Puppets of Nostalgia: The Life, Death, and Rebirth of the Japanese Awaji Ningyō Tradition 069102894X, 9780691028941

Citation preview

PUPPETS OF NOSTALGIA

P U P P E T S OF N O S T A L G IA

T H E LIFE, DEATH, AN D R E B I R T H OF T H E JA PANESE AW AJI NINGYO TRADITION

Jane Marie Law

P R IN C E T O N

U N IV ER SIT Y PRESS

PRIN CE TO N , NEW JERSEY

Copyright © 1997 by Princeton University Press Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, Chichester, West Sussex All Rights Reserved

Library of Congress Cataloging-in-Publication Data Law, Jane Marie. Puppets of nostalgia: the life, death, and rebirth of the Japanese Awaji ningyo tradition / Jane Marie Law. Includes bibliographical references and indexes. ISBN 0-691-02894-X (alk. paper) 1. Puppet thcater-Japan-Awaji Island. 2. Performing artsReligious aspects. 3. Japan-Religious life and customs. 1. Title. PN1978.J3L39 1997 791.5'3'09521 87-dc20 96-33487 Publication of this book has been aided by a subvention from the Hull Memorial Publication Fund Committee This book has been composed in Galliard Princeton University Press books are printed on acid-free paper and meet the guidelines for permanence and durability of the Committee on Production Guidelines for Book Longevity of the Council on Library Resources Printed in the United States of America by Pnnceton Academic Press 10

9

8

7

6

5

4

3

2

1

For Adam, Samuel, and Tamar

with love

If we should laugh at and insult the memory of the Puppet, we should be laughing at the fall that we have brought about in ourselves, laughing at the Beliefs and Images we have broken. -E. Gordon Craig (1912)

C ontents

Acknowledgm ents

xi

I n tr o d u c ti o n . O f S tories and F rag m en ts

3

O ne In th e Shape o f a P erson: T h e Varieties o f R itual U ses o f Effigy in Japan

17

Two K adozuke: T h e O u tsid e r a t th e G ates

49

T h re e A C rippled D eity, a P riest, an d a P u p p et: K u g u tsu and E bisu-kaki o f th e N ishinom iya Shrine

89

Four A D ead P riest, an A n g ry D eity, a F isherm an, and a P uppet: T h e N arrative O rigins o f Awaji N ingyd

137

F ive P u p p ets o f th e R o a d , P u p p e ts o f th e Field: Shiki Sanbaso, E b isu -m a i, and P u p p e try Festivals o n Awaji

164

Six P u p p ets and W hirlpools: Ico n s, N ostalgia, R egionalism , and Id e n tity in th e Revival o f Awaji N ingyo

204

E p ilo g u e

265

Notes

267

Bibliography

301

Index

313

Acknowledgments

first tim e I walked into the Awaji Puppet T heater in Fukura in 1978 until this final m anuscript was prepared for publication, I have been indebted to m any agencies, teachers, and friends in b o th Japan and the U nited States for support, guidance, and assistance. O n Awaji, my first thanks go to Fujim oto Yasurd and his family— Seiko, FIatsuho and M izuho— for being my hom e during many extended peri­ ods o f fieldwork. T he people o f their neighborhood o f N akanochd in Fukura all became a special p art o f this project, and I will never forget their kindness and interest in my work. At the theater, the Awaji Ningyd Joruri Kan, the puppeteers, chanters, and musicians were m ost generous with their tim e and knowledge. Naniwa Ayako and M atsuyama M itsuyo took care o f endless details w ith equally endless good spirits. Nakanishi H ideo, director o f the high school puppetry club at M ihara Fligh School and a scholar o f Awaji puppetry, was always helpful and patient, and I acknow l­ edge his m any contributions to this work. F udo Satoshi, an avid A w n ji ningyd enthusiast, grandson o f a puppeteer and son o f the first “revivalist” o f the tradition, was always available to drive me to new locations and discuss ideas. Fie also generously gave me perm ission to use rare photos from his father’s collection in this book. M ori M asaru, president o f the Awaji Puppetry Preservation Society, was m ost generous w ith financial support during m any stages o f my research. Tam ura Shoji o f the Awaji Historical Archives in Sum oto was always gracious in giving m e access to rare docum ents. Two very special people m erit a separate accolade. U m azum e M asaru, director o f the Awaji N ingyd Joruri Kan, and Naniwa Kunie, its resident grandm other and spirit, gave me m ore inspiration and assistance than I can ever acknowledge. I d o u b t I shall ever again have such enchanting experiences in field research as I had w orking w ith these two fine people. For assistance in setting up my research trips to U sa, I am grateful to the Shingon priest Iwatsubo-sensei on Awaji and the A ndd family in Usa at D airakuji. Kabata Yoshimizu o f the O ita Prefecture Buraku H istory Study G roup gave me an excellent historical and them atic overview o f the issues relating to puppetry and status in the O ita area. Research for this project was funded by different agencies. From 1 9 8 7 89, I was a M om busho scholar w ith affiliation at Tsukuba University in the Division o f Philosophy. D uring my stays at Tsukuba, Professor Araki F ro m t h e

M ichio and Miyata N oboru were m ost supportive. Also at Tsukuba, I was assisted with many aspects o f this research by Komatsu Kayoko, Yoshimine Kazu, and M iyamoto Ybtaro. After I came to Cornell in 1989, I received summer subsidies from the Cornell East Asia Program to con­ duct the research for the final chapter o f this work. Many teachers at the University o f Chicago had a major influence on this project. Mircea Eliade’s enthusiasm for the religious dimensions of puppets and his interest in my research helped form the early stages o f my work. Lawrence Sullivan and David Tracy were both careful and critical readers, and I acknowledge their assistance and support. My deepest gratitude for my teachers goes to Professor Joseph M itsuo Kitagawa. H e was a m odel o f hum ane academic integrity, and I can never express all he has done for my educational developm ent. I miss him greatly, and although my research has gone in different directions, his voice informs all my reflections on the history o f religions, rem inding me to look for the subtle and gentle gestures o f people. At Cornell, many colleagues and friends gave me m uch help. John McRae read parts o f this work and offered helpful criticism and advice. Kyoko Selden’s careful attention deserves a special note o f thanks, and I am deeply grateful for her endless support and assistance over the years. Leslie Peirce, Rachel Weil, and Chris Franquem ont read early drafts o f chapters and helped me find my own voice in this writing process. I am particularly grateful to Chris for reading through the entire penultim ate draft and making many helpful stylistic suggestions. Diane Kubarek o f Cornell Inform ation Technologies was especially helpful in showing me how to digitize the drawings included in this book. Preparing a m anuscript for publication and being a working m other with two tiny children requires more than the ability to do several things at once. One needs special friends who can step in and help out. To Karen Davison, U em ura Yasuko, and Peng Chenna, I extend gratitude for kind­ ness and assistance I can never repay in full. Ivy Mauser, Kay Reynolds, Amelia Massi, Louise Raimando, Rebecca Schwed, and Bettina Ehrm ann were a great support to me as I juggled my several roles. I owe a special thank you to my copyeditor, John LeRoy, who liberated many o f my cumbersome sentences from the clenched fists o f verbosity. At the Princeton University Press, I thank H elen H su for her careful at­ tention to endless details, and Ann Wald, my editor, for her support and clarity. Janet Snoyer prepared the index. My father-in-law, Ron Law, read chapters o f the book, and made many helpful stylistic suggestions. My own parents, Alfred and Delores Swanberg, were also a great support along the way. My husband, Adam, and my two children, Sam and Tamar, have earned the dedication o f this book

m any tim es over. I could n o t have w ritten this b o o k w ith o u t th eir su p ­ p o rt, and I ow e th em my love and gratitu d e. W hile all o f these people m ade m any co n trib u tio n s to this w ork, it goes w ith o u t saying th a t any errors are my ow n. I have given surnam es first in Japanese, unless referring to w orks by Japanese au th o rs published first in E nglish. Any p h o to g ra p h s in th e w ork, unless otherw ise n o te d , are m y ow n.

PUPPETS OF NOSTALGIA

HONSHU

N IS H IN J

OSAKA BAY

AWAJI

TOKUSHII

SHIKOKU PACIFIC OCBAN

Map o f Awaji as situated in Japan.

AWAJI IN CONTEXT

Introduction O f Stories and Fragments

Place and Time: The M an w ith the Pictures The m an was in his sixties, going bald, with bits o f gray hair around his temples. H e sat on the bench outside the Awaji puppet theater, on the prom enade overlooking the swirling N aruto Straits. O n the other side o f the enorm ous whirlpools, the island o f Shikoku could be seen through the bright sunshine. It was late afternoon, and the direction o f the whirlpools indicated the water o f the Pacific Ocean was flowing into the Inland Sea o f Japan. In several hours, it would flow back o u t once again. T he cape at the southern tip o f the small island o f Awaji, w here we sat, form ed one side o f the opening o f the Setonctikai. O n his lap, the m an held a pile o f old black and white photographs, a tattered and yellowed handw ritten m anuscript, and a collection o f papers. These were docum ents collected by his father in the 1940s w hen he had tried to save the tradition o f Awaji puppetry from extinction. This man had come to the theater to show me the docum ents, and as he did so, he told me the following story: “My grandfather [on my father’s side] was a Ddkumbo mawashi—he m anipulated sacred puppets at N ew Year, weddings, and sometimes at other small festivals. O f course, this wasn’t his only occupation. We were poor, and so he opened up a little store. You could call it a mini superm ar­ ket, I guess. “I was about ten w hen the war started. I rem em ber that tim e clearly. Like so many other people in Japan, we were soon starving as Japan m oved further into the war. N o one had any money to spend in the store. Things were really tough. So one day, to feed the family, G randfather decided to sell his puppets. N o sooner had he made the decision, it seemed, than several m en from Osaka showed up in the courtyard to look them over. His puppets were really beautiful. H e had Sanbaso, Okina, Senzai, Ebisu, all the little flutes and drum s and hand rattles—the whole bit. H e even had nice little costum es for all o f them —in pretty good shape, too. “The m en from Osaka paid my grandfather a small am ount o f money, packed the puppets into the car, and drove away.

