the 823 regulation mentioned in note 55 to make an estimate of 469,000, a possible hint of the toll repeated epidemics and famines had taken on the population of the island. 55. For a brief synopsis in English, see Farris, Population, Disease, and Land, 146, 198. The 823 regulation may be found in Ruijū sandai kyaku, Remonstrance of the Council of State, Kōnin 14 (823) 2/21, 434 – 437. It is, of course, one more sign of the harsh times in Kyushu. 56. On private rice loans, see Nihon kōki, Enryaku 18 (799) 2/15; on the Sanuki case, see ibid., Enryaku 18/5/28; on the aid to Awaji, see ibid., Enryaku 23 (804) 1/23; on the growing disparity between rich and poor, see ibid., Daidō 1 (806) 5/6; Nishimura and Yoshikawa, Nihon kyokō shi kō, 29 – 31. 57. See Nishimura and Yoshikawa, Nihon kyōkō shi kō, 31 – 35. For 808, see Nihon kōki, Daidō 3 (808) 4/21; for 812, see ibid., Kōnin 3 (812) 5/18. For 814, see Nihon kōki, Konin 5 (814) 7/21 as well as Nihon kiryaku, Kōnin 5/6/3. The famine of 817 is chronicled in Nihon kōki, Kōnin 8/5/21, while that of 818 appears in Nihon kiryaku, Kōnin 9/3/19. 58. For the 823 policy toward the capital see Nihon kiryaku, Kōnin 14/2/1, 3/16; for the rice doles, see Ruijū sandai kyaku, Order of the Council of State, Tenchō 10 (833) 7/6. Also note Nishimura and Yoshikawa, Nihon kyōkō shi kō, 38 – 42. 59. On the irrigation works, see Shoku Nihon kōki, Jōwa 7 (840) 6/13; for the Dewa peasants, see ibid., Jōwa 8 (841) 2/13; for the Wakasa rankholder, see ibid., Jōwa 14 (847) 5/22; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 45 – 48. 60. Shoku Nihon kōki, Kashō 2 (849) 10/23; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 49. 61. The text of this difficult wooden tablet may be found in Hakkutsu sareta Nihon rettō: 2001, 51. Also on possible famine victims for the mid-ninth century, see Mikami Yoshitaka, “Kodai ‘henkyo’ no minshū haaku.” The original texts of Mikami’s lacquer documents are contained in Akita jōseki, 1 – 19. 62. On these wealthy farmers, see Cornelius Kiley, “Provincial Administration and Land Tenure in Early Heian,” 279 – 282. 63. On the situation in Owari, see Nihon Montoku tennō jitsuroku, Saikō 1 (854) 2/15; on the aid in Iwami, see ibid., Saikō 1/3/23; on the victims in Hizen and Bungo, see ibid., Saikō 1/4/26. On the problems in Dewa, see ibid., Saikō 2 (855) int. 4/19, 10/19; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 50 – 51. 64. Nishimura and Yoshikawa, Nihon kyōkō shi kō, 52 – 57. For 862, see Nihon sandai jitsu
304 | william wayne farris roku, Jōgan 4 (862) 5/23; for 863, see ibid., Jōgan 5 (863) 1/27; for 866 and 867, ibid., Jōgan 8 (866) 2/16 – 10/8, and Jōgan 9 (867) 2/13 – 4/22. The Ever-normal Granary is mentioned in the same source, Jōgan 9/4/22, and the quotation comes from the same date. 65. Nishimura and Yoshikawa, Nihon kyōkō shi kō, 57 – 66. In all cases, the information is drawn from Nihon sandai jitsuroku. See Jōgan 11 (869) 7/2; Jōgan 12 (870) 5/26 and 10/25; Jōgan 13 (871) 3 – 8/7; and Jōgan 14 (872) 7/24 – 25; Jōgan 15 (873) 12/2; Jōgan 16 (874) 7/29 and 16/11/27; Jōgan 18 (876) 3/21 and 7/11; Gangyō 1 (878) 7/3 (contains the story about Jingū’s tomb). For the grain relief from Harima, see Gangyō 2 (879) 1/15. On the jōhei sho, see Gangyō 2/1/27. For the information on Hōki, see Gangyō 6 (882) 6/25. 66. Nihon sandai jitsuroku, Gangyō 4 (880) 2/25; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 67. 67. Nihon kiryaku, Engi 8 (908) 9/9; ibid., Engi 13 (913) 9/9; ibid., Engi 17 (917) seventh month, on the robbers; ibid., Engi 17/12/19; on the robbers of 931, see Fusō ryakki, Shōhei 1 (931) 2/8; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 70 – 75. 68. On the distribution of cash in 942, see Honchō seiki, Tengyō 5 (942) 4/9; Nishimura and Yoshikawa, Nihon kyōkō shi kō, 76 – 78. 69. Nishimura and Yoshikawa, Nihon kyōkō shi kō, 82; Nihon kiryaku, Tentoku 2 (958) int. 7/9. 70. William McNeill, Plagues and Peoples, 5 – 6. Microparasitism refers to bacteria and viruses that find a source of food in human tissues.
12
d Charlotte von Verschuer
Life of Commoners in the Provinces The Owari no gebumi of 988
T
he elites of Kyoto — emperors, court officials, monks, literati, and poets — have been the focus of many studies of Heian Japan. What is often overlooked is that the local population in the provinces provided the food, clothing, and various supplies the elites needed to perform their daily duties. And although the commoners rarely appear in the historical sources or the literary works of the age, they were of central importance to the livelihoods of the central elites. This chapter focuses on provincial life and examines Heian society from a material point of view in an attempt to explore this codependence. For this purpose, the Petition of Owari Province in 988 (Owari no kuni gunji hyakushōra gebumi, hereafter referred to as the Owari no gebumi) provides an informative window into the lives of district officials, local notables and managers, corvée workers, and peasants.1 The petition, consisting of thirty-one articles, details the mismanagement of the incumbent Owari provincial governor (zuryō), Fujiwara no Motonaga (dates unknown), and was submitted to the imperial court by district officials (gunji) and local notables (hyakusei).2 The court examined the petition in the second month of the following year, and Motonaga was dismissed from office two months later. This protest was part of a series of provincial grievances against abuses by provincial governors during the Heian period. Eighteen similar petitions were submitted by twelve different provinces between 971 and 1041, of which six resulted in the dismissal of provincial governors. These petitions likely alerted the court to severe problems in the provinces, resulting in the implementation of new measures to contain such abuses. Of the eighteen petitions, only the submission by Owari Province has been preserved in its entirety, causing it to receive attention by scholars since the Tokugawa age (1600 – 1868). Still, some historians have questioned its authorship because of its “elaborate writing style” (meibun) and the unlikelihood that the local district officials who supposedly authored the report would have such literate training. They have accordingly attributed its authorship to political rivals of Motonaga at the imperial court, provincial literati (hakase), or literate individuals, suggesting that the petition was forged for political aims and is not indicative of material conditions in the province.3 However, this seems to be a misevaluation, and it is more likely that the petition was indeed authored by local notables of Owari.
306 | charlotte von verschuer
Map 12.1. Owari Province. Source: © Mikael Adolphson. Adapted from Kodai chimei daijiten, 2:223.
This study discusses relations between the court and the provinces by examining the contents of the Petition of Owari Province. For my translations I have used the three oldest manuscript copies (as the original manuscript is not extant): the Sōbon manuscript of 1281 in Waseda University printed in Nihon shisō taikei, the Tōbon/Tōdaibon of 1311 at the Historiographical Institute (Shiryōhensanjo) of Tokyo University, and the Shinbon of 1325 of Shinfukuji in Nagoya.4 In this essay, I have chosen to limit my focus to entries that center on the communication, taxation, and transportation systems.