“ I can see that day like it happened yesterday. I was watching the trans­ action from a little bit behind my grandfather. H e stood in the courtyard and looked down the road until long after the car with the men in it had rounded the corner and the dust it raised on the road had settled. H e just kept standing there and standing there. In my mind, I rem em ber him standing there all day, until the sun w ent down. “Then he came inside, and I tell you, he never said another w ord about those puppets. They were never m entioned again. “I was only a kid, but I felt really sad. I didn’t know what made me sad then. But thinking back on it now, I guess I felt like a part o f Awaji had been packed into the back o f a car and stolen away to Osaka. O r maybe I felt som ething had died. I don’t know what I felt, really. “Many years later, after my grandfather died, we were going through his things, and we found all these little sketches he had made o f Sanbaso. They were quite detailed— they showed the puppet’s eyes, his cos­ tum e, how his m outh worked, his hands and feet. They were just lovely. Do you suppose he was trying to rem em ber the puppets he had sold away to the collectors from Osaka? I d o n ’t know, but I still have the sketches.” This story was told to me in July 1988, while I was conducting fieldwork on the island o f Awaji concerning the history o f A w aji ningyo shibai (the Awaji puppetry tradition). Far from being a puppet theater for children, Awaji puppeteers perform ed rituals that maintained rigid ritual purity codes and kept dangerous spiritual forces properly channeled and appeased. O ut o f these ritual roots, a rich dramatic joruri (ballad) tradi­ tion developed.1 This form o f puppet theater later became the classical Bunraku for which Japan is famous. This book, based primarily on my fieldwork on Awaji in the village o f Fukura in 1984, from 1987 through 1989, in the summers o f 1990 and 1991, and in the fall o f 1993, is a study o f the dynamics o f tradition form a­ tion and revival in Awaji ningyo shibai. M ost studies o f Japanese puppetry by Japanese and foreign scholars focus on ningyo joruri, the ballad tradi­ tion, giving only passing attention to the ritual underpinnings o f this strong theatrical tradition. I have chosen to take a different approach. I am m ost interested in w hat is called shinji (or kamijyoto, an alternate reading, for “sacred m atters” ), the ritual use o f puppetry in this tradition. In the following chapters, I explore the history, development, and meanings o f rites perform ed by itinerant puppeteers in Awaji. I show how these rites used beings once removed from the hum an realm—puppets—to mediate the dichotomies o f order and chaos, purity and pollution, danger and safety, good and evil, the hum an and the divine, and in recent years, the ordinary and the exotic. The existence o f this ritual system carried o u t by

outcast itinerant puppeteers challenges com m on assumptions in ritual theory that posit “a shared system o f m eaning” as the basis for ritual effi­ cacy and significance. The man who told me the above story, whose grandfather sold his puppets to survive, was also the son o f the man who started the movement to preserve Awaji ningyo in the 1940s. I open this work with this particu­ lar story because it succinctly expresses the paradox contained in almost any study o f a Japanese “folk perform ing art” (minzoku geino) in the 1990s: While one may be trying to study the past o f a tradition, any understanding o f that past is refracted through tw o powerful lenses: the threatened extinction o f these folk perform ing arts and the nostalgic discourse supporting their revival. When I began my preliminary field­ work on Awaji puppetry in 1984, my initial interest was in how Awaji ritual puppetry developed and how its formation and dynamics could shed light on larger issues in Japanese nonecclesiastical religious practice and worship. I w ent to the village o f Fukura in southern Awaji n o t only b e­ cause it was the hom e o f the only surviving troupe on Awaji, the Awaji Ningyo Joruri Kan, b u t also because this was the best base from which to access archives relating to Awaji puppetry’s past and because it was close to where m ost o f the surviving puppeteers from the pre-W orld War II period lived. But studying the past o f Awaji puppets during the 1980s m eant that I was also witnessing the coincidence in time and place o f num erous currents in what could only be described as the dusk o f this once vital performance tradition: grief many people felt over the pass­ ing o f Awaji as a vital center for puppetry performances and life on Awaji, public expressions o f nostalgia for the “good old days” when visits by ritual performers and puppet festivals were com m on, manipulations o f these strong sentiments o f nostalgia and sadness for various political and economic ends, and in the late 1980s and early 1990s, the subsequent “revival” o f this tradition as a lucrative financial endeavor. People working at revival o f Awaji ningyo took w hat appeared to be a trajectory toward extinction and rerouted it into som ething else. My fieldwork coincided with—and in some strange and ironic ways contributed to —the revival o f this tradition. In a very real sense, this study begins with the crisis o f the tradition’s threatened extinction. The man in this particular story clearly expressed how he felt an essential part o f his identity had been lost when the Awaji puppets o f his childhood were sold to collectors from Osaka. I heard many stories with this same theme. O ne woman told me that near the end o f the war, when she was desperate, malnourished, and suffering from disease, she and her husband took the puppet heads that her father, like his father before him, had handled as a D bkum bo mawashi and burned them to heat

bath water to wash their sick, dirty children. She laughed wryly when she thought o f how m uch one o f those kashira (puppet heads) would bring today as an example o f “folk art.” Discussions o f Awaji puppetry often became narratives about where people were when the last performances took place and how they felt about the demise o f the tradition. I have heard this same man tell his story about the collectors from Osaka many times, and in each telling the story changed considerably, becoming more concise and affecting. Initially I was tem pted to dismiss this continuous transform ation o f the narrative as the fabrications o f an unreliable informant, but I came to see that reviving a tradition entails reworking life stories. Stories are more than just data. For people on Awaji, telling stories about one’s grief over the threatened death o f the tradition becomes a way o f legitimating one’s right to partic­ ipate in its revival. Revival constitutes a retelling o f the past, including the tradition’s “near death experiences.” Because these stories reveal so much o f the feelings that people on Awaji have for their puppetry tradition, careful sculpting and constructing o f narratives becomes a way o f process­ ing conflicting feelings o f grief, loss, confusion, and guilt. They have seen their island change dramatically in the last few decades, and they know that they have been unable to hold on to what they feel made Awaji a unique place in Japan: Awaji ningyo. Som ething about Awaji puppets has captured the imaginations o f many people living on the island, and for them Awaji ningyo has become deeply synonymous with a certain part o f their identity, tied up with what they have come to regard as a collective and intricate system o f meaning, which, they insist, people on the island all share. To tell the story of the death and rebirth o f Awaji puppets is to tell a story about a part o f themselves and their relationship to both the present and the past. Awaji has been a center for puppetry from at least the sixteenth century, when documents describe the activities o f puppeteers living in the district o f Sanjo in present-day M ihara-gun. Stories involving a puppet or a pup­ petry performance included striking hum an experiences, often told with a charged sense o f nostalgia or a strain o f grief. These stories seem as much a part o f the traditions as the facts about performances and puppets in the historical texts. W hat memories was the old grandfather seeking to recap­ ture when he made his tiny sketches o f the ritual puppets that he had sold to the collectors from Osaka? W hat emotions was the old man who told me the story trying to express when he continually revised this poignant event from his childhood into an ever tighter story? This story and others like it convey, in a way that accounts o f troupes being revived today in Japan cannot, the em otion and sense o f attachm ent people feel about puppetry performances, ritual or dramatic.

A Series o f Q u estio n s, an In teg ra tio n o f Sources The Awaji ningyo tradition as a whole raises a range o f historical and structural questions in the history o f Japanese religions: ( I ) the signifi­ cance o f the use o f effigies o f the hum an form in ritual appeasem ent and purification rites; (2) the ritual dynamics o f itinerant perform ance and the significance o f the “outsider” in the m aintenance o f the ritual purity codes o f medieval and early m odern Japan; (3) the social status problem s o f Japan’s “outcasts,” the category to which puppeteers belonged and from which their descendants have struggled to free themselves; (4) the sig­ nificance o f Sanjo districts as “geographies o f otherness” ; (5) the history and developm ent o f the kugutsu groups (w andering perform ers, both male and female), m entioned in literary texts from the m iddle o f the H eian period on; (6) the m ovem ent from sacred rites (shinji or kam igoto) to “secular” entertainm ent (in this case, joruri ballad recitation enacted by puppets) and the continued interaction o f the two; and (7) the rediscov­ ery o f the ritual origins o f this tradition and the political implications o f the revival o f Japanese “folk perform ing arts” w ith its ironic retrieval o f outcast arts as expressions o f the “authentic and the valuable.” W hile it is readily apparent th at each o f these im portant issues is inherent in the Awaji tradition, w ritten sources for a comprehensive study o f each issue are either fragm ented or nonexistent. In preparing this book, I have com ­ bined textual studies o f prim ary and secondary sources relating to both Awaji puppetry and other forms o f ritual and entertainm ent in Japanese history with extensive fieldwork on Awaji and around Japan. D oing so has led me to question the facile distinction drawn betw een “prim ary” and “secondary” sources. Every perform ance, every personal story about the past and the present, and every recorded interview becomes a new and different prim ary source, as do supposedly secondhand accounts from people living on Awaji. I first studied Awaji puppetry while preparing a doctoral dissertation on Awaji puppetry as it relates to Japanese folk religion. After I finished my doctoral studies, I continued to visit Awaji and participate firsthand in discussions about the extended revival o f puppetry th ro u g h o u t Japan. This book is a result o f both periods o f research, and although it draws on some o f the research o f my doctoral dissertation,2 it addresses decidedly different questions. M y earlier concern was w ith how a study o f Awaji ningy5 revealed larger dynamics in Japanese folk religion, an enterprise for which I adm it a declining enthusiasm . In this study, I am interested in how ritual puppetry in Japanese history has continually been in a process o f tradition form ation, attem pting to forge a tradition for itself, even as

that very tradition was continually being absorbed and appropriated by the larger dynamics o f Japanese religious and political life. D uring my predoctoral research on Awaji, conducted for one m onth in 1984 and again for several extended periods from 1987 to 1989,3 I en ­ gaged in four very different types o f research. First, for nearly tw o m onths I spent every day at the Awaji Ningyo Joruri Kan w atching rehearsals, perform ances, and daily activities at the theater, and talking with visitors who w ould come by to chat, see w hat was happening, and ask questions. This theater continues to be the hub o f both contem porary Awaji p u p ­ petry activities and ongoing scholarship into the history o f the tradition. In lengthy talks w ith perform ers and the director, U m azum e M asaru, I learned about the training experiences they had undergone and about how they understood puppetry’s uniqueness as a dram atic and ritual m e­ dium . I also interviewed the younger m em bers o f the theater separately, to understand how they perceived the tradition from the perspective o f the next generation. F or them , Awaji puppetry was about the present and the future, and I was curious how they related to the tradition’s past. While at the theater, I interviewed audiences w ho came to the Awaji Ningyd Joruri Kan on tours to see the brief perform ances. O ften I handed o u t questionnaires prior to perform ances. This work gave me a healthy distrust o f questionnaires as an exclusive fieldwork m ethodology in Japan. A lthough the questionnaires provided me w ith the occasion to converse w ith people in the audiences, the answers I received were largely polite attem pts to second-guess w hat I w anted to hear: “Lovely perform ance.” “Very nice dolls.” “W hat a lovely settin g .” “G am batte n e !” (G ood luck!) I used conversations with the audience, the occasional helpful question­ naire, discussions w ith m em bers o f the theater and the com m unity, and my ow n observations o f perform ances to reconstruct how people under­ stood the genre o f puppetry as a ritual and theatrical m edium . Second, during this period and on num erous other lengthy research trips in 1988 and 1989,1 talked w ith people directly connected w ith p u p ­ petry today. In 1987, m em bers o f the Awaji Ningyo H ozonkai (Awaji Puppetry Preservation Society) were just beginning a new wave o f activity to revive the num erous arts relating to puppetry, particularly puppet head carving, and costum e and stage prop production. I was able to m eet with O e M inosuke, the famous puppet-head carver for the Bunraku theater and the last master o f kashira carving alive today, in his hom e in N aruto, across the N aruto Straits from Awaji on the island o f Shikoku. Local scholars on Awaji were also starting new collaborations in research to re ­ vive Awaji ningyo and understand its history. Third, on several trips to Awaji from 1984 th ro u g h 1989, I spent m ornings at the ceremonial center o f Awaji ningyo, the O m ido H achim an Daibosatsu shrine in Sanjo. This shrine claims to be the birthplace o f

Awaji ningyo, and in the past all itinerant puppeteers setting o u t from Awaji w ould present themselves before it to venerate the patrons o f p u p ­ petry enshrined here: H yakudayh, D okum bo, Ebisu, and Akiba. Today, every m orning at ab o u t ten o ’clock local people from Sanjo (m ostly el­ derly farm ers) take a break from their work and play gateball, a gam e n o t unlike croquet, under the shady trees at the shrine. These leisurely m o rn ­ ings afforded me the o p p o rtu n ity to talk at length w ith people w ho had grow n up in a small village th at less than fifty years ago had been the hom e o f several active puppetry troupes. N o one ever w anted m e on their gate­ ball team , b u t I heard a lot o f stories, m ost o f them told m ore th an once and w ith great em bellishm ent over the years. Finally, I spent tim e at the Awaji Rekishi Shiryokan (Awaji H istorical Archives) in Sum oto, using m anuscripts and secondary sources difficult if n o t impossible to get elsewhere in Japan. T h ro u g h each o f these research strategies, I intended to observe the local response to the trad itio n , to understand how people o n Awaji u n d ersto o d th e m eanings o f Awaji p u p ­ petry, and to get som e picture, th ro u g h b o th historical research and stu d ­ ies o f the present situation, o f w hat prew ar Awaji pu p p etry was like. M y concern at th at tim e was one o f retrieval and reconstruction o f the religious rites perform ed by itinerant puppeteers, particularly the Sa-nbasd rite. Sanbaso ritual is the m ost sacred rite in the Awaji tradition. D erived from the larger Okina tradition in earlier Japanese perform ing arts such as N o h and Kagura, this rite represents the direct m anifestation in the hum an realm o f visiting deities (“old m en ” ) from the o th er w orld, and is perform ed to purify hom es, rem ove pollu tio n , and bring blessing from the sacred realm at the period right after the N ew Year. This ritual forms the m ain pillar in the religious use o f pu p p etry in Japan, and so th e focus o f my study was on this rite before its dem ise in the tw entieth century. Awaji puppeteers had traveled all over Japan perform ing these rites, and m any o f them had settled in o th er regions and started their ow n p uppetry troupes, considering them selves to still be “Awaji pu p p eteers.” Som e o f those Awaji lineage centers still have rem nan t perform ances presented at shrines as p art o f a ritual calendar. I was interested in using th e Sanbaso perform ances o f these o th er Awaji lineage theaters to reconstruct a picture o f the Sanbaso ritual on Awaji at an earlier tim e. In 1988, I traveled to N agano prefecture and atten d ed the International P uppetry Festival in Iida-shi. T h en I w atched p uppetry perform ances at the K uroda N ingyo Za (presented o u td o o rs in the precincts o f the H achim an shrine), and inter­ viewed the puppeteers at th a t th eater and at the Waseda N ingyo Za nearby. T he pu p p etry traditions in b o th o f these shrines were started by an Awaji puppeteer in the early eighteenth century. In the m idst o f my studies, an exciting discovery in the tow n o f M orioka in Iwate prefecture b ro u g h t to light puppets, texts, and ritual

implem ents from the Awaji tradition. An Awaji puppeteer by the name o f Suzue Shirobe had perform ed in M orioka around 1641 and had estab­ lished an Awaji lineage theater in the area, m anipulating his puppets for the Lord o f N anbu and even the Em peror. There has n o t been an active theater in M orioka since before the Meiji Restoration in 1868, b u t I wanted to examine the puppets and docum ents found there as a reference point in reconstructing the types o f puppets used in the early Awaji tradi­ tion. In February 1989 I spent several days in M orioka examining the findings. At the end o f my fieldwork on Awaji in 1988, I spent nearly two m onths traveling thro u g h o u t the Inland Sea and around Kyushu, exam in­ ing references to the mythical founder o f Awaji puppetry, Hyakudayu, and trying to unravel the relationship betw een this nam e, ritual puppetry, ap­ peasem ent o f m alevolent spirits, and worship o f the deity Hachim an. This period o f focused fieldwork prior to writing my doctoral disserta­ tion impressed upon me the need to understand the “local know ledge” about religious activities, to listen to the stories o f the people involved, and to attend to their issues. Research trips from the northernm ost part o f H onshu in Iwate to the tip o f Kyushu indicated the wide scope of this im portant puppetry tradition in Japan. This perspective impressed on me the im portance o f understanding the road as sacred geography in this tradition. As I was involved in the research o f reconstructing Awaji puppetry’s past, I had a series o f persistent questions that I consider herm eneutically significant. First, I was continually interested in the activity o f memories, even painful m em ories, as a lens through which this tradition is recon­ structed. While I viewed nostalgic discourse w ith some suspicion, I was interested in understanding w hat kind o f creative and even healing process was being served by this “painful longing” for the past and by the effort to retrieve what was rapidly vanishing. H ow are we to understand this revival o f a ritual and dramatic tradition? M y second question was, W hat was I doing working alongside people for w hom this tradition was p art o f their past, trying to sort through fragm ents o f evidence to piece together a coherent history? Why Awaji and n o t Kalispell, M ontana (or Silver Spring, M aryland, or H am pstead, England or wherever we are from)? Why their past and n o t my own? Nostalgic discourse and the telling o f stories about the past is a m ultitextured field. The interplay o f nostalgic discourse and tradition form ation is perhaps best seen initially in som eone else’s tradi­ tion, but it is capable o f being generalized to shed light on activities closer to hom e. This book addresses these fundam ental issues and confronts the related problem o f fragm ented evidence and the fabrication o f seamless “histories” by scholars for a variety o f ends. Although the first several chapters are about the history o f ritual pup-

p e try in Japan and the rise o f Awaji as a center for ritual p u p p e try in th e early seventeenth century, my focus in these studies has shifted from my earlier interests. I am less interested in reco n stru ctin g the past o f ritual p u p p e try in Japan than in show ing how the traditions o f ritual p u p p etry and worship have been continually reco n stru ctin g the past and forging new identities and traditions. M y studies on Awaji in the sum m ers o f 1990 and 1991, and again in the au tu m n o f 1 9 93, have provided the o p p o rtu n ity to explore these issues m ore directly, since d u rin g those trips I was fo rtu n ate to participate in th e discussions and developm ent o f the forum for reviving p u p p e try aro u n d Japan, th e triennial AU Japan P u p ­ p etry Sum m its (Z enkoku N ingyd Shibai S am itto), w hich occurred from 1990 to 1993. This book directly addresses the issue o f nostalgia and the revival o f folk culture in Japan today. Far from being a m o d ern p h e n o m e n o n , n o s­ talgic discourse is p a rt and parcel o f the form ation o f trad itio n th ro u g h ­ o u t history, even in th e case o f ritual puppetry. F inding a past and m aking it co h eren t is at one level a m ost fundam ental h u m an activity. A p p ro p riat­ ing th e past is p art o f th e narrative process, and narratio n is a special kind o f know ledge th a t people can have a b o u t them selves. N arrative c o n ­ stru ctio n , alm ost by definition nostalgic at som e level, can even have a healing effect.4 In chapter I I lay th e g roundw ork for a discussion o f ritual p u p p etry in Japan by considering p u p p e try as a ritual m edium and by exploring th e threads o f influence from various m agical uses o f dolls, effigies, and puppets th a t find expression in Japanese ritual. T his discussion shows the variety o f ritual uses o f puppets in Japanese religions and also indicates how it is n o t possible to argue for a single trad itio n o f Japanese puppetry. I p o in t o u t th e m ultivalence o f the term ningyo, w hich I translate as “p u p ­ p e t.” I suggest th a t in each case o f th e ritual use o f dolls, effigies, and pu p p ets, there is a concretization o f a narrative event in th e life o f a person o r a com m unity. C hapter 2 is a study o f th e issues o f sameness and difference in Japanese ritual. I t reveals the need to have outsiders, in this case outcasts, to p er­ form certain ritual functions in Japanese society. T h e discussion explores the developm ent o f th e ritual specialist as O th er, and th e natu re o f the kadozuke (literally, “attached to the g a te ” ) perform ance tra d itio n , in w hich itinerant specialists w ere responsible for m ediating ord er and chaos, the realms o f th e hu m an and the divine, and pollution and purity th ro u g h ritual perform ance. I explore the history o f tw o form s o f itin eran t perfo r­ m ance, Senzu m a n zu i and M utsu-bayashi, to reveal the stru ctu re o fitin e rant perform ance appropriated by nonecclesiastical specialists discussed later. Since the Awaji ningyo tradition was based u p o n itin eran t ritual p er­ form ance, this chapter lays the g roundw ork for those th a t follow.