Communicating between the Center and the Periphery To maintain control of the provinces, the imperial court deployed governors to the sixty-eight provinces, sent orders by official messengers, and collected information on administrative matters by regular envoys. From the eighth century, four types of envoys assumed a central role in the court’s control of the provinces: envoys in charge of yearly accounts in the provinces (shōzei chōshi), envoys in charge of taxes in kind (kōchōshi), envoys in charge of household registers (keichōshi), and envoys in charge of reports on provincial personnel (chōshūshi). These and many other messengers traveled between the capital and provinces on a network of official highways that formed what I shall call the infrastructure of the state communication system.5 According to the ritsuryō regulations (the Taihō Codeof 701), the entire
Life of Commoners in the Provinces | 307
realm was connected to the central region (Kinai) by seven official roads. Three highways — the Hokurikudō, Tōsandō, and Tōkaidō roads — ran northeast and stopped at Sadō and today’s Sendai region (see map 1.1). Three other highways — the San’indō, San’yōdō, and Nankaidō roads — ran southwest, and the Saikaidō road covered Kyushu and extended to the Tsushima and Iki islands. During the Heian age, the first six roads met at a crossing south of the capital, linking the center to its peripheries. The conditions of these highways can be gleaned through several archaeological discoveries, which include a finding in 2001 near Kanazawa city in Ishikawa Prefecture. Archaeologists in the Prefectural Bureau of Cultural Property excavated this site, which probably corresponds to the ancient Fukami post station situated on the Hokurikudō highway, on the way to the Notō peninsula. They unearthed eighty-five meters of this road, which was made of consolidated earth bordered on both sides by a water channel up to one-half meter wide and forty centimeters deep. The road was seven meters wide in the eighth century, but some repair work was done in the early ninth century, probably because of insufficient water drainage. The channels were widened and deepened to one meter on the east side. At the same time the road’s width was reduced to five meters.6 The Taihō Code also mandated a system of post stations (umaya or eki or yaku) that were managed by provincial authorities along the highways.7 The Engi shiki, a collection of court proceedings compiled in 927 C.E., stipulated that post stations were to be situated at a distance of about thirty ri (ten miles or sixteen kilometers) from each other and to provide lodging, food, and horses for traveling officials.8 In all, some 402 stations across the country provided 3,487 station horses (ekima or yakume) from their own cattle breeding or purchased from other pastures, and a number of districts in each province provided another 687 relay horses (tenma or denme) obtained from military pastures or bought from other pastures. Each station had at its disposal a plot of rice fields assigned to cover the costs of its management. There were also a fixed number of tax households (ekiko or yakuko) who worked on the station’s rice fields and pastures. Governors and other officials with imperial messages who traveled from the capital to the provinces received tokens noting the number of horses allowed according to their court rank and distance traveled. The provincial government also provided tokens to envoys, who delivered reports on local affairs to the capital. The time it took to travel long distances varied depending on purpose. For example, express messages to the capital to report rebellions, movement of provincial armies, epidemics and droughts, fires in public buildings, omens of good fortune, shipwrecks, and the return of Japanese embassies were high priority and could reach the capital in little time. One of these express messages suggests their travel time. In 836 a messenger announced that a Japanese mission to China had shipwrecked shortly after its departure from Kyushu. At a distance of eight hundred kilometers from Dazaifu to Kyoto, the messenger took eleven days (covering seventy-two kilometers a day).9 However, in cases of large groups carrying taxes to the capital, the same journey took twenty-seven days, and fourteen days to return without any goods.10
308 | charlotte von verschuer
As for sea routes, the Taihō Code stipulated that travel from Dazaifu to Kyoto should take thirty days, and provincial governors assigned to Shikoku and Kyu shu traveled by sea from the early eighth century.11 To facilitate such travel and transports, several ports were important along the route: Hakata harbor in Kyu shu, Ōtsu on Lake Biwa, and Naniwatsu on the Inland Sea. From Naniwatsu, which served as the nexus of the Inland Sea routes, ships traveled up the Yodo River to the Yodotsu port south of the capital, and travelers proceeded on foot by the San’yōdō route. Sea travel, however, was not free of risks, and shipwrecks were common. Storms and pirates caused much insecurity, and in the case of the latter, Fujiwara no Sumitomo (? – 941) ravaged the Inland Sea with his fleet of several hundreds of ships. For example, Ki no Tsurayuki (? – 945) in the Tosa nikki was concerned about pirates near Murotsu during his return to Kyoto from Tosa Province after ending his appointment as governor. As for Owari Province, which was situated on the Tōkaidō route, we know that its governors received traveling food supplies for trips to the capital in accordance with the Taihō Code but had to provide their own horses because horses were not provided for journeys from provinces within close geographic proximity to the capital. By contrast, officials traveling from Kyoto through Owari Province and farther eastward were provided with horses at three stations in the province — Umazu, Niimizo, and Futamura — as well as at the relays of Ama and Aichi districts in Owari.12 The eleventh entry of the Petition of Owari in 988 contains a specific article addressing problems associated with the travel network. Request for an official decision: [the governor] fails to provide traveling food supplies (denjikiryō) to all stations, as well as the income (atai) from 156 chō of tenant rice fields (kubunden) [assigned to the stations] and due to support the station employees. The corvée dues are heavy in our province, but nothing exceeds the [hardships of] corvée at the stations. Since ancient times food supplies have [always] been provided for official envoys traveling to and from the capital, and the income from the [station] rice fields has [always] been assigned as food and allowances for station employees. There are 12 chō of assigned rice fields (ryōden) for each station and another 16 chō of rice fields for the district relays, totaling 52 chō of [station and relay] rice fields. However, the incumbent governor [Fujiwara] Motonaga no ason seized the income of both for three years and did not assign anything [to the stations]. Nothing is more important than the management of the territory of a province. [For instance] when the [eastern provinces] send [the horses as tribute to the court], the inspectors of the pastures (kenboku no tsukai) who travel between the provinces and the capital take advantage of the authority entrusted to their mission in connection to the [tribute] horses, and they ignore the hardships of local corvée workers. They pretend that the food supplies [at the stations] are inadequate and would look for [problems as small as] wounds between the horses’ hair [and inflate them]. Or they would complain about the inadequate welcome treat and [create prob-
Life of Commoners in the Provinces | 309
lems by] piercing the [horses’] skin, making them bleed. They collect many bribes and feed the tribute horses with fodder and linger [at our expense]. They obtain at their will [all sorts of] local products (dosan) and drive the horses all over by whipping them with a horsewhip. If the governor were a good official, and would refrain from withholding food supplies [for the traveling inspectors], the district officials (gunji) and the local notables (hyakusei) would not have such troubles. We therefore request an official decision in order to rapidly secure the condition of the station employees. The grievances thus focused on the maintenance of the stations. Local authorities complained that Motonaga seized the income from rice fields allocated to cover the expenses of station employees. As noted in the petition, these rice fields consisted of a plot of 12 chō (1 chō = 1.13 hectares or 2.79 acres) for each of the three stations and 16 chō for the horse relays of two districts, for a total of 52 chō. As Motonaga seized the income for three years, this amounted to 156 chō.13 Thus, from the point of view of the petitioners, the management encountered great difficulties because the governor seized for his personal benefit the dues earmarked for the maintenance of the stations. Motonaga’s malfeasance caused additional problems for local officials. Among the envoys, the entry specifically mentions those who inspected the pastures of the northern provinces. The central government administered two kinds of pastures: military pastures, which provided the horses for regional militias spread over eighteen provinces, and imperial pastures, which raised horses for the imperial court and were situated in four northern provinces: Kai, Shinano, Kōzuke, and Musashi. These provinces sent altogether 240 horses to the capital each year. They served at annual ceremonies such as archery competitions, horse races, and the celebrations of the new horses (aouma no sechie). Horses were also used for imperial pilgrimages and processions, and by various officials such as imperial bodyguards (konoe), the palace gate guards (emon), and imperial police guards (hyōe). During the ninth month of each year, in preparation for the annual delivery of horses, a member of the Office of Horses (Meryō) inspected the imperial pastures in the four northern provinces and compiled a list of the fouryear-old horses to be delivered to the court in the eighth month of the following year. Horses from Shinano and Kōzuke provinces were escorted on the Tōsandō road, while those from Kai and Musashi provinces traveled on the Tōkaidō road through Owari Province. These provinces were required to supply fodder and horse guides from one station to another, in addition to providing food and lodging for inspectors and escorts. By the ninth century, the court noticed problems in the horse registers and in the late delivery of horses. An 862 decree also noted for the fourth time since 758 that the guides were letting the horses run loose and that accompanying envoys caused disturbances at stations.14 The 988 Owari petition mentions similar problems. The horse escorts claimed that supplies at the stations and welcome banquets were inadequate, and they caused problems by letting the horses run loose, even injuring some of them in order to obtain bribes and extort local products. In addition, the escorts lingered
310 | charlotte von verschuer
at the post stations, exhausting the limited supply of fodder, instead of promptly continuing their trip to the capital. In the view of the petitioners, all this was the fault of the governor, Motonaga, because he retained the food supplies for himself. Finally, it is worth noting that these complaints are expressed by the district officials and the local notables who acted as spokespersons for the station employees and corvée workers. Thus the eleventh entry of the report gives details about the local management of the communication system. As noted above, the communication system was an essential tool for the administrative control by the imperial court of the provinces. In addition, the tax income from the provinces was essential to the material survival of the imperial court.