Chapter 3 turns to a discussion o f clearly m anipulated ningyo as o p ­ posed to ningyo as static objects. I take up the issue o f the late H eian kugutsu groups (itinerant ritual puppeteers). The discussion presents the major evidence we have about these groups, as well as a th o ro u g h discus­ sion o f the activities o f ritual specialists several hundred years later, also called kugutsu, attached to the Nishinomiya Ebisu shrine. The deity Ebisu, usually represented as a fat, happy-go-lucky deity with a fish under his arm, is in fact understood to be capable o f causing epidemics, disasters at sea, and other natural calamities. Ritual puppeteers were understood to mediate and transform the spiritual powers o f this enigm atic and danger­ ous deity into a deity o f luck (fukushin) as Ebisu is com m only understood today. These Nishinomiya Ebisu-kaki (bearers or carriers o f Ebisu) trav­ eled the countryside dissem inating the power o f the Nishinomiya center and perform ing ritual appeasem ent for Ebisu by using a puppet under­ stood to be its spirit vessel (goshintai). Since this is the ceremonial center from which Awaji puppeteers broke away, and against which the Awaji tradition comes to define itself, understanding the Nishinomiya kugutsu and Ebisu’s liminal nature is a necessary step in appreciating the form ation o f the Awaji ningyd tradition. It is clear that by the middle o f the sixteenth century, the ties that bound the Nishinomiya kugutsu to this ceremonial center were w eaken­ ing as these puppeteers struck o u t on their own, engaged in m ore dra­ matic styles o f entertainm ent, and became increasingly popular. Chapter 4 studies the rise o f Awaji puppetry as a distinct tradition developing o u t o f its Nishinomiya origins. While puppeteers may have been active on Awaji prior to the sixteenth century, only then did Awaji puppetry develop a unique identity in contrast to the Nishinomiya kugutsu. T he discussion focuses on how one text, Ddkumbo D enki, became a central work in the form ation o f the Awaji ningyo tradition. Being a tradition means having a com m on story about the past. I present evidence showing how this nar­ rative gave the Awaji tradition a coherent center and thereby encouraged the spread o f Awaji puppetry th ro u g h o u t Japan. C hapter 5 presents an in-depth study o f the two sacred performances Shiki Sanbaso and Ebisu-mai, their use in itinerant perform ance, and their inclusion in larger puppetry events on Awaji called ningyo mutsuri. The discussion includes a translation o f each piece along with a detailed discus­ sion and an interpretation o f its content. I focus on the m ultitextured alterity o f these rites. Puppets, puppeteers, and deities are all “outsiders” in the Japanese order o f m eaning, and these rites become powerful invo­ cations o f these forces from outside the everyday realm, capable o f revital­ izing and purifying com m unities. I show the interaction betw een the developed joruri performances and the ritual perform ances, and I also discuss a theatrical spectacle unique to the Awaji tradition know n as ddgu-

Jfaeshi (literally, “changing stage sets” ). N ingyo m atsuri, presented either in the spring right after planting or in the fall right after harvest, were understood to be festivals for the kam i (deities) o f agriculture. We explore the dynamics o f these elaborate ritual and theatrical events, perform ed over several days on tem porary stages or in shrine precincts. T he final chapter is a discussion o f the contem porary state o f Awaji puppetry. Today, Awaji ningyo appears to have m ade a com eback, and the Awaji N ingyo Joruri ICan has becom e a successful financial venture, w ith the troupe m aking overseas tours at least once a year and giving daily perform ances at the new, specially built indoor Fukura theater. In this discussion, I trace the history o f this com eback and discuss the coinci­ dence o f regionalism , revivalism, and nostalgia in Japanese society in the last tw o decades. I discuss the status o f Awaji ningyo’s ritual tradition in this retrieval process, and the irony o f its initial exclusion and later in ­ clusion in the revival. Finally, I explore the role o f the foreign scholar in this process.

M etanarratives in th e Study o f Japanese R eligious H isto ry This book deals w ith an expression o f nonecclesiastical, popular religion in Japanese history. It is centered on the activities o f a group o f people w ho occupied the m argins o f Japanese religious life and society, while perform ing ritually integral activities to keep the central systems o f m ean ­ ing functioning. This study adds to the grow ing num ber o f m onographs th at argue against a singularity o f experience in Japanese popular religious life and ask us to see Japanese history as radically heterogeneous, even th o u g h there have been strong attem pts b o th w ithin and outside Japan to posit a n o tio n o f a singular identity. U ntil recently, m ost scholarship on Japanese folk religion has posited a unity o f experience at the popular level. This idea o f a uniform “ folk cu ltu re,” articulated in the U nited States by the w ork o f R obert Redfield, was in Japan m ost ardently advocated by the folklorist Yanagita Kunio. Yanagita argued th a t people shared a unified experience, half conscious at tim es, parallel to and equally as im p o rtan t as the Japanese imperial line. “T he folk” in H okkaido shared a strata o f consciousness w ith “the folk” in Kyushu. Yanagita constructed w hat has becom e the rudim entary fram ework o f all studies o f Japanese folk religion, although it ignores groups such as Korean im m igrants living in Japan, outcasts, and o ther people whose experiences som ehow deviated from his norm . Yanagita, hoped to give voice and identity to people who could n o t speak for th e m ­ selves, and to invest the everyday life ( nichijo seikatsu) o f “ folk culture” w ith a prestige it had previously n o t enjoyed. To a certain extent, had it

n o t been for the ethnographic enterprise spawned by Yanagita’s agenda and his enthusiasm for it, m uch o f the data used in this study w ould have been perm anently missing from the historical record. But another result o f his enterprise was the construction o f an identity for the imagined masses in Japan, and a m etanarrative o f Japanese religious history that has becom e hegem onic and oppressive. A ccording to Yanagita, the folk were rural and to a certain extent ahistorical. They shared a com m on heritage and set o f concerns and religious sensibilities. His im agination o f the folk transcended description and became a prescription for an authentic Japanese identity. Yanagita’s view o f folk religion was expanded upon and systematized m ost directly by his son-in-law, the folklorist H ori Ichiro. H ori argued for a unity o f Japanese religious experience: “In this context we may point o u t that there is good reason to speak o f Japanese religion as an entity.”5 H e identified six “com m on tendencies” he claimed could be considered continuous in Japanese folk religion: em phasis o n filial piety (ko) an d an c esto r w orship co n n e c te d w ith th e Japanese family system ; em phasis on on (d eb ts an d favors given by superiors) an d boon (th e re tu rn o f ori)\ m u tu a l b o rro w in g an d m ixing o f d ifferen t religious tra d i­ tio n s (o r syncretic te n d en c y ); b elief in th e co n tin u ity b etw e en m an an d deity, o r easy deification o f h u m a n beings; coexistence o f d ifferen t religions in on e family o r even in o n e perso n ; stro n g b elief in spirits o f th e dead in c o n n e ctio n w ith an c esto r w orship as well as w ith m o re anim istic co n c ep tio n s o f m alev o len t o r b en ev o len t soul activities.6

This approach to Japanese popular religion was echoed by Joseph Kitagawa, who also proposed that Japanese religions shared a unity o f consciousness. While he never addressed the issue o f “folk religion” di­ rectly, he m aintained that it is possible to speak about Japanese religion in the singular (just as he w ould m aintain that one can learn a great deal about various data by understanding Japanese religions in the plural). As well, Kitagawa identified a series o f com m on tendencies: ( I) a nonsymbolic understanding o f symbols, or the tendency for the represented to participate in its representation (he writes th at “the m eaning o f each being was sought not in itself but in it m utual participation, continuity and cor­ respondence to and w ith others w ithin the total framework o f the m onis­ tic world o f m eaning” );7 (2) viewing past, present, and future in a fluid relationship to one another, evidenced by the tendency to interpret past events in term s o f present realities, or w hat Kitagawa calls (inverting the Augustinian axiom) “a past o f things present” ; and 3) the tendency to view the given world as the ultim ate w orld, and the Japanese nation “as the measure o f all things.”*

Kitagawa goes one step further than H o ri in m aintaining th a t these tendencies all p o in t tow ard a Japanese sensibility o f “seam lessness.” This seamlessness, dem onstrated in th e Japanese conception o f the hum an re ­ lationship to the natural w orld, m atter, symbols, and sacredness, co m ­ prises w hat Kitagawa calls a “ unitary m eaning stru c tu re ” : “As far as one can tell, one o f the basic features o f the early Japanese religious u n i­ verse was its unitary m eaning structure, a structure w hich affirmed the belief th a t the natural w orld was the original w orld, and w hich revolved around the n o tio n th a t the total cosm os was perm eated by sacred o r k a m i nature. E verybody and everything in the early Japanese m onistic u n i­ verse, including physical elem ents such as fire, water, w ood and stone, as well as animals and celestial bodies, w ere believed to be endow ed w ith kami n a tu re .”9 Yanagita, H o ri, and Kitagawa can be seen as representative o f a p articu­ lar approach to Japanese popular religion which, th o u g h often insightful, posits a continuity across consciousness (Yanagita), geography (H o ri), and tim e (Kitagawa). O ne w onders just w hat tattered seams in society are concealed by such an adam ant insistence on a single piece o f fabric. Such an approach is problem atic, particularly in ou r discussion o f ritual puppetry, in th a t it tends to blur differences and exclude persons and data th at do n o t fit w ith the currents o f th e grand narrative o f “ Japan” as an idea.10 R ecent scholarship guided by the zealous instincts o f p o stm o d e rn ­ ism has directed our gaze tow ard those people w ho for different reasons have stood outside the grand narrative o f Japanese history. This same cur­ ren t o f scholarship also warns us n o t to be to o strident in o u r claim to speak for those w ho have n o t been spoken for in history. H . D . H a ro o tu n ian writes as follows ab o u t Yanagita’s agenda to give voices to the “folk” : “H e com m itted him self to the im possible task o f trying to speak for the folk outside the language o f pow er and reason th a t had concealed them from view. H e m isrecognized his task, failing to see the m isfortune o f the folk— the interm inable m isfortune o f their silence and their failure to se­ cure representation for them selves— lay precisely in the fact th a t as soon as a person attem pted to convey their silence, he passed over to the side o f the enemy, the side o f history and reason, even if, at the same tim e, he puts the claims o f the enem y into q u estio n .” 11 In this study, I do n o t claim th a t a unitary structure o f m eaning u n d e r­ girds the various and w idespread uses o f puppets in ritual practice in Japan. N o r do I claim to have som ehow given voice to the unvoiced “o u t­ siders” o f Japanese religious history. My claim is m ore hum ble: to present the fragm ents o f evidence we have for a study o f a particular case in Japa­ nese religious history, and in the process provide insight into a recurring dynamic in Japanese religious history— the interaction betw een the cen-

tral religious authorities, for w hom a unified structure o f m eaning is a useful device, and those people on the margins, whose very ritual actions both challenge and m aintain the power o f the center. This study o f ritual puppetry is about one group o f “others” in Japanese religious history. To a large extent, the lives and stories o f ritual p u p ­ peteers and their activities have been left o u t o f the historical record. O ften ail we have are relics (even icons) from the past, pieces o f physical o r textual evidence with few clues about how they were used or under­ stood or what they m eant. It is n o t possible to tell their story for no one knows what it is. This book, however, presents the evidence that we do have. By putting it in one place and examining it, perhaps we will hear other voices in the history o f Japanese religious life, telling stories other­ wise lost to us. I situate this concern within a larger intellectual movem ent o f breaking away from grand narratives o f the center, a m ovem ent that dem ands we “activate the differences” and promises to reflect a m ore textu red picture o f Japanese religious histories.12

I In th e Shape o f a Person: T h e Varieties o f R itual U ses o f Effigy in Japan

I F IR S T SAW a p erform ance o f Awaji ningyo in the fall o f 1 9 7 8 , w hen I h ap p en ed in to th e dilapidated th ea ter o n th e w a te rfro n t o f F u k u ra Bay, in a small fishing village in th e so u th e rn p a rt o f th e island. T h e th e a te r was in a large hall located over a souvenir shop. A few d o z e n folding chairs were set up in row s for th e audience. D ow nstairs, one could buy seaw eed from th e nearby In lan d Sea, and trinkets w ith a regional flair. T h a t day I saw th e pilgrim age scene from Keisei A w a no N a ru to , th e story o f a y o u n g girl w h o , separated from h er p arents since the age o f th ree, goes lo o k in g for th em w earing th e g arb o f a pilgrim on th e Saikoku Pilgrim age ro u te . In the scene, she arrives at th e h o m e o f a w om an w ho quickly recognizes th e ch ild ’s sto ry a n d realizes th a t she is h er d aughter. A h eart-w re n c h in g dram a unfolds as we learn o f th e events th a t led to the separation o f th e p a re n t and child. F o r political reasons, th e w om an is forced to conceal her identity as th e girl’s m o th e r, and in th e end sends her away to co n tin u e h e r search for her parents. T h e m an ip u latio n o f sym bols im p a rte d a tragic feeling to the p e rfo r­ m ance. T he audience becam e aware th a t at som e level b o th m o th e r and child know they have b een reu n ited , b u t m u st experience th eir separation all over again as th e child is th ro w n o u t th e d o o r and it is slam m ed sh u t behind her. It was a pow erful perform ance. Like the B unraku th e a te r for w hich Japan is fam ous, th e p u p p ets used in the p erform ance w ere each m an ip u ­ lated by th ree p e o p le .1 T he p u p p eteers w ere clo th ed in black and covered w ith h o o d s, called kuroko. U nlike the B unraku p u p p ets, th e p u p p ets used in this p erform an ce s to o d well over a m eter h igh, giving th em an eerie quality as th ey w ere m oved th ro u g h th eir paces by shadow y figures w ho co n tro lled th eir destinies in this tragic dram a. T h e large size o f Awaji p u p ­ pets, one o f th eir distinguishing features, developed from th eir ro o ts as rural dram a p resen ted in o u td o o r theaters an d m akeshift spaces. B igger pu ppets are easier to see. T h e sm aller, delicate dolls o f th e B unraku th e a te r w ere developed to suit th e aesthetics o f an in d o o r theater. C o n c e rn in g th e large size o f Awaji ningyo, th e scholar o f Japanese p u p ­ p etry N agata Kokichi c o m m e n te d , “You k n o w th a t Awaji ningyo are a little larger th an B unraku puppets. T h e reason is th a t Awaji p u p p e ts a re n ’t

meant to be stage puppets primarily. They are puppets o f the road, m eant to be perform ed in people’s entryways and courtyards, by the sides o f roads, in shrines, wherever there is enough space and people to put on a show. Bunraku puppets are lovely, but those are puppets o f the stage. If you take Awaji puppets and stick them only inside a theater, it is like p u t­ ting a wild animal in a zoo. It loses all its wildness. I like Awaji puppets because as they have developed they have maintained their yaseimi [innate wild nature].”2 The story o f this girl and her m other was in the joruri ballad style. Be­ cause the action was presented with puppets, the text was recited by a single chanter, whose forceful interpretation was punctuated by the sharp notes of a shamisen, a three-stringed instrum ent played by an accompanist. Al­ though the piece I saw had only two characters, there were eight perform ­ ers on stage throughout, with o f course the unseen crew backstage manag­ ing the backdrops. While the performance was undeniably lovely, there seemed to be something intentionally inconvenient about using puppets. I was struck by a very basic question about this theatrical medium: Why go to all this trouble? Surely having three people manipulate one large doll with the characters’ lines and even stage asides recited by a chanter is more costly and cumbersome than using hum an actors. (H ow much more cumbersome it m ust be when a perform er is itinerant and has to haul these puppets around, as was the case in the purification rituals o f the Awaji tradition!) W hat is it that a puppet can express that a hum an perform er cannot? This book is in part an exploration o f that question. O n the one hand, I have been interested in the general phenom enology o f puppetry as a theatrical medium and in the discussions by theater specialists, puppe­ teers, and ritualists around the world who struggle with this same ques­ tion: why puppets? O n the other hand, I have tried to discover how these large figures made o f paulownia wood and fabric are understood in Japan. Clearly a broader understanding o f the power o f effigies, dolls, and body substitutes in Japanese religion has contributed to the developm ent o f this ritual tradition. In Japanese ningyo is written with two characters m eaning “person” and “shape.” I have chosen to translate this term as “puppet” although in other contexts it could as easily be translated as “doll” or even “effigy.” My decision to translate this term this way was uncomplicated: In the first place, most scholars writing in western languages about Japanese ninjyyo shibai (drama using ningyb) translate the term this way, or see an affinity between this type o f performance and other puppetry forms around the world. Second, I maintain th at the abstract idea o f the puppet as a descrip­ tive category goes the furthest toward helping us understand the decision to use these nearly life-size beings in rituals o f appeasement and purifica-

tion, and eventually dram atic theater. T h ro u g h o u t the w orld, the deci­ sions to use effigies rather than hum an actors in ritual share som e co m ­ m on religious concerns. In this chapter I w ant to set Awaji ningyo in its larger context. First, I explore w hat a com parative study o f p uppetry can d o for o u r u n d e rsta n d ­ ing o f this particular case. Second, I survey o th er ritual uses o f ningyo that inform th e rise o f ningyo shibai, particularly on Awaji, in the m id-six­ teen th century. In h ere n t in the follow ing discussion is an abstraction: in different effigies o f th e hum an form in a variety o f cultural contexts aro u n d th e w orld, there is a category we have com e to regard as “ p u p p e t.” A lthough in this book I present m aterials concerning the realities o f p u p ­ pets in one region, I m ust begin by talking ab o u t “p u p p ets” in general. By looking at the category o f p u p p et “ bracketed” (as phenom enologists say) o u t o f its context, I feel it is possible to becom e sensitized to th e deeper m eanings in h eren t in specific cultural cases. W hile I am n o t looking for the “ essential p u p p e t,” I think a prelim inary phenom enological approach lays the groundw ork for a descriptive understanding o f the specific Awaji m aterials o f later chapters. This m ethodology, th o u g h unable to account for specific m eanings o f one cultural context, can train o u r eyes to see im p o rtan t aspects o f the data. Studying ritual pu p p etry in Japan, I am continually am azed by the p ro ­ vincialism o f Am erican th eater audiences and even scholars w hen it com es to the issue o f puppets. Because in ou r ow n culture puppets have been relegated to the playroom and because gifted puppeteers m ust ply their trade presenting poorly developed skits for children (w ho, one can only assume, m ust be partly offended by adult assum ptions ab o u t their tastes), it is often difficult for Am ericans to realize th a t in o th er cultures puppets are used to stage serious dram as for m ature audiences. I cannot c o u n t the tim es I have had to defend the legitim ate rig h t o f a historian o f religions to take Japanese puppets seriously. A large p a rt o f studying p u p p etry is spent apologizing, justifying and continually explaining this choice to otherw ise broad-m inded and intelligent people. M ost scholars o f pu p p etry are forced to m ention this problem at the outset. C on fro n tin g this bias as I studied Japanese puppetry, I found m y­ self in a strange so rt o f kinship w ith the sentim ents o f th e flam boyant early tw en tieth-century British playw right and d irector E. G o rd o n Craig. H e considered the p u p p et an example o f tru e and total th eater (while suggest­ ing actors could be done away w ith). In his fam ous essay “T he A ctor and the U b e rm a rio n e tte ” he expressed this m ost vehem ently: To speak o f a p up pet w ith m ost m en and w om en is to cause them to giggle. T h ey think at on ce o f the wires; they think o f the stiff hands and the jerky m ovem ents; they tell m e it is “a funny little d o ll.” But let m e tell them a

few things ab o u t these Puppets. L et me repeat again th a t they are the descen­ dants o f a great and noble family o f Im ages, Im ages w hich w ere m ade in the likeness o f G od; and th a t many centuries ago these figures had a rhythm ical m ovem ent and n o t a jerky one; had no need for wires to su p p o rt them , n o r did they speak th ro u g h th e nose o f th e hidden m anipulator. (P oor P u n ch , I mean n o slight to you! You stand alone, dignified in your despair, as you look back across th e centuries w ith painted tears still w et u p o n your ancient cheeks, and you seem to cry o u t appealing!y to your dog, “Sister A nne, Sister A nne, is nobody com ing?” A nd th en w ith th a t superb bravado o f yours, you tu rn the force o f your laughter [and my tears] upon yourself w ith th e heartren d in g shriek o f “ O h my nose! O h my nose! O h my nose!” ) D id you think, ladies and gentlem en, that those puppets were always little things o f b u t a foot high? . . . I f we should laugh at and insult the m em ory o f the P u p p et, we should be laughing at the fall that we have b ro u g h t ab o u t in ourselves, laughing at the Beliefs and Im ages we have broken.3

O verw rought th o u g h his prose may be, C raig’s com m ents point to an im p o rtan t elem ent o f pu p p etry traditions around the world. N o t infre­ quently puppets have been used in religious rites to represent and em body w hat are seen as divine forces. T he discussion th at follows— o f puppetry as it is viewed around the w orld and o f the wide range o f ritual and dram atic uses o f puppets in Japan— will properly context ualize the u nfo rtu n ate cul­ tural bias o f recent Am erican audiences and scholars, which reduces p u p ­ pets to pathetic com edians in a second-rate theater.

W h at Is a Puppet? In com m on usage, the w ord English w ord “p u p p e t” has a wide variety o f referents. Webster’s Third Intern a tio n a l D ictionaryprovidcs the following inform ation: [M E popet, fr. M F poupette,] la . little doll, dim . o f assumed poupee, doll, fr. assum ed VL puppa, alter scale figure o f a hum an or o th e r living being often co nstructed w ith jointed limbs appropriately painted and costum ed and usually m oved on a small stage by a rod o r hand from below, o r by strings o r wires from above— see m arionette, lb . an actor in a play o r a pantom im e, 2. archaic: ID O L , 3. archaic: a vain, gaudily dressed person, 4. one w ho acts or is co ntrolled by an outside force or agent and is no longer the arbiter o f his ow n situation.