Provincial Taxes and the Economic Viability of the Court The Nara and Heian courts obviously depended on the provinces for their income. According to the Taihō Code and the Engi shiki regulations, there were three main types of taxes: land tax (so) in rice, per capita taxes in kind (chō or mitsugi), and labor services.15 The land tax system was based on the allotment of plots of tenant rice fields to cultivators. An average household of ten people would cultivate more or less one hectare of this kind of public rice fields. The cultivators grew other grains and vegetables on other farmland, but only rice fields were taxed. The land tax was 1.5 sheaves of rice ears or 1.5 to (12.75 liters) of rice grains per tan (1 tan = 0.279 acres) As one tan produced between fifteen and fifty sheaves of rice, the tax theoretically amounted to 3 – 10 percent of the total rice production,16 but the burden was de facto higher because of losses of land (by canals, field dikes, etc.) and crops (due to natural disasters), as well as other additional levies in rice. In addition to the land tax, cultivators were burdened with a compulsory public rice credit tax (suiko). During the spring, when little rice remained, local authorities provided loans to farmers at an interest of 30 – 50 percent to be collected after the autumn harvest. This rice credit amounted in the ninth century to 10 – 100 sheaves per adult man (or per farm) or a minimum 3 sheaves per tan of land. Hence it was not uncommon for cultivators to have lost one-third to one-half of their crops after the payment of taxes and interest. These losses were exacerbated in times of climatic fluctuations that in turn resulted in further accumulated debts.17 The land tax and rice credit equally provided the annual income for local governments and were utilized within the provincial and district administrations for operational costs and as a reserve resource for the rice credit. It is important to note that only a limited amount of the rice income collected was sent to the imperial court. Let us consider some figures from Owari Province to understand the complaints of the petitioners in 988. According to the register of annual resources of Owari Province dated 734, one of only twenty-five (fragmentary) registers still extant today in the Shōsōin collections, Owari Province had a land tax income of 10,000 koku (1 koku = 85 liters = 10 sheaves) of rice grain and an accumulated reserve (or capital) of 258,440 koku. In addition, it gave a total credit of 225,704
Life of Commoners in the Provinces | 311
sheaves (or 22,570 koku) and earned 97,946 sheaves (or 9,794 koku) as interest, suggesting an average 42 percent interest rate. This was indeed a heavy burden, and it may have contributed to the death of 256 farmers that year due to food shortage. In 734 the province paid the imperial court a total 20,000 sheaves of rice (or 2,000 koku unhulled rice). This amount remained in practice up to the tenth century, as 2,100 koku are still prescribed for Owari in the Engi shiki regulations in 927.18 Apart from the land tax and the rice credit, provinces gained another source of rice revenue through the lease of surplus public farmland to tenants. The rent (jishi) amounted to one-fifth of the harvest and was assigned to all sorts of public services both in the provinces and in the capital (militia, guards, stations, etc.). In 799, twenty-one provinces delivered part of the collected rice rent to the capital, and in 914 there were still fixed amounts of provincial rent incomes due for delivery to the imperial court. At that time in 914, the amount from Owari was 150 koku or 1,500 sheaves.19 It can be assumed that, at the end of the tenth century, Owari still levied the above-mentioned several kinds of rice taxes from the farmers. But in 988 the governor overcharged the population of his province and even seized part of their own rice earnings. The third entry of the petition states as follows: Request for an official decision: [The governor] seizes at his will a land tax of 3.6 to of rice per tan of land, exceeding the official rule (kanpu). The land tax (sokoku) is regulated by official rules. For several generations former administrators recognized — although with regret — some losses of the harvests according to precedents [and reduced taxes. But the present governor has ignored losses and] levied the full amount of taxes. [Previous] governors collected 1.5 to, others more than [less than]20 2 to. Yet the current governor Motonaga no ason charges us 3.6 to. This is contrary to precedents. First of all, the way of government is comparable to simple cooking; [a leader with] a heart filled with compassion for the people will [easily] tame [any wild birds, even] pheasants. However, [in our province many cultivators] are giving up tilling the soil in spring and are ready to run away (hokumin). Moreover, a lord of a castle has to behave a certain way; and a shepherd of a whole region even more has to encourage agriculture. [We know from Chinese classics that] during the growth of rice plants in spring and at the moment of their ripening in summer, [the benevolent administrator] Liu Kuan used only a [soft] whip to punish lazy farmers who left their hands idle; and [the rich landlord] Wang Dan rewarded with [feasts and] wine those who endeavored to work with pride. [In contrast, the governor of our province] every year during the tilling period in the fourth and fifth months sends all sorts of envoys to our places urging us by saying, “The rice you received in accordance with precedents as payment for the purchased tax products (kōeki zōmotsu), you have to rapidly hull (tsuku) it and deliver it [to the provincial government].” [Upon this] the district officials and local notables (hyakusei) have no means [to refuse] and have to assume the burden in spite of their difficulty. They hull [even] the soaked rice seeds
312 | charlotte von verschuer
and deliver the rice to the government grain storage. In the meantime, farmers leave their spades and interrupt tilling; women who breed the silkworms forget [to collect] the mulberry [leaves] and stop spinning. Is this not enough cause for the people (hyakusei) to grief? Also, this will enhance the lack in preparing the tribute for the court. We demand that the mandate of Motonaga no ason be terminated and that a new and just administrator be appointed. Excessive land tax is thus the main complaint against Motonaga. As indicated by the petition, the land tax rate fluctuated during the tenth century and no longer adhered to the rate of 1.5 sheaves (or 1.5 to) per tan of land set by the Taihō Code. Hence governors were provided with broad discretion to set the tax rate. From the end of the ninth century, provincial precedents (kokurei) gradually emerged according to local customs, which governors followed upon taking office. It is likely that the petitioners are referring to these provincial precedents when making note of “official rules” and “precedents.” According to the petition, 3.6 to of rice per tan exceeded the prevailing practice in Owari Province, which the petitioners claim to be either 1.5 or about 2 to of rice.21 The petitioners also note that former governors lowered levies in times of reduced harvest, a practice introduced by the Taihō regulations, which exempted cultivators from the land tax if the region sustained losses from half its harvest.22 The entry indicates that the governor levied an excessive amount of land taxes and even seized rice earnings from the local elites of the districts. The governor also harassed the peasantry in the busy period of the year — in the fourth month, when the rice seedlings grow and the water level has to be adjusted, and in the fifth month, when the seedlings are transplanted to the main fields. The local notables feel forced even to hull the rice seeds that have been soaked in water for about ten days in the second month before being sown.23 Apart from the land tax and other rice taxes, the government levied per capita taxes. From the eighth century, there were three types of per capita taxes in kind payable in craft or food products: tribute, labor tribute (yō), and young people’s tax (chūnan sakumotsu). According to theTaihō Code, tribute comprised alternatively plain woven silk taffeta (kinu), silk thread (ito), silk floss (wata), and hemp cloth (nuno), or various products such as iron ingots or hoes; salt; or seafood such as finned fish, shellfish, and seaweed, which were either dried, cooked, salted, or marinated for conservation. Other products were later added, and items for each province are listed in the Engi shiki. By the tenth century, tribute included more sophisticated weavings such damask (aya) and twill (ra) silk, which were likely produced in collective local workshops. The labor tribute basically consisted of hemp cloth with some alternatives depending on the court’s needs. For some provinces, the Engi shiki stipulated products such as brocade (nishiki), indigo dyes (ai), kunugi oak bark, wooden chests, and food products such as rice and fish. Young men of seventeen to twenty years were exempt from the corvée but paid a supplemental tax (soemono) or young people’s tax. It included plant dyes such as Indian madder (akane) and Cape jasmine (kuchinashi); various plant fibers; plant oils such as sesame (goma) oil and perilla (egoma) oil; food products;
Life of Commoners in the Provinces | 313
and crafts products such as straw mats, bamboo trellises (sunoko), kegs, paper, lacquer, feathers, and roebuck horns. Precious silk fabrics, silver ore, ceramics, bamboo boxes, wooden tubs (ancient oke, modern magemono), other straw mats, animal skins, oils, and food products were later added.24 In addition to these individual taxes, other products were levied at the collective level, including administrative village specialties (nie), district specialties(domō), and provincial contributions (kōkenmotsu), with the last two being later combined as purchased tax products (kōeki zōmotsu), mentioned in the two petition entries above and below. They consisted of raw materials, crafts, and various foodstuffs purchased with its rice earnings by the provincial government from the local workshops in the districts. In practice they comprised some of the earlier tribute taxes stipulated in the Taihō Code.25 In addition to all these taxes the Engi shiki makes note of various special contributions such as medicinal herbs, an unfermented milk product (so), fruits, and other local products.26 The public management of the tax collection underwent several changes from the eighth to the tenth centuries, basically shifting from a per capita levy to a land-based levy.27 For example, rice credit was granted not to persons but according to acreage. From the tenth century onward various taxes fell into two basic categories: tax products (kanmotsu) and various corvée services (rinji zōyaku). By this time, not only rice payments on allocated plots and rented farmland but also payments in handicraft and food products were levied at a land rate (den ritsu), meaning that they were calculated according to acreage of land and not the number of individuals. From mid-eleventh century, the corvée services were also levied at a land rate and melted together with the other payments. This merging of dues resulted from difficulties in accurately gauging the number of residents as population registers ceased to be drafted in the early 900s. Cultivators were dispersed on allotted plots or public rented land; some even fled their homes. Some of these independent vagrants (furō or rōnin) managed to open new lands or to collaborate with private estates (shōen). A group of rich farmers emerged. These rich cultivators gained control of larger stretches of farmland, assembled manpower, reached arrangements with local authorities, or increased their income by selling credits of their surplus rice. They came to play a key role in the management of public lands and taxes at the local level, and were used by provincial governors as contract agents on units of land tenures (myō) and appear in the sources as managers (fumyō) who were responsible (fu) for levies of taxes and rents on their unit of tax assessment. These agents seem to correspond to the hyakusei of the districts in the report of Owari. On the provincial level the administration developed in a similar manner. Governors (kami) who had represented the imperial court as provincial administrators in the eighth century became provincial managers or a kind of contract governors, who managed tax collection and sent a fixed amount to the capital, from the early tenth century. The court granted these governors the inspection right of rice fields (kendenken), which in turn gave them the authority to determine the yield per acreage and, in some cases, set the land tax. Governors established the amount of the tax in rice at their discretion but for the most
314 | charlotte von verschuer
part adhered to provincial precedents, although in practice, governors did not necessarily follow prescribed guidelines. While the taxation rights of governors were initially designed to secure the court’s income, they inadvertently gave rise to various abuses. Only in the middle of the eleventh century did the court fix a limit of 3 to of rice per tan of land (kōden kanmotsu rippō) in order to control such abuses. But before such measures were implemented, governors retained extensive tax authority and were not yet subjected to limitations imposed by the court. As such, a four-year tenure provided many provincial governors an opportunity to enrich themselves at the expense of the local population. This system also paved the way for opposition between the local district authorities who provided the tax products and the province governors who levied them. This tension is evident in the seventh entry of the Petition of Owari in 988: Request for an official decision [on the following matter]: By pretending to levy the tax of purchased [tax products (kōeki zōmotsu)] [the governor] is cheating us by taking silk taffeta, handwoven hemp cloth, Shinano hemp cloth, regular hemp cloth, [lacquer28], oil, China grass (karamushi), cloth, Indian madder, and silk floss. Among the purchased tax products mentioned above, there is a limit to annual furnishings of silk to the [provincial] government. However, the total amount collected in our province [from the districts] reaches several thousands of bolts (hiki). [The fabrics] are delivered between the middle of the fifth month [June] and [the end of] the ninth month [October]. [The governor] takes [the fabrics from our workshops] at a price of 40 to 50 sheaves of rice for a [bolt] of silk taffeta, 8 sheaves or more for [a bolt of] handwoven hemp cloth, and 5 or 6 sheaves for [a bolt of] Shinano hemp cloth, [and of regular hemp cloth29]. The other products’ prices are negligible. After [the prices are set], he claims that they went down and pretends that there is a remnant [of undelivered fabrics to be furnished by our workshops]. He sets [the price difference of undelivered] silk and hemp cloth on the basis of the originally fixed price. Upon this the tax commissioners arrive with many followers and seize the specialties (domō): 1.5 [1.6] koku of rice per [undelivered] bolt of silk and 4 to [5 to] of rice per [undelivered] bolt of hemp cloth.30 Other products follow the original prices [and are not charged extra]. [The total] amounts to three or four times [the originally fixed quantities]. And that is not all. It is impossible to calculate the expenses for food supplies and equipment [charged to us at each visit by the tax commissioners]. In order to pay our debts caused by these illegal furnishings, we have to sell the property of our ancestors and to endanger the survival of our children and grandchildren. We also have to sell even the clothes of husbands and wives and are unable to protect our beloved children from the heat [in summer] and the cold [in winter]. Because of the greed of just one person the lives of local notables (hyakusei) are in danger. Therefore the people (jinmin) of our realm frown their eyebrows [in anguish] and are crying and moaning. The vagrants (rōnin) of our area are unable to join their heels (kakato o katamukeru) [or to stand upright, because of their misfortune]
Life of Commoners in the Provinces | 315
and are filled with sadness and despair. All this is due to the incapacity of the governor [Motonaga]. We request a decision [of the court] in order to soon be freed from his illegal (hihō) harassing. This entry concerns purchased tax products, a collective tax paid by Owari province to the imperial court. According to the Engi shiki, these purchased tax products varied according to each province’s specialty and for the most part consisted of textiles, animal skins, cereals, various plants, seaweed, foodstuffs, handicrafts, lacquer, and other raw materials. In the case of Owari, the Engi shiki prescribed 150 bolts of silk taffeta, 12 bolts of white silk taffeta, 3 koku of oil, and 100 bolts of cloth of China grass, as well as animal skins, grains, seaweed, and glazed ceramics.31 Although the calculation of the prices indicated in the seventh entry is difficult to follow, one can imagine the following scenario underlying Motonaga’s deceit. He first collected silk at a price of 45 sheaves or 4.5 koku of rice per bolt of silk that he then allegedly reduced to 1.5 koku or 15 sheaves per bolt, pretending that he paid for three bolts of silk but received only one, hence leaving two bolts of silk undelivered. Then, instead of taking two more bolts of silk in kind, he charged the price in rice and seized 3 koku of rice for the undelivered silk. Motonaga’s cheating consisted in extra charging for the so-called undelivered silk. As for the price reduction, from the early tenth century, the governors of several provinces experienced difficulties in providing the full amount of purchased tax products (such as silk) to the court, claiming that revenues from provincial rice harvests were insufficient. It seems that they were sometimes allowed to reduce the quantity or even the prices of the products.32 In this respect Motonaga’s initiative of a price reduction does not seem contrary to precedents, but the rate was excessive. In the tenth century, governors retained some latitude in managing the taxes and were not yet controlled by court restrictions introduced in the middle of the eleventh century. Evidently some governors took advantage of this situation, and our entry shows how they were able to enrich themselves through the tax system. Here again, the petitioners are not commoners who work in weaving workshops but local elites such as the independent vagrants (rōnin) noted above. These vagrants might have played a role in providing the purchased tax products to local authorities.33 Let us now examine the various taxes in kind delivered from Owari and their use at the Heian court. Owari rice supplies amounted to about 2,000 koku a year. According to the Engi shiki, the Office of Grains (Ōiryō) each year received a total 17,330 koku (1.5 million liters or roughly 1,200 tons) from twenty-two provinces. High-quality hulled (or 50 percent hulled) white rice (hakumai) was delivered directly to the Office of Grains; less-hulled black rice (kokumai) was sent to the grain stores of the Ministry of Court Affairs (Nakatsukasashō). The rice of both kinds was distributed to the court’s bureaus and offices as food allowance for officials on duty. It was also served at all court banquets accompanying the annual court ceremonies (nenjū gyōji) and was presented as an offering at many annual rituals. The rice from the provinces was therefore used for everyday needs as well as for all official events. Only the rice served to the emperor, the empress, and the
316 | charlotte von verschuer
imperial prince was produced in government fields (kanden) run by the Ministry of Palace Affairs (Kunaishō). Deliveries from the provinces became more irregular from the end of the ninth century, and some court offices cultivated their own fields, but the court continued to depend on rice deliveries from the provinces, as can be seen in a case in 1038 when the Chancellery (Kurōdo dokoro) noticed that there was no rice left at the Office of Grains and had to resort to other resources to prepare the offerings for the Ise Shrine.34 Silk taffeta constituted another tax product, and as suggested by the ninth entry in the petition, 404 bolts were delivered to the capital each year. Other provinces also provided yearly deliveries of silk to the court. According to the Engi shiki regulations, twenty-eight provinces delivered a total of 8,503 bolts of silk as purchased tax products, and another twenty-nine provinces provided an unknown quantity of silk as regular tribute (chō). These silk products were primarily woven taffeta silk (kinu) although there were other varieties of precious silk such as brocade and damask, as well as those of a rougher quality such as ashiginu. Among all varieties, the largest quantity needed was taffeta silk: 3,478 bolts were used to dress the emperor, and 700 bolts were used as undergarments, shirts, and vests for the imperial consort. Taffeta also dressed guardsmen and low-ranking officials on duty, as well as Buddhist monks and Shinto priests. It was also a prime material to fabricate furniture such as screens, and it was used for a variety of other purposes such as sieves for flour. In addition, taffeta served as offerings although not as often as the lesser-quality ashiginu. Every official event concluded with distributions of gratifications for its participants consisting of silk clothes for the guests and taffeta silk fabric for musicians, escorts, and lowranking officials. Fujiwara no Michinaga alone mentions such distributions about a hundred times in his journal. Taffeta silk was in high demand because it served as a kind of currency in barter for rice at rates fixed by the Engi proceedings.35 The petitioners also mention Shinano hemp cloth, which was originally woven in Shinano Province, but in the tenth century was also produced in Suruga and Owari provinces. This cloth was used in the Heian court as material in the imperial workshops, and large amounts were distributed to court officials. All officials received semestral stipends consisting of ashiginu, silk thread, hemp cloth, and iron hoes. The regulations stipulated that Shinano hemp cloth should be used for these distributions. The hemp cloth stipends amounted to 36 bolts for high officials, 10 bolts for officials of the fifth rank, and 3 bolts for officials of the eighth rank according to the Taihō Ritsuryō in the eighth century. Although the figures for the tenth century are unknown, it can be assumed that the biannual rewards amounted to thousands of bolts. Shinano cloth was also used for gifts and gratifications, especially to temples and various monks. For instance in 1038, a gift of 200 bolts was provided in honor of Kyōmyō (965 – 1038), the late abbot of Enryakuji. The aforementioned Michinaga made wide use of Shinano cloth. In 1001 he gave Fujiwara no Senshi (961 – 1001), the retired imperial dowager, 1,400 bolts as a celebration for her fortieth birthday, and from 1004 to 1015 he distributed a total of 4,023 bolts to officials and employees.36 All these examples show that
Life of Commoners in the Provinces | 317
Shinano cloth was a highly appreciated item within court circles, and it is worth noting that Owari was one of the three provinces to provide this product. An Owari product not mentioned in the petition but stipulated by the Engi shiki regulations consisted of glazed ceramic (shi no utsuwa or jiki) cups, bowls, plates, and vases. Nagato Province was the only other provider of this type of pottery to the imperial court. Only several dozen pieces of this product were delivered to the capital each year, and they were used exclusively by the imperial family. In the ninth century, when unglazed hajiki and sueki pottery were the main productions and glazed high-temperature ceramics were imported from China, the kilns of Owari were the first in Japan to make pottery at a high temperature (1100 – 1200°) through a process of burning the ceramics twice and glazing them in a gray or greenish color. It can be assumed that these pieces were influenced by the Chinese celadons. Many pieces of the Owari production, known as the Sanage kilns, have been excavated from archaeological sites in Aichi Prefecture.37 Whereas Shinano hemp and glazed ceramics were specialties of Owari Province, other tax supplies, including lacquer, oil, and Indian madder, were provided to the Heian court by many other provinces. The lacquer was stocked in the storehouses of the Ministry of Treasury (Ōkurashō) and used in the imperial workshops, which produced lacquer dishes, furniture (chests and frames for screens), palanquins, and other equipment for the imperial family.38 Oil was also used at the imperial craft shops. Owari sent to the court 3 koku (255 liters) of oil as purchased tax products and additional sesame oil as a young people’s tax. The provinces generally pressed oil from sesame seeds, perilla seeds, hemp seeds, walnuts (kurumi), viburnum (hemi, yabudemari), camellia (tsubaki), and an unidentified plant called hosoki. Sesame and hosoki oils were the main lamp fuels, and all oils were used in the court’s kitchens.39 Indian madder was a primary dyeing material. Mixed with a lithospermum (murasaki) it produced a bright red color. The court needed every year respectively 319 large pounds (214 kilos) and 140 large pounds (94 kilos) of Indian madder just for dyeing the silk of the clothes of emperors and the imperial consorts.40 Dyes, oils, ceramics, silk and hemp fabric, rice, and vegetables as well as the earlier mentioned horses are but a few examples of the yearly deliveries from Owari Province to the Heian court. They demonstrate the imperial court’s needs for provincial supplies for food, clothing, and all equipment to sustain daily life and official duties. Yet all the products had to find their way to the capital by means of an organized transport system. Once again, the petitioners of Owari provide a glimpse of its actual management.