W hile this dictionary definition gives us the com m on uses o f the word, it provides no criteria to justify grouping the m any kinds o f puppetlike objects un d er one heading. In Japan, w hat in English we w ould translate as “ puppets” can be nearly life-size dolls, w ith elegantly carved and care-

lacquered heads having m oving eyes, m o u th s, and even eyebrow s, exquisitely co stu m ed and realistically m anipulated by as m any as th ree persons. In th e T o h o k u district, small heads o f horses o r w om en are a t­ tached to th e ends o f sticks and m anipulated by sham ans. O n Sado Island tiny dolls act o u t com ical skits betw een m ore serious pieces, always peeing on so m eone in th e audience and draw ing a few laughs before the p e rfo r­ m ance retu rn s to th e stuffier h ig h b ro w dram a. O r p u p p ets m ay be small jo in ted dolls said to be e n te red by deities in a sacred sum o m atch c o n ­ d u c te d to appease m alevolent spirits. In In d o n esia, Turkey, S outheast Asia, and C hina, w ords com m only translated as “p u p p e t” refer to flat pieces o f leath er cu t a n d painted in the likeness o f any n u m b er o f characters, held up in fro n t o f a lam p to cast shadow s o n a clo th screen. A “p u p p e t” can also be a doll held by a ro d o r m oved by h and. T he w ord also refers to the m ario n ette, a doll m an ip u ­ lated by a com plicated series o f strings or wires. A glove w ith a face pain ted on it is also a p u p p et. In th eaters in th e Soviet U n io n d u rin g the 1960s, th e w ord referred to th e chairs and tables m oved a b o u t by actors as p a rt o f a perform ance. In W est G erm any’s experim ental th e a te r in th e late 1960s, the w ord was given to a series o f geom etric shapes w hich danced to m usic, joined in th e final act by th e cello— also a p u p p e t. T h e a te r directors in the Soviet U n io n and W est G erm any consciously in te n d e d these exam ples as deco n stru ctio n s o f the c o m m o n assum ptions th e audience had a b o u t p u p p e try as representations o f th e h um an. T hey th ereb y h o p e d to m ove the theatrical m edium to a new level o f vitality. Each o f these exam ples has a special nam e in its ow n cultural co n tex t, yet we have com e to call th em all p u ppets in English. Likewise, scholars in Japanese perform ance w ould consider th em all o f ningyo shibai, th e p u p ­ petry trad itio n . O u r unconscious categories are divorced from th e issue o f m echanical complexity. H o w have we com e to take this step? W hat is com m on to all these examples? T h e A m erican p u p p e te e r Bil Baird had an answer: “A p u p p e t,” he w ro te, “is an inanim ate figure th a t is m ade to m ove by h u m an effo rt b e­ fore an aud ien ce.”4 A ccording to this definition, a p u p p e t is n o t m erely a doll. I f a child begins to m ove a doll and give it speech, the doll is still n o t a p u p p e t. B ut if this child m oves th e doll to anim ate it and presents the perform ance before an audience o f p arents o r friends, he o r she is partici­ p ating in a basic form o f puppetry. In this sam e vein, a p u p p e t is n o t the m echanized m an o r bird th a t pops o u t o f a clock at the strike o f an hour. Such anim ation lacks the elem en t o f h u m an effo rt behind th e m ovem ent. P uppetry, according to B aird’s widely accepted definition, m ust sim u lta­ neously co n tain an inanim ate object, a h u m an e ffo rt, and an audience. A ccording to this helpful definition, the co n tex t o f perform ance is an es­ sential elem ent in w h at constitutes puppetry. fully

A lth o u g h B aird’s definition does provide lim itin g criteria fo r th e ca te­ gory, it is still in ad eq u ate. I f o n e waves a pencil in th e face o f a spectator, does th e pencil th e n becom e a puppet? Such action conform s to his d efini­ tio n . T w o m o re in g red ien ts are necessary: th e in te n tio n to co m m u n icate so m eth in g m eaningful (even if th a t m eaning is th e m eaninglessness o f life) and som e idea o f rep resen ta tio n . O u r linguistic usage o f th e w o rd pu p p et in such a w ide variety o f contexts indicates a general aw areness o f th is schem a. W h at is fascinating a b o u t th e ritual use o f p u p p e try in th e Awaji case is h o w this tra d itio n ’s fo rm u latio n o f each o f these p o in ts— inan im ate o b ject, h u m an e ffo rt, audience, intentionality, and re p re se n ta ­ tio n — challenges m any o f o u r assum ptions a b o u t ritual and theater.

T h e A ppeal o f P uppets: Five Perspectives I have always b een in tere ste d in h ow people w ho have seen serious p u p p e t th e a te r feel a b o u t th e m ed iu m . R esponses te n d to be polarized: people eith er love o r h ate pu p p ets. A m ong th o se w ho h ate th e m , th e reason m o st co m m o n ly given is th a t p u p p ets give peo p le the creeps. T h ere is so m eth in g u n settlin g a b o u t im ag in in g th a t inanim ate hum anlike figures are actually h u m an and th e n b ein g b ro u g h t back to th e aw areness th a t these effigies are in fact n o th in g m o re th a n w o o d an d p ain t. F o r o th ers, this is precisely w h a t is pow erfully appealing a b o u t p u p p etry . S eeing a g o o d p u p p e try p erform ance is like w a tch in g m agic at w ork. A p u p p e t th e a te r p resents an ap p reh en sio n o f th e bo u n d aries b etw een reality and illusion, b o d y an d soul, h u m a n and n o n h u m a n , th e living an d th e dead, and th e m aterial an d im m aterial w orlds. P u p p ets, I argue, elicit an in ­ tensity o f response in an audience precisely because th e use o f an in an i­ m ate o b ject allows fo r an ex ploration o f th e h id d e n processes o f th e im ag in atio n , an d in th e case o f ritual p u p p etry , th e in n er w orkings o f th e spiritual life. Below, I p re se n t five d iffe ren t perspectives describing th e appeal o f p u p p ets as a ritual o r th eatrical m ed iu m , draw n prim arily from W estern language sources. Each reveals so m eth in g o f th e experience th a t people have o f p u p p ets as a th eatrical m edium . ( I ) P u p p ets have allow ed for a p articu lar freed o m o f expression n o t possible w ith h u m an actors (u n til th e ad v en t o f special effects in film ). (2) T h e use o f p u p p ets enables h u m a n beings to im agine an escape from a seem ingly inescapable fate, an d to create b ey ond the co n strain ts o f the h u m a n co n d itio n . A perso n can assum e th e ro le o f th e creato r an d c o n tro ller o f beings. (3) T h e tra n s­ fo rm atio n o f an in n ate, m aterial ob ject in to a living an d b re ath in g charac­ te r satisfies a creative spiritual an d psychological need. (4) P u p p ets are m o re convincing because they rem ain w h a t they are— they d o n o t “ repre-

IN T H E S H A P E OF A P E R S O N

23

sent” their p a rt and th en revert to som eone else w hen they walk o ff stage. T h ro u g h th e intentional inconvenience o f having to “ rep resen t” the h u m an being, puppets m etaphorically figure the hum an condition. (5) Puppets represent an awareness o f the relationship betw een the m aterial and spiritual realm s and are able to becom e vessels for visiting spiritual forces.

Puppets an d the Freedom o f Expression A d o m in an t feature o f a th eater using puppets instead o f hum an actors is th e possibility o f greater freedom o f expression. Puppets can be m ade to do things w ith ease th a t hum an actors can do only w ith great effo rt and elaborate theatrical devices. Puppets can fly, change in to anim als, ghosts, and dem ons, dism em ber them selves and one another, and enact terrific physical transform ations. A beautiful m aiden can reveal herself to be an ugly d em o n o r a terrifying fox spirit, and suddenly change again in to the m aiden— a device com m on in the classical jo ru ri p u p p e t theater, using a g a b u (trick head). Just as a person may have flashes o f insight in to so m e­ one else’s tru e n atu re, so to o p u p p et th eater audiences gain fleeting reve­ lations th ro u g h th e use o f gabu. In p u p p e t th eater all over the w orld, fu rth erm o re, a lth o u g h extensive speaking parts may be given to child characters, the actions o f a p lo t are narrated by an adult (such as the d h a la n g in Javanese shadow th eater or the tayii in Japan). H u m an child actors may have difficulty rem em bering many lines, and th eir delivery may lack the perform ative force dem anded by a m ature audience o f serious dram a. T here is also th e appeal to the playw right th a t his o r h er w ork will n o t be tam pered w ith extensively by haughty actors. I t is often n o te d th a t the great playw right C hikam atsu M onzaem on, w ho w rote m any o f th e great dram as o f the Japanese p u p p e t th eater, m uch preferred this m edium . H e did n o t have to c o n te n d w ith the extravagant egos o f actors insisting on their ow n in terp retatio n s o f characters. C hikam atsu’s alleged attitu d e, perhaps apocryphal, is p a rt o f the u n d erstanding people w orking w ith puppets have ab o u t the appeal o f th eir work. T he a rt o f the classical p u p p e t th eater in Japan consists o f three in te r­ dep en d en t arts: dram atic recitation o f the jb ru ri, o r ballad, by a tayii (chanter), sham isen accom panim ent (shamisen hiki), and p u p p e t m an ip u ­ lation ( ayatsuri) by a pu p p eteer ( ningyo tsukai). W ithin the a rt o f joruri recitation, there is a level o f artistic force called hakusei (w hite voice). This term refers to th e c h a n te r’s voice th a t is said to transcend th e c h a n te r’s individual character o r personality, allow ing the spirits o f th e joruri char­ acters to appear in the perform ance. This aesthetic ideal expresses the view

T h is fox spirit ka sh in i (p u p p e t h ead) is concealed in th e h ead o f a specially crafted fem ale kashira. W hen a strin g is p ulled o n th e kashira, this fox appears.

T h is d e m o n h ead ( g a b u ) w orks th e sam e w ay as th e fox sp irit kashira.