Transporting Goods to the Capital Beginning in the eighth century, many convoys of porters carried cargoes of foodstuffs, raw materials, and craft products from the provinces to the capital every year. According to the Taihō Code, these convoys departed the provinces
318 | charlotte von verschuer
in the eighth month and arrived in the capital between the tenth and the twelfth months. The porters were corvée laborers levied every year from different households. A district official guided each group, and the whole convoy was directed by the envoy in charge of the taxes in kind (kōchōshi), who was a representative of the province. Textile fabrics were packaged in bolts, food products were wrapped in hemp cloth and straw mats or filled in jars, and ceramics were likely carried in wooden cases. Each bundle was marked with a wooden label, many of which have been excavated in Nara since the 1970s in the form of wooden tablets with inscriptions (mokkan). Written on one of these wooden tablets is the following inscription — “Owari Province, Chita district, Nieshiro village, Asakura hamlet, Wanibe no Irotachi, household chief, 3 to [25.5 liters] of tribute salt, first year of Tempyō [729]” — meaning that Irotachi, who was living in Asakura, had gathered salt as tax on behalf of his household and delivered it to the administrative village Nieshiro, which was managing the tax collection of all inhabitants of the village.41 Food provisions to the porters were provided by the households of the porters themselves, but as the households often were incapable of doing so and porters would fall ill or even die of hunger, the provisions soon came to be supplied by the provincial grain stores. Even so, there were many cases of sickness and many deaths en route. From the ninth century, the administrative villages were in charge of looking after porters who fell ill. The official rate of the food supply was 2 shō (1 shō = 0.85 liters) of rice and 2 shaku of salt per person per day on the way to the capital and half the amount on the return to the province. For example, we know from the annual account (shōzeichō) of Tajima Province that 530 porters went back and forth between Tajima and Nara in the year 737, suggesting that generally the main tax convoys from the provinces may have counted between 200 and 400 men and additional smaller convoys between 50 and 200 men.42 Horses gradually came into use as the caravans came to be composed of both porters and horses with their guides walking next to them. Rice was carried on horseback from the provinces on the Tōkaidō route, but by sea from the provinces on the Japan Sea and the Inland Sea and from Kyushu. Scroll paintings from the twelfth and thirteenth centuries show cargoes carried by horses as well as by oxen or on oxcarts. Transport on horseback and by oxcarts spread mainly in the central provinces from the ninth century, and by the tenth century there were also paid transport services by ship and by carts available in the central region.43 The tax convoys from the various provinces probably traveled on the official highways but did not benefit from any services at the post stations as these were reserved for traveling officials. The Engi shiki regulations distinguished between far, medium far, and close provinces and fixed the number of traveling days. Owari was a close province, and tax convoys were to take seven days to the Heian capital, four days to return. However, many caravans traveled more slowly than prescribed and arrived late in the capital. Although the regulations fixed punishments of up to one hundred whips of the lash for the caravan leaders, the delays were so frequent that the authorities found themselves obliged to extend the delivery time limits by several months for some provinces. In addition, part
Life of Commoners in the Provinces | 319
of the cargo was often stolen or lost en route. In such cases, district leaders and the porters were held responsible. When they reached the capital, the kōchōshi, who represented the province, delivered the cargo and the registers of the porters and of the tax products to the Ministry of Population (Minbushō), which, together with the Ministry of Treasury, checked the documents, the accounts, and the quality of the products. The handicraft products were then delivered to the storehouses of the Ministry of Court Affairs, and the rice was stored at the granaries of the Office of Grains and of the Ministry of Population.44 In the case of Owari, the governor misused these tax convoys as indicated in the twentythird entry of the petition. Request for an official decision: By infringing upon ancient precedents (kyūrei), [the governor] obliges the delivery men of the provincial government (zōshi kinin) and the population (jinmin) of the region to provide porters (tanpu) and horses and to carry various products both to the [Heian] capital and to Asa tsuma. The service of porters and horses is regulated by provincial precedents (kokurei). However, during the winter months and agricultural season, not one month, even ten days pass without frequent transports [of goods] to the capital. There is a distance of ten post stations [from our province] to the capital; and on the way back to the province, [the porters] have to climb steep peaks high into the clouds on myriads of ri. Therefore the porters suffer from inflammation of their shoulders caused by the shoulder stick as they carry [both their cargo and] their grief. The animals’ hoofs are injured and with pain they carry the cargo on their packsaddles. The porters are not even able to return to their home province, because by then the food supply is depleted and their strength exhausted. The animals are close to collapse, as the grass is dried out and the water is frozen. Sometimes there is not a single porter left in the whole province, not one transport animal left in any inhabitant’s household (hyakusei no kamado). For delivering his rice earnings (chinmai), the governor imposes [loads of] 1.2 koku for porters and loads of 2 koku for animals. It is impossible to estimate the trouble caused to the men and the damage caused to the animals. We demand a decision in order for men and animals to be free from unjust burdens that have nothing to do with the transport of [legitimate] tax cargoes to the capital. The petitioners complained that there were too many transports, that the burdens were too heavy, and that the cargoes were used for private purposes. It seems that the rice was not forwarded to court granaries but instead sent to the governor’s residence in the capital, and the same thing happened with wheat deliveries, as noted in the fifteenth entry. To a large degree, this practice is indicative of the permeability between public obligations and private interests common among Heian-period officials. The line between public and private property was not clearly demarcated. In one case, for instance, in 1038, the widow of the governor of Mutsu was granted a certain amount of rice stocked in the provincial granary during her husband’s tenure in office for her personal use. Governors were indeed able to enrich themselves oftentimes at local expense, but on the
320 | charlotte von verschuer
other hand they also fulfilled their duties to the imperial court at their personal expense.45 In addition, the petition indicates that rice was delivered not only to the capital (ten stations from Owari) but also to the port of Asatsuma, in modern Maibara on the eastern shore of Lake Biwa (see map 12.1). While the reasons for using two routes remain unclear, we can assume either that Asatsuma served as a transit port for deliveries to the capital through the Ōtsu port or that its market offered Motonaga opportunities for trade. The convoy consisted of men and horses, and in the case of the latter, when animals’ hoofs were injured — a common complaint as evidenced in the petition above — it is likely that such wounds were caused by a lack of iron horseshoes, which interestingly enough were not generally used until introduced by the Dutch in 1729. Even then, as suggested by woodblock prints such as Yotsuya naitō shinjuku, dated to 1858, by Hiroshige, horses continued to be equipped with straw sandals.46 The Engi shiki regulated the weight load of horses. A load of five to or 0.5 koku (42.5 liters) of rice was to be packed into a straw sack, with three sacks or 1.5 koku (127.5 liters) for one horse load. If 1 koku (85 liters) of unhulled rice weighed roughly 70 kilograms, one horse carried a load of approximately 105 kilograms. In the case of textiles, one horse had to carry 70 bolts of taffeta silk, or 30 bolts of hemp cloth, or 100 pounds kin (67 kilograms; 1 kin = 670 grams) of copper ore. We know a concrete example from the previously mentioned convoy of Tajima in 737 C.E. At that time one set of 26 koku of soybeans was distributed on sixteen horses each loaded with 1 koku and twenty porters each carrying 0.5 koku. Another cargo of 5 koku of fermented fish (sushi) was carried by twenty-eight porters in fourteen jars. Two men carried one jar (which was probably attached to a shoulder stick). The burden was 0.38 koku for two people plus the much heavier ceramic jar itself. It therefore seems that the average loads amounted to 1 or 1.5 koku for horses, 0.5 koku for men (or 1 koku for two men).47 Based on these examples we can understand the complaints of the petitioners of Owari: its porters carried 1.2 koku of unhulled rice weighing about 84 kilograms, distributed on two porters who suffered from an”inflammation of their shoulders.” Horses were saddled with 2 koku or 140 kilograms, an amount much heavier than the usual load. It goes without saying that these cargoes were challenging for a sevenday journey from Owari to the capital. The people of Owari thus appear to have suffered under the tax management of Governor Fujiwara no Motonaga and his retainers. Still it should be kept in mind that I have focused on only four of the total thirty-one complaints, which concern the communication, taxation, and transportation systems, as a means to examine interaction between the court and the provinces. The next section focuses on the petitioners and the nature of their grievances.
Life of Commoners in the Provinces | 321
The Petitioners and Life in Owari Province While the petitioners identify themselves as district officials (gunji) and local notables (hyakusei), it is important to note that their complaints are stated on behalf of peasants (in many entries), tax porters (entries 22 and 23), post-station employees (entry 11) and rich vagrants (entry 7). The petitioners also claim to represent subordinate employees of the provincial government such as delivery men (zōshikinin, entries 21 and 23), clerks (shosei, entry 21), and third- and fourthgrade assistants and secretaries (entry 20). Governor Motonaga, his son Yorikata (dates unknown), retainers (rōtō, rōjū, entries 15, 21, and 26 to 31), tax commissioners (chōshi, or “pressuring envoys” [seme no tsukai/sekishi in many entries]), and the vice-governor (mokudai, entries 5 and 22) are the adversaries. The two opposite parties are therefore the district officials and local notables together with some subordinate province officials and employees on one side, the upperclass province officials on the other side. While the authors of the Petition of Owari Province must have been literate, it is unclear how these district officials and local notables learned to write literary prose, as they were hired among the local population and were not appointed among Heian or provincial officials. In addition, the only official school (kokugaku) was at the provincial government.48 However, a stylistic analysis of the petition will give us the key to identifying the authors. The entries in the Owari report were written in an administrative kanbun style, which while suggesting some literary training indicates that its authors lacked the experience of Heian court officials, who were trained at the Bureau of Higher Learning (Daigakuryō). The authors follow the Chinese custom of writing parallel sentences but only seldom quote from the Chinese classics (as in entry 3), and some Chinese quotations are incomplete. In addition, certain expressions such as chinmai in the twenty-third entry, kakato in the seventh entry, and hokumin in the third entry, to name a few, do not correspond to the conventions of the standard language. Often the contents do not correspond to the title. For example, the title of the tenth entry is food aid to travelers, but it deals with the destitute and homeless, and the title of the eleventh suggests tenants rice fields (kubunden) but more accurately concerns assigned rice fields (ryōden). The petition includes many repetitions such as the price of taffeta silk, which is repeated in the fifth, seventh, and ninth entries. As for the administrative language, many terms such as tax products (kanmotsu in entries 17 and 23) adhere to conventional usage. However, the terms “purchased tax products”(kōeki zōmotsu or kōeki) and “specialties” (domō) in the third, fifth, seventh, and ninth entries do not correspond to usage commonly found in Heian-period documents. Originally domō meant “district specialties” and were part of purchased tax products, but in the Owari petitions, domō is used as a general term for silk or rice seized by tax commissioners. Furthermore, “purchased tax products” are not mentioned in administrative documents after the eighth century except in the Engi shiki proceedings.49 In the twenty-eighth entry, the authors also provide an unusual defi-
322 | charlotte von verschuer
nition for zōshikinin as being deliverymen, although they were normally known as office clerks in the court’s nomenclature. In short, these features illustrate the various inaccuracies in the language of the petition. Another issue centers on the prominence of the term hyakusei, which in this context refers to local elites, managers, or notables.50 Yet in some entries, the term is used in its traditional meaning of peasants or commoners (entry 3), or peasant households, hyakusei no kamado (entries 23 and 24), which otherwise, as in the case of the twenty-eighth entry, were referred to as “people’s households” (jinmin no kamado). In general, it is common for the words jinmin, min/tami, and shomin to refer to people or commoners (entries 7, 9, 15, 17, 21, and 23). Hence a closer examination of the language of the petition reveals many inconsistencies, suggesting that it was not written by an official of the Heian court or by a person of the nobility as these individuals would have mastered Japanese terminology and the Chinese classics and style of writing. In light of these various issues, the Petition of Owari Province cannot be assessed as a well-written literary document (meibun), as has sometimes been claimed. It has also been valued as a pedagogic model by some authors because it is quoted in various court manuals such as the Hokuzanshō (collection from the Northern Hills; first half of the eleventh century) and has often been copied and possibly used as a reference by estate (shōen) managers in later centuries.51 However, in my opinion, this was more likely due to the structure and composition of the petition rather than to its literary style. The thirty-one entries follow a methodological order: different kinds of taxes (entries 1 – 5), purchased tax products (6 – 9), transport (10 – 13 and 19), administration and management (14 – 18), financial support and burdens of peasants and monks (20 – 25), ill treatment by the retainers (26 – 30), and the handling of court decrees (31). Each entry mentions the subject of the complaint in the title, states in what way matters are contrary to rules and precedents, provides concrete examples or figures, stipulates how matters should be and how they contrast with the present governor’s attitude (describing his misdeeds often in very ornate language), and ends with a moving picture of the local people’s grief and suffering, thereby hoping to elicit compassion among Heian court administrators. Each entry finishes with the same request — the withdrawal of the incumbent provincial governor. As such, the general composition of the report might have been considered as a useful model for writing petitions by later generations, especially because it was in the end successful in removing the governor. A closer examination of the information from the petitions provides further evidence that its authors were not aristocrats from the capital. Let us examine this assertion in light of the concrete figures given in the entries. The eleventh entry refers to the acreage of the rice fields affected and to the management of the post stations and the district relays. Although the number of the stations and relays and the calculations in the entry are accurate, the acreage of the rice fields given in the entry is higher than the standards set by the Taihō Code. The same is true for the above-mentioned figures of the transport burdens: they slightly exceed the regulations. According to the seventh entry the price paid by the
Life of Commoners in the Provinces | 323
governor for one bolt of taffeta silk was 40 or 50 sheaves of rice. The price was, however, 50 or 60 sheaves in the neighboring provinces (entry 9), which corresponds to the official price fixed for the silk of Ise Province, a neighbor province of Owari. In many other provinces, including those of Kyushu, the price was 80 sheaves for a bolt of silk, and in 989 the Kanzeonji temple in Chikuzen Province paid 100 sheaves. Compared to these figures, the price paid by the governor of Owari was lower than the standard prices. For Shinano hemp, the governor paid 5 or 6 sheaves of rice, which was less than the regulated price of 8 sheaves. The petitioners mention thousands of bolts of silk, counting 1,212 bolts in three years in the ninth entry, or 404 bolts a year. This exceeds the fixed amount of silk to normally be supplied to the Heian court by Owari Province every year and is therefore suggestive of the governor’s abuse of the tax system.52 All these figures are concrete and detailed. It seems unlikely that a court official in Heian would have such local details at hand; and, if the petition were forged, the complaints would likely have been couched in more general terms. The concrete figures thus indicate that the petition was written by somebody who knew the living conditions in the province. Other figures in thePetition of Owari Province deal with the various taxes payable in rice, provincial expenses for the maintenance of the irrigation systems, repair of provincial temples, stipends for provincial monks, allowances drawn from the rice fields of the province officials, amount of hulled rice, and prices of horses. It is worth noting that some data concern the districts and other data the provincial government. This leads to the question whether the authors are only district elites (as they claim for themselves) or if they include some province officials. There are other examples of such incongruities. For instance, several entries mention the provinces’ expenditures for three years but no single district’s expenditures. The thirty-first entry states that the government of Owari received nine decrees from the Heian court but failed to bring six of them to the attention of the districts. How could district authorities know about the documents received by the governor without obtaining this information from members of the provincial office? This disjunction suggests that the petition was authored by district authorities in close collaboration with some province officials. In fact, two entries give us a clue for identifying them. According to the twenty-first entry, the province clerks (shosei) suffered from being deprived of their daily food allowances, and according to the twentieth entry the province assistants and secretaries did not even receive their salaries. They were low-ranking province officials and employees who worked under the governor and the vice-governor.53 It seems understandable that these province employees agreed to support the district elite’s complaints. Interestingly enough, the members of the province government are mentioned in the text as low-ranking officials (sanbu ika honkan ijō) instead of being identified by their titles such as assistant (jō), junior assistant (sakan), and secretary (shishō) possibly to deflect attention from potential opponents. These assistants and secretaries probably had the opportunity to receive some training at the provincial school and gain at least a basic knowledge of Chinese classics and
324 | charlotte von verschuer
Japanese terminology. They might have held the brush when writing the petition while agreeing on its contents with the district elites. This would also explain both the somewhat unsophisticated style of the text and its contents providing figures for the province while also describing the situation in the districts. Such an interpretation concurs with Katō Tomoyasu’s treatment of lowranking province employees. Katō argues that they were recruited among district authorities and had the same social background, which would explain the alliance between the two.54 Katō also acknowledges three leading groups of provincial society in tenth-century Japan: province officials, district officials, and local elites. This once again points to the identity of the petitioners of Owari. They included not only district authorities and low-ranking province officials, but also local elites, called “local notables” (hyakusei) or “vagrants” (rōnin). The local elites belonged to a new class of people that emerged in the tenth century from among vagrants and other peasants who succeeded in becoming small proprietors and who sought alliances with local officials. Yet this class of people remained silent throughout the Heian period, as they were basically illiterate. We know them indirectly through the brush of provincial officials. The Petition of Owari Province serves as a unique medium for the voice of local notables, offering a window into local society in Owari Province. The petitions chosen for examination in this chapter also provide important insight into the interaction between the province and the capital, between a periphery and the political center of Heian Japan.
Between the Capital and the Provinces Broadly speaking, we can distinguish between two classes in Heian society: the court aristocracy and the population of commoners including peasants and craftsmen. The aristocracy lived in the capital and its surroundings, and its members were temporarily transferred to the provinces as governors and vicegovernors. After the tenth century, many governors completed their terms of office in the capital, opting to relegate their responsibilities to local officials. It might be a special feature of the Heian period that its elites were concentrated in the capital. In later periods, there were more elites, who came to be more dispersed throughout the country. Geographically, Owari Province was situated at the periphery of the central region but not at the far periphery of the realm, which was represented in particular by Mutsu and Dewa in the east and by the outlying islands to the west. Through an examination of the Petition of Owari Province of 988, we have been able to examine the communication, taxation, and transportation system between the province and the capital. The communication system consisted of a network of highways and post stations throughout the country. This infrastructure served as a conduit between the provinces and the capital, providing the means for central control and collection of information. Yet the burden of maintaining this network and providing transportation convoys rested with the provinces. In fact, the court depended on the provinces’ management of this network. As for taxation and material goods, the examples from Owari Prov-
Life of Commoners in the Provinces | 325
ince indicate that the imperial court was heavily dependent upon the provinces for various supplies to sustain both its official duties and the material daily life of the imperial household and the court officials. In terms of practical life, the relationship between the capital and the provinces can be summarized in the following way: the court issued orders and admonishments to the provinces, and the provinces provided information and material supplies to the court. Economically, the provinces could live in autarky, but the court could not survive without contributions from the provinces.