IN THE SHAPE OF A PERSON

25

The child character from Keisei Awa no Naruto:Junrei no Dan at the Awaji Theater. Using puppets rather than human actors makes it possible to give major dramatic roles to child characters, as the lines are recited by the tayu (chanter). This puppet is manipulated by three people.

that sometimes the personality of a particular artist can be a hindrance to the depth of a performance, and the artist's transparency (if not outright invisibility) becomes necessary. Last, a theater with puppets can present horror in a way that is at once abstract and concrete. In the Japanese theater, examples abound of what can only be considered scenes of gratuitous violence. Dismemberment, with body parts being thrown about as a jilted lover flies into a murderous rage, decapitation, and the grizzly scenes of warriors hungry for blood on the battlefield were often presented on the puppet stage . They were popular partly because they allowed for a cathartic release of tensions in a society where civil war was a recent part of its history. According to a convention developed in the theater, those in the audience who found particularly violent scenes upsetting would avert their eyes from the stage, look toward the tayu's platform, and watch his face while he chanted the lines. Tayu came to use these scenes, during which they receive much attention from the audience, to further their own careers. This convention

developed from th e realization th a t staging violence w ith puppets can be highly affecting. In the same vein, w orking w ith puppets m ay allow for a distancing from the realities o f violence and horror. Precisely because p u ppets are n o t ac­ tual hum an bodies, they allow for an exploration o f the realities o f h o rro r and violence and for an abstract consideration o f th e basic ideas o f finitu d e, aversion, and tolerance. R ecent avant-garde th eater in L o n d o n and N ew York has explored the use o f exacting realism using hum an actors. T h e directors o f these pieces discuss their w ork differently, b u t they seek to make the audience experi­ ence a perform ance as if they w ere actually p resent at a m urder, a rape, o r even a mass killing. People atten d in g these perform ances have o ften found them to be so u p setting th a t they have required counseling. F o r the direc­ tors, this is evidence o f th e success o f their p roductions. G o o d theater, they m aintain, should tear asunder the boundaries people erect betw een a rt and life. T h e use o f p u p p etry in scenes o f violence argues for a diam etrically opposite n o tio n o f theatrical experience. U sing puppets as a device makes th e scenes o f violence less im m ediate, allow ing people to explore their em otions w ith o u t feeling as if they they w ere them selves th e victims— to have experiences w ith o u t feeling overw helm ed. In sh o rt, th e p u p p e t th eater enables staging a full range o f im aginative experiences, which m ay exceed the boundaries o f actuality, possibility, and even tolerance. W ith th e advent o f film and special effects, it is tru e , hum an actors have a new freedom ; p u p p etry n o longer has the great ad ­ vantage it did d u ring th e m any m illennia before the invention o f celluloid.

A H u m an Being as an A ll-Pow erful God It has also been suggested th at p a rt o f the urge to create and use puppets arises from the desire to becom e the m aster o f o n e ’s own destiny, even if only w ithin the confines o f a perform ance space. T h e p u p p et theater d irector and playw right A ndrei Sokoloff w rote: “ O u t o f an urge for the artistic freedom o f his creative will, m an invented puppetry. By this inven­ tio n , m an freed him self from his belief in an inescapable fate; he created a w orld o f figures determ ined by him self and thus consolidated his will, his logic and his aesthetic— in sh o rt, he became a little god ruling his own w o rld .”5 Ironically, th e prom inence o f tragedies in the p u p p e t th e a te r in the end tells us th a t even w hen given the chance to becom e G od, h u m an beings repeat G o d ’s m istakes (and, one could argue, G o d ’s sadism ). W hat kind o f god w ould drive its creations to th e tragic ends often staged in pup-

petry? P e rh a p s th r o u g h th e iro n y o f th e p u p p e t, th is q u e s tio n o f th e o d ic y is re p e a te d ly b e in g asked a t so m e level. T h e classical Jap an ese p u p p e t th e a te r, B u n ra k u , is fam o u s fo r its e la b o ­ rate p u p p e ts , e ach m a n ip u la te d by th re e p e o p le . T h is style o f m a n ip u la ­ tio n , d e v e lo p e d in th e late s e v e n te e n th cen tu ry , w as a resp o n se to th e d e m a n d s o f th e a te r au d ien ces w h o w a n te d m o re stylized a n d su b tle m o v e ­ m e n t. T h e h ey d ay o f p u p p e try in Ja p an c o in c id e d w ith a p e rio d o f rig id social c o n tro l by a to p -h e a v y m ilita ry g o v e rn m e n t, w h ic h literally s trip p e d away m o st social fre e d o m s. I t is te m p tin g to a rg u e th a t th e p o p u la rity o f p u p p e try d u r in g th is p e rio d w as in p a rt d u e to th e id e n tific a tio n m any au d ien ces felt w ith th e w o o d a n d la c q u e re d b ein g s o n th e stag e. J u s t like th e a u d ie n c e s, th ese p u p p e ts w ere m o v e d th r o u g h trag ic ev en ts by s h a d ­ ow y o r invisible fo rc e s.6 T h e re is an aw areness in a th e a te r w ith p u p p e ts th a t o n e is w itn essin g a m e ta p h o r fo r real life o n th e stag e, a n d in th e case o f T o k u g a w a Ja p a n , th is m e ta p h o r c o u ld have b e e n o n e o f m axim al social re stric tio n a n d m a n ip u la tio n o f th e p e rs o n in society.

The T riu m ph o f L ife over D eath in C reation P e rh ap s o n e o f th e m o s t c o m p e llin g rea so n s w hy p u p p e t th e a te r is so a p ­ p ealin g is th e h u m a n n e e d to see a n d p a rtic ip a te in th e reversal o f d e a th . W h at is d e a th b u t th e tra n s fo rm a tio n o f th e an im a te in to th e inanim ate? In p u p p e try , th is d ire c tio n is reversed. T h is p o in t w as m ad e by th e d ire c ­ to r o f th e B u d a p e st P u p p e t T h e a tre , D e z so Szilagyi, w h e n h e w ro te: T h e a u d ie n c e a t a p u p p e t sh o w w itn esses a c tio n w h ich satisfies an u rg e p re se n t since tim e im m e m o ria l. O n th e p u p p e t sta g e , b e fo re th e s p e c ta to r’s eyes, th e su p rem e act o f c re a tio n is ta k in g p lace— lifeless d e a d m a tte r is tu rn e d in to life. In his o w n activity m a n , as a ru le , achieves th e o p p o site . In o rd e r to create a n y th in g , h e has to w a tc h p a r t o f his living e n v iro n m e n t su ffer d e a th . T o c lo th e him self, to m ake a chair, to p u t his ideas d o w n o n p a p e r, to re p re se n t th e w o rld w ith b ru s h a n d p a in t— to d o any o f th e se th in g s, h e m u st tu rn living o rg an ism s in to lifeless m in erals. A t th e sam e tim e his y e a rn in g a n d w ish to c re ate life are in fact far s tro n g e r th a n th e c o m p u lsio n o f his d e stru c tiv e in stin c t. T h is creative u rg e is tra n s la te d in to o th e r sp h eres a n d satisfied by th e p u p p e t b o u g h t to life.7

M y th s a b o u t th e o rig in s o f p u p p e try reflect th is m o tiv a tio n . O n e o f th e m yths fro m th e C h in e se sh a d o w p u p p e t th e a te r n a rra te s th e o rig in s o f p u p p e try as a m agical re sp o n se to th e iro n clad fin itu d e o f d e a th . H e r e is th e s to r y :8 A n e m p e ro r a n d his b e lo v e d w ife w ere d eep ly in love a n d s p e n t m any h o u rs in each o th e r ’s com pany. O n e day, th e w ife fell ill. H e r h u s b a n d te n d e d h e r day a n d n ig h t, b u t a s h o rt tim e later, she d ie d . T h e e m p e ro r

28

CHAPTER 1

could not be consoled. Day after day and night after night he grieved the loss of his love and companion. H e grew pale and thin. H e was despondent. Affairs of state were neglected. O n e day, a sorcerer arrived at court and told the emperor that he could bring his wife back. T h e emperor's heart quickened. "What must I do t o have her back?" he implored. "You must promise me three things," responded the sorcerer. "I must have a c o t t o n screen, a lamp, and your promise that you will stay on your side of the screen. You must never try to look at your wife's face." T h e emperor agreed. And so, with the help of the sorcerer, the loving wife returned t o her husband night after night. They spent long hours in conversation— the emperor on one side of the screen and his wife on the other. They discussed their love for one another, events of the day, and their past and future. T h e emperor was at last consoled. His strength returned. H e was once again able to see to the duties of his office. But one night, in a moment of passion, the emperor could contain himself n o longer. H e rushed to the screen and tore it down. To his great h o r r o r and anguish, he f o u n d n o t his wife b u t the sorcerer. N o w he discovered the truth: T h e sorcerer had been imitating his wife's voice all along during those long and intimate nights, and casting her shadow from the lamp on the screen with a leather p u p p e t intricately carved in the likeness of her profile. T h e Chinese sources provide two endings t o the story. They are always presented together, and the reader can choose which one is m o r e fulfilling. In one ending, the emperor is so outraged that he orders his guards t o put the sorcerer to death. T h e other ending tells us that the emperor is so grateful t o have his wife back, even if only "behind the screen," that he p r o m o t e s the sorcerer t o marshal and gives him a fortune. And this, so the sources tell us, is the origin of Chinese shadow p u p p e t theater. Even today the cotton screen in the theater is commonly called " t h e veil of death." This myth reveals a potential in p u p p e t r y t o enact symbolically one of the deepest (and most hopeless) of h u m a n yearnings: that the finality of death can be u n d o n e . By insisting o n the double perspective t h r o u g h both endings, the story suggests the h o p e that puppetry can magically accomplish the impossible, provided one can live with the illusion.