Notes 1. Owari no kuni gebumi is cited in essays by Cornelius Kiley (“Provincial Administration and Land Tenure in Early Heian Japan”) and Dana Morris (“Land and Society”) in Donald Shively and William McCullough, eds., The Cambridge History of Japan, vol. 2, The Heian Period, in their general discussion of the land tenure system and the tax system of the state’s rice economy. By contrast, this essay focuses on local craft and agriculture from the point of view of the taxpayers at the village and district levels. I am grateful to Simone Mauclaire and Detlev Taranczewski for their constructive comments in the process of writing this article. In this article I used the following editions of Owari no kuni gunji hyakushōra gebumi: Abe Takeshi, Owari no kuni gebumi no kenkyū; Owari no kuni gunji hyakushōra no ge in NST; Owari no kuni gunji hyakushōra no ge in Heian ibun; Shinshū Inazawa-shi shi: Shiryō hen 3; and Owari no kuni gunji hyakushōra ge gebumi (facsimile of the Tōdaibon manuscript). The text of my translation of the entries ii, 3, 7, and 23 can be found in each edition under the same entry number. 2. Early postwar scholars in Japan assumed that hyakushō refers to commoner-peasants (see, e.g., Kiley, “Provincial Administration,” 167) as in modern Japanese, and therefore named the document under scrutiny here Owari no kuni gunji hyakushōra no gebumi. Although I have chosen to maintain this title, I concur with the more prevalent opinion today that in the ninth to twelfth centuries it refers either to commoners or to peasants or to local notables, depending on the context, and that the contemporary reading was hyakusei. 3. Katō Tomoyasu, “Fujiwara no Motonaga wa naze uttaerareta ka,” 140 – 151; Watanabe Shigeru, “Owari no kuni gebumi kyōju no shokeitai,” 73 – 81; Miki Masahiro, “Kyōkunsho Chūmonsho no sekai, ge,” 21 – 55. See here after the section “The Petitioners and Life in Owari Province.” 4. From the Sōbon manuscript (printed in Nihon shisō taikei, vol. 8) only entries 1 – 14 and 31 are extant; and of the Tōdaibon manuscript (on the facsimile of the Shiryōhensanjo) entries 1 – 4 are missing. The Shinbon copy has been printed in Heian ibun (HI), vol. 2, and in Abe Takeshi, Owari no kuni gebumi no kenkyū. A thorough study of all three manuscripts including printed versions and a photograph of the Shinbon manuscript can be found in Shinshū Inazawashishi. 5. On the transportation system, see Tanaami Hiroshi, Kodai no kōtsū (Tokyo: Yoshikawa kōbunkan, 1969); Kodama Kota, Nihon kōtsūshi (Tokyo: Yoshikawa kōbunkan, 1992; rpt. 1994). On the envoys, see Francine Hérail, Yodo no tsukai, and Seiji Yōryaku, ch. 56, p. 404 – 414. For lists of envoys and registers, see Dai Nihon komonjo (DNK) 1:597; and Ruijū sandai kyaku, ch. 18, p. 584. 6. Ishikawa ken maizō bunkazai sentā and Hirakawa Minami, eds., Hakken!, 4 – 5, 22 – 26. 7. On the system and management of stations, see Ritsuryō, Denryō 33, p. 247, Kyūmokuryō
326 | charlotte von verschuer 14 – 22, pp. 416 – 419, Kushikiryō 42 – 43, pp. 392 – 393, Buyakuryō 19, p. 256; Ruijū sandai kyaku, chap. 18, pp. 579 – 584. All subsequent references to the Taihō Code of 701 are derived from the Nihon shisō taikei version of the Yōrō Code of 718. 8. Engi shiki, ch. 26, pp. 665 – 668, ch. 28, pp 711 – 717, ch. 48, pp. 973 – 981. 9. Charlotte von Verschuer, Les relations officielles du Japon avec la Chine aux VIIIe et IXe siècles, 319 – 323. 10. Engi shiki, ch. 24, p. 619. 11. Shoku Nihongi, Jingi 3 (726) 8/30; Engi shiki, ch. 23, p. 583, ch. 11, p. 327. 12. Shoku Nihongi, Jingi 3 (729) 8/30; Engi shiki, ch. 11, p. 327, ch. 28, p. 711, ch. 24, p. 604. 13. The ritsuryō codes stipulated a total of 12 chō of post station rice fields for medium-size provinces such as Owari (Ritsuryō, Denryō 33, p. 247), but Owari later became an importantsize province (Engi shiki, ch. 22, p. 560). 14. Ritsuryō, Kyūmokuryō,1 – 28, pp. 413 – 420; Ruijū sandai kyaku, ch. 18, pp. 575 – 579 (decree of 862); Engi shiki, ch. 48, pp. 973 – 983 (sa-u meryō), ch. 28, p. 708; Francine Hérail, Fonctions et fonctionnaires, 507 – 521. 15. For the taxes, see Sonoda Kōyū, Nihon kodai zaisei shi no kenkyū; Dana Morris, “Land and Society,” 183 – 235. I will discuss only the land tax (hereafter) and the per capita tax in this essay. Apart from these two there were the corvée or labor services. They included sixty days (later reduced to thirty days) of service to the provincial government each year for every male adult, as well as three years at the capital for two adult males from each administrative village, and one to three years (not full time but in shifts of ten days) of military service as guards at the capital or in the provincial militias for one out of every three to four adult males of twenty-one years or more. 16. Ritsuryō, Denryō 1 – 3, pp. 240, 570 (tax in sheaves sotō); Engi shiki, ch. 23, p. 585, ch. 26, p. 654 (tax in grains sokoku, and the yield of 15 – 50 sheaves). For measurements of the eighth century, see Iwanami Nihon shi jiten, 1431. 1 chō = 1.13 hectares or 2.79 acres; 1 tan = 11.3 ares or 0.279 acres; 1 koku = 85 liters. 17. Charlotte von Verschuer, Le riz dans la culture de Heian, 267, 268. 18. DNK 1:607, 608, 613; Engi shiki, ch. 23, pp. 584 – 585; Sonoda, Nihon kodai zaisei shi no kenkyū, 96, 100 – 104. 19. Seiji yōryaku, ch. 53, pp. 308 – 311; Kōnin shiki, 18, 19; Engi shiki, ch. 26, p. 654. 20. Discrepancies between the extant manuscripts. For sokoku, see note 16. On the third entry, see Miyahara Takeo, “Owari no kuni gebumi no sozei to densei,” pp. 32 – 34. 21. The Shinbon manuscript states “less than 2 to” while the Sōbon manuscript indicates “more than 2 to.” See notes 4, 20, and HI 2:474. 22. Ritsuryō, Buyakuryō 9, pp. 253 – 254; Nishiyama Ryōhei, “Ritsuryō sei shakai no hen’yō.” 23. See Verschuer, Le riz dans la culture de Heian, 41 – 47, for details regarding the ricegrowing process. 24. Ritsuryō, Buyakuryō 1, 4, pp. 249 – 252, Shikiinryō 40, p. 178; Engi shiki, ch. 24, pp. 597 – 622. 25. Ritsuryō, Buyakuryō 7, 29, 35, pp. 253, 259, 260, Shikiinryō 7, 33, pp. 163, 175; Engi shiki, ch. 23, pp. 590 – 594, ch. 39, pp. 875 – 877. 26. Engi shiki, ch. 37, pp. 828 – 843, ch. 23, pp.586 – 590, ch. 33, p. 779. 27. Nishiyama Ryōhei, “Ritsuryō sei shakai no hen’yō.” For a comprehensive summary of the historical development of tax management, see Kiley, “Provincial Administration.” 28. Lacquer is listed in only one of the manuscripts. 29. Listed in only one manuscript. 30. Quantities differ between the manuscripts. Four or 5 to correspond to 0.4 or 0.5 koku.
Life of Commoners in the Provinces | 327 31. Engi shiki, ch. 23, p. 591. For other tax products of Owari, see ibid, ch. 23, pp. 584 – 586, 590, 591, ch. 24, p. 604. 32. Seiji yōryaku, ch. 57, p. 434, Engi 11 (901) 2/25; Nihon kiryaku, Tenryaku 1 (947) 11/11; Gonki, Chōhō 4 (1002) 2/19. Motonaga’s cheating with silk has generated many interpretations; see Miyahara Takeo, “Owari no kuni gebumi no chōyō to kōeki,” 11 – 17. Miyahara points out that the “several thousands of bolts of silk” seized by Motonaga probably comprises other taxes in kind here called domō. 33. Ruijū sandai kyaku (ch. 8, pp. 333 – 334, Jōgan 13 [871] 8/10) mentions abusive business with purchased tax products by vagrants in Kyushu. 34. Engi shiki, ch. 23, pp. 584 – 585 (nenryō shōmai); Hérail, Fonctions et fonctionnaires, 394 – 398; Shunki, Chōryaku 2 (1038) 10/8, in idem, Notes journalières de Fujiwara no Suke fusa, 1:173. 35. Engi shiki, ch. 23, pp. 591 – 594,ch. 24, p. 601, for taxes; ch. 14, pp. 396 – 397, ch. 15, pp. 423 – 424, for imperial clothes; ch. 46, p. 964, for other clothes; ch. 39, p. 871, for sieves; Francine Hérail, Notes journalières de Fujiwara no Michinaga, 3:760, for distributions; Engi shiki, ch. 26, p. 663, for barter rates. See also Heian jidai shi jiten, 628. 36. Engi shiki, ch. 30, pp. 734 – 735, ch. 17, pp. 457 – 459; Ritsuryō, Rokuryō 1, pp. 304 – 305; Shunki, Chōryaku 2 (1038) 10/11, in Hérail, Notes journalières de Fujiwara no Sukefusa, 1:175; Heian jidai shi jiten, 1128 (for Michinaga). 37. Engi shiki, ch. 23, p. 590, ch. 39, pp. 868, 872, ch. 11, p. 893; Charlotte von Verschuer, “L’habitat rural du Japon ancien,” 39, 40. 38. Engi shiki, ch. 17, pp. 448 – 463, ch. 30, p. 733. 39. Engi shiki, ch. 23, p. 591, ch. 24, p. 604, ch. 36, pp. 809, 810, 813; Verschuer, Le riz dans la culture de Heian, 234 – 237. 40. Engi shiki, ch. 14, pp. 401 – 404, ch. 15, pp. 423 – 444; Verschuer, Le riz dans la cul ture de Heian, 234 – 237, and the Annex, “Inventaire des plantes comestibles et industrielles,” 345 – 359. 41. Tanaami, Kodai no kōtsū; Kodama, Nihon kōtsū shi, with the mokkan text on 55; Ritsuryō, Buyakuryō 3, p. 251. 42. DNK 2:65. The calculation of 530 men and the estimations are my own. 43. Ruijū sandai kyaku, ch. 6, pp. 275, 276, ch.14, p. 396, ch.19, pp. 623, 624; Engi shiki, ch. 22, p.568, ch. 23, p. 582, ch. 26, pp. 666 – 668. For transport by horses or carts in the eighth century, see DNK 2:65, 155, 171 – 173, 557 – 560. 44. Engi shiki, ch. 24, p. 604, for the number of traveling days, idem, ch. 22, pp. 567, 568, ch. 23, pp. 582, 594, and Ruijū sandai kyaku, ch.19, p. 624, for delays, losses, and punishments; Engi shiki, ch. 22, p. 560, ch. 24, p. 604, for distances; idem, ch. 23, p. 585, for rice storage; idem, ch. 20, p. 568, and Hérail, Yodo no tsukai, 91 – 95, for procedures in the capital. 45. Shunki, Chōryaku 2 (1038) 10/8, in Hérail, Notes journalières de Fujiwara no Sukefusa, 1:173; Terauchi Hiroshi, Zuryō sei no kenkyū, 83 – 110. 46. Charlotte von Verschuer, “De la houe à la sandale chevaline dans le Japon ancien,” 437, 443. 47. Engi shiki, ch. 50, p.995, ch. 26, p. 668; DNK 2:65. 48. According to Momo Hiroyuki (Jōdai gakusei no kenkyū, 425), the provincial schools served to educate province officials but sometimes also trained district officials in the eighth century. Heian ibun, 9:3522, even mentions a district school (gakkō) in Gunma-gun, Kōzuke Province, in 1028. 49. The Heian ibun gives domō in seven documents until the twelfth century with the meaning of rice payments, and gives kōeki zōmotsu only in two eighth-century documents of Owari. But the purchase (kōeki) of various tax products (zōmotsu) was practiced in the ninth century.
328 | charlotte von verschuer See Ruijū sandai kyaku, ch. 8, pp. 334, 336, Jōwa 10 (843) 3/15, Jōgan 13 (871) 8/10, and other occurrences. 50. Kimura Shigemitsu, “Chūsei hyakusei no seiritsu,” analyzes the change of meaning of hyakusei, tato, fumyō, and shujin from the tenth to twelfth centuries. 51. Kokushi daijiten, 2:961; Watanabe Shigeru, “Owari no kuni gebumi kyōju no shokeitai,” 76; Katō, “Fujiwara no Motonaga wa naze uttaeta ka,” 148 – 150; Miki, “Kyōkunsho Chūmonsho no sekai, ge,” 21 – 35. 52. Engi shiki, ch. 28, p. 711, for the stations; ch. 26, p. 663, and HI 1:335 (Dazaifu chō an of 989), for the prices of silk in the provinces. Engi shiki, ch.23, p. 591, prescribes 150 bolts as purchased tax (see above, note 32). Engi shiki, ch. 26, p. 663, for the price of Shinano hemp. 53. See the province diagram in Ritsuryō, Shikiinryō 70 – 73, pp. 192 – 194. 54. Katō, “Fujiwara no Motonaga wa naze uttaetaka.” See also Yamaguchi, “Jū seiki no kokugun gyōsei kikō”; and Hirano Takemi, “Owari no kuni gunji hyakushōra gebumi ni tōjō suru nin’yō kokushi ni tsuite.”
13
d Karl Friday
Lordship Interdicted Taira no Tadatsune and the Limited Horizons of Warrior Ambition
I
n 1028 — the very middle of the self-professed era of “peaceful tranquility” — the central aristocracy’s self-complacence was ripped by reports that Taira no Tadatsune (967 – 1031), a maternal grandson of the infamous Taira no Masakado (? – 940), had attacked and ravaged the provincial government compound (kokuga) in Awa. This incident, and the events that followed, rank among the most dramatic episodes in the early history of Japan’s warrior order. Masakado’s insurrection, some seven decades earlier, had climaxed with the protagonist’s claiming for himself the title New Emperor. Tadatsune’s reach did not extend so far, but his grasp held the provinces of the Bōsō peninsula — Kazusa, Shimōsa, and Awa — for the better part of three years, and left much of the region in ruin (map 13.1). His timing was also intriguing, coinciding with both the apex of the Fujiwara regency in the capital and with what some historians have identified as an era of transition between old and new forms of property holding in the countryside.1 It is small wonder, then, that historians have long cited the “Taira no Tada tsune Rebellion” as a signal moment in Heian history. For the most part, it has been presented as an early — albeit incongruously unsuccessful — harbinger of things to come, a portent of the Genpei War of the 1180s, the Jōkyū War of 1221, and the Nanbokuchō wars of the late fourteenth century that, step by step, ushered in the medieval era of localism and warrior rule. Closer analysis indicates, however, that Tadatsune’s career and circumstances were more typical of his class, and his fate more emblematic of the structure and inherent stability of the Heian polity, than the drama surrounding his insurrection might suggest.2 The following study examines the events of 1028 – 1031 in detail, with a special eye toward the light they shed on the relationship between the capital and the countryside that held in the mid-Heian period.
Warrior Society in the East Life in the eastern provinces during the mid-Heian period was increasingly dominated by a competition for wealth and influence between multiple groups, including provincial-resident elites, provincial officials, and the temples, princes, and officials (shoin, shogū, ōshinke) of the court.3 At the axis of this competition were
330 | karl friday
Map 13.1. Bōsō peninsula
the middle-ranked court nobles whose careers centered on appointments to provincial government offices. To ensure a continued succession of posts, such career provincial officials (zuryō) forged alliances with the lofty aristocrats (kugyō) above them. At the same time, many zuryō found that they could use the power and perquisites of their offices, and the strength of their court connections, to establish landed bases in their provinces of appointment and to continue to exploit the resources of these provinces even after their terms of office expired. Central government complaints of problems caused in the hinterlands by “officials who have finished their terms of office and by sons and younger brothers of princes and court officials” appear as early as 797.4 Such figures, we are told, were “settling down in their [former] areas of jurisdiction, where they hindered agriculture,
Lordship Interdicted | 331
gathered up the peasantry like fishermen harvesting fish, and constructed plans for their own evil gains.”5 This trend was both widespread and enduring. The practice of nobles “settling down” in the provinces is a well-known phenomenon of the Heian period, but expressed in these terms the concept is a bit misleading, for few central aristocrats actually abandoned life in the capital for a provincial existence. Typical exurban provincial officials concluded marriage and other alliances with local figures and held packages of lands scattered about the countryside, which provided them with income. At the same time, they maintained extensive contact with political affairs in the capital and often kept homes there. Some individuals and branches of families became more thoroughly committed to a provincial existence than others, but most were still careful to keep their ties to the capital alive. They could not afford to do otherwise, for to cut oneself off completely from the court was to sever oneself from the source of official appointments and personal connections — and thereby end all hope of maintaining one’s social and political position. To provincial governors and their families, Heian-kyō was the source of the human and physical resources that made their provincial business activities possible, as well as the marketplace for the goods they brought from the country.6 Many provincial officials, compelled by a need to defend themselves and their prerogatives against outlawry and armed tax resistance, as well as by the desire to expand their income by all means fair or foul, included “warriors of ability” among the personal entourages that accompanied them to their provinces of appointment. A significant number of zuryō also cultivated military skills themselves.7 Fighting ability was useful in winning the respect of, or intimidating, armed residents of their provinces, and it could also speed the advance of an official’s career by opening doors to posts in court military units or to the private service of important court figures. By the tenth century, military service at court and service as a provincial official had become parallel and mutually supportive careers for many zuryō, resulting in the emergence of the group Japanese scholars have dubbed the miyako no musha, or “warriors of the capital”: men of the fourth or fifth court rank who curried the patronage of the higher nobility, and recognition by the state, by serving as bodyguards, police, and soldiers.8 Provincial warrior leaders were, broadly speaking, of two main types of pedigree: descendants of cadet branches of miyako no musha houses — the Seiwa Genji, the Kanmu Heishi,