Puppetry and the Power of

Generalization

Another appeal of puppets is their ability to stay what they are: puppets d o n o t have real lives offstage. An audience may have t o suspend its disbelief t o transform a w o o d e n p u p p e t into a character, but that is all. I t does n o t have t o suspend its awareness that the stage character has an offstage life, as is true for a h u m a n actor. While t h e transformation of a person into a

IN T H E S H A P E OF A P E R S O N

29

stage character is magical, there is always the sense of "representation," or worse, a sort of " p r e t e n d i n g " inherent in role-playing. Craig made this point most dramatically at the turn of the century. Appalled by what he t h o u g h t was the decay of the British theater t h r o u g h an obsession with the particular talents of individual artists over the totality of theater as an art form, he called for a radical reinvention of theater. His discussion was intended to focus attention on the problem of interpretation and what he called a debased realism r u n n i n g amuck in the theater and ruining performance. At the heart of his argument was the (hugely hyperbolic and rhetorical) insistence that actors must be replaced by puppets, a stance that so radically distanced him from the theater that he was never able to reclaim a place in the very art form he sought t o revitalize. His championing of the puppet as an "ideal actor" is revealing of a certain aesthetic sensibility: As I have written elsewhere, the theater will continue its growth and actors will continue for some years to hinder its development. But I see a loophole by which in time actors can escape from the bondage they are in. They must create for themselves a new form of acting, consisting for the main part of symbolic gesture. Today they impersonate and interpret; tomorrow they must represent and interpret; and on the third day, they must create. By this means style may return. Today the actor impersonates a certain being. He cries to the audience, "Watch me! I am now pretending to be so and so, and I am now pretending to do so and so," and then he proceeds to imitate as exactly as possible, that which he has announced he will indicate. For instance, he is Romeo. He tells the audience that he is in love, and he proceeds to show it, by kissing Juliet. This, it is claimed, is a work of art; it is claimed for this that it is an intelligent way of suggesting thought. . . . The actor looks upon life as a photo-machine looks upon life; and he attempts to make a picture to rival a photograph. He never dreams of his art as being such an art for instance as music. He tries to reproduce Nature. As I have said, the best he can do when he wants to catch and convey the poetry of a kiss, the heat of a fight, or the calm of death, is to slavishly copy, photographically—he kisses—he fights—he lies back and mimics death—and when you think of it, is not all this dreadfully stupid? . . . The actor must go, and in his place comes the inanimate figure—the Uber-marionette, we may call him, until he has won for himself a better name. . . . Today in his least happy period many people have come to regard him as rather a superior doll—and to think he has developed from the doll. This is correct. He is a descendent of the stone images of the old temples—he is today a rather degenerate form of a God. Always the close friend of children, he still knows how to select and attract his devotees. When anyone designs a puppet on paper, he draws a stiff and comic looking thing. Such a one has not even perceived what is contained in the idea which we

now call M arionette. H e m istakes gravity o f face and calm ness o f body for blank stupidity and angu lar deform ity. Yet even m o d ern p u p p ets are ex trao rd in ary th ings. T h e applause m ay th u n d e r o r dribble, th eir hearts b eat n o faster, n o slow er, th e ir signals d o n o t grow h u rrie d o r confused; an d th o u g h d ren ch ed in a to rre n t o f b o u q u ets and love, th e face o f th e leading lady rem ains as so lem n , as beautiful and as rem o te as ever. T h ere is so m e th in g m ore th a n th e flash o f genius in th e m a rio n ette , an d th e re is so m eth in g in him m ore than th e flashiness o f displayed personality. T h e m a rio n ette appears to m e to be th e last echo o f som e no b le an d beautiful a rt o f a past civilization.9

Craig’s com m ents about the hum an stage may be overstated, b u t his insight into the potential o f the puppetry stage is valid: the puppet has a unique power o f generalization capable o f transcending the finitude o f any one person’s experience. It can portray a hum an being in general be­ cause it is in fact no single hum an being. The famous Soviet puppeteer and theater critic Sergei Obraztsov o f the State Central Puppet Theatre in M oscow echoes this sensibility: W h at is it th e n , th a t an inanim ate p u p p e t can express th a t a flesh an d b lo o d acto r cannot? W hat is its power? S trange as it may seem , its very p o w er lies in the fact th a t it is inanim ate. I f an acto r o n a stage sits d ow n in an arm chair and hitches up his tro u ser leg so th a t his knees d o n o t spoil th e crease, th e audience may well n o t notice it. B ut if th e same m ovem ent is m ade by a p u p p et, the audience is likely to b u rst in to applause because th e p u p p e t has m ade fun o f all the m en w ho m ake this m ovem ent. O n the stage, a m an may p ortray an o th e r m an, b u t he c a n n o t p o rtra y m an in general because he is a m an himself. T h e p u p p e t is n o t a m an an d for th a t very reason it can give a living portrayal o f m an in general. . . . T h e p u p p e t is a plastic generalization o f a living being: m an, rein d eer, dove. T h e p u p p e t is a sculpture, th e p u p p e te e r creates a dram atic gen eralizatio n o f a living being; m an, reindeer, dove. T h e process by w hich this in an im ate b e ­ com es anim ate seems to th e audience to be a real m iracle.10

H um an actors can certainly portray “ the hum an being in general” on stage. But what is compelling about O braztsov’s com m ent is the fact that puppets spare us from the awareness o f the representational and imitative quality o f ritual and theatrical performance.

Puppets as Vessels f o r O ther Forces

The Japanese religious use o f puppets carries the representational capacity o f puppets to a higher level. In some cases, puppets are understood to become actual abodes for visiting deities and wandering spirits. While ac-

tual hum an sacred specialists som etim es assume this role, hum an beings are able to en ter into such altered states only th ro u g h a radical break with their hum an personae. A p u p p et, however, has no o th er persona to shed and can serve as a tem porary residence for sacred forces repeatedly. Such puppets, th ro u g h their continual contact w ith sacred forces, becom e sa­ cred objects in their ow n right. T he puppets used in the Sanbaso ritual (discussed in later chapters) are an example o f these sacred objects. An exam ple from Indian m ythology, often cited as an origin tale, re­ veals the m ythical use o f puppets as vessels for supernatural forces: A long tim e ago, w hen Indian dolls were crude blocks m ade o f w ood w ith painted faces, there was a certain doll m aker w ho m ade dolls w ith m ovable limbs and lovely carved heads. T he deity Shiva and his wife Parvati loved to go to this doll m aker’s shop to see his various creations. O ne day Parvati happened to see som e beautiful dolls fashioned w ith exquisite detail. She was so enchanted by them th at she begged Shiva to allow the tw o o f them — Shiva and Parvati— to en ter the dolls and bring them to life. Shiva granted h er request, and soon the dolls were dancing ab o u t the shop, m uch to the joy o f their creator. Finally after m uch fun and dancing, both Shiva and Parvati becam e tired. They w ithdrew them selves from the dolls and were a b o u t to leave, w hen the doll m aker grabbed them by the arm and pleaded, “ H aving given life to my dolls, how can you bear to leave them lifeless and just go away?” To this Parvati responded, “Since you m ade these dolls, it is for you to give them life.” A nd th en th e deities left the doll m aker alone w ith his w ooden creations. H e was so determ ined to bring them back to life th a t he attached strings to their lim bs to help them dance. A nd so, it is to ld , the m arionette was b o rn .11 T he religious m o tif th a t this story elaborates— divine beings en ter in an ­ im ate objects and brin g them to life— is a central elem ent o f Japanese ritual puppetry. P u p p et bodies becom e spirit vessels for divine spirits, spir­ its o f th e dead, and even souls o f those still alive. N agata Kbkichi discusses this aspect o f Japanese pu p p etry as one o f th e origins o f the stro n g trad i­ tion o f using effigies in Japanese ritual: Essentially, th e hum an body is vacant. Em pty. T he M a n ’yoshu abounds with such references, indicating this early Japanese sense ab o u t th e hum an body. We find a poem th a t asks, “W ould th a t we were only able to borrow the discarded shell o f the cicada as a body.” F u rth er, the classic the K onjaku in­ cludes the lines, “the shell o f th e locust is em pty.” This em pty hum an body becom es the dynam ic lod g in g place for th e soul. In o th e r w ords, the hum an being is th a t essence inhab itin g the carcass o f the “m aterial” body. Spirit and flesh are tw o .12

A discussion o f the phenom enology o f puppets leads us im m ediately into the perception o f th e relationship betw een the body and the soul, the m aterial and the spiritual. A study o f Japanese puppets reveals various

Japanese understandings o f the spirit world and human souls ( tama-shii) and the power o f a human agent to harness and control these forces. This was a pronounced feature o f the Awaji tradition. Rituals using puppets were designed to invoke, manipulate, and contain spiritual forces in the bodies o f puppets.

In th e Shape o f a Person: M eanings o f th e Word N ingyd The Japanese word ningyd has a wide variety o f referents, and can not simply be translated as puppet. Ningyo is the sinicized reading o f a twocharacter com pound, and it can also be read hitogata when using indige­ nous Japanese pronunciation. In the context o f the Awaji tradition, ningyo is best translated as puppet, according to the above five-part defini­ tion in which a puppet is an inanimate object moved by a human agent before an audience with the intention o f communicating something through the process o f representation. In the larger context o f Japanese ritual, however, ningyo encompasses a wide range o f objects, including both static and manipulated ritual objects. H ere I present materials which argue for a Japan-specific reading o f the term. As a general category using the Chinese readings for the two characters “person” and “shape,” the word ningyd appears in the Japanese language relatively late, toward the end of the Heian period. Prior to this, all the cases considered below that can be considered examples of ningyo had spe­ cific names in their own contexts: htmiwa-, hitogata, kokeshi, kugutsu, etc. The connotations o f this general term ningyo and the phenom ena to which it refers have undergone a great deal o f transformation throughout Japanese history. A num ber o f elements, however, seem to recur. M ost apparent is the simultaneity o f ningyo and shamanic magic and the use o f ningyd as mediators between the human and divine worlds. Within this general shamanistic tendency, we see a num ber o f recurring motifs: (I) ningyb are used as spirit vessels for deities or spirits summoned by a shamanic figure; (2) ningyb are tangible references to a particular person or animal used as representational equals; (3) ningyo are understood to have the power to ward off danger when placed at the entrance to tombs or other strategic places; (4) ningyo can serve as surrogates for human beings in rites o f purification and healing; (5) ningyo are understood to be substitutes for unborn children and small infants, protecting them from epidemics and sickness, and appeasing their spirits should they reach an untimely end; (6) ningyo are used in appeasement rites to reenact the calamities that brought about the malevolent actions o f a spirit, or to serve as vessels for the malevolent spirits when they are summoned to

th e rite; (7) ningyo serve as su b stitu te bodies for possession w hen th e pow er o f the spirit being invoked is to o pow erful to e n te r a living h u m an being. H e re , I survey som e o f these uses o f ningyo in Japanese religious history.

Haniwa Figurines: Surrogacy a n d Protection O n e o f th e m o st u b iq u ito u s im ages to stare m utely at us from the h istory o f Japanese archaeology is th a t o f th e haniwa figurines. Clay effigies o f anim als (boars, dogs, birds, horses, etc.), o rn am e n te d w o m en , m usicians, and soldiers have been exhum ed from to m b s th ro u g h o u t Japan, and th e styles o f these objects vary from th e m o st sim ple and abstract to the highly detailed. As w ith o th e r data from th e archaeological reco rd , we can only sp ecu ­ late a b o u t how they were used and u n d e rsto o d . P erhaps th e haniw a m e d i­ ated for the deceased while in the to m b . M any o f th e figurines w ere fash­ ioned after w hat appear to be fem ale sham ans.13 T hey w ear th e regalia o f possession seen in later religious practice and appear to be in a trance. Perhaps these sham anic figures w ere u n d e rsto o d to serve as po in ts o f c o n ­ n ectio n betw een th e living and th e dead. Perhaps they w ere ancient guardians, em b o d y in g th e sacred pow ers to w ard o ff evil from th e to m b . P erhaps th e presence o f haniw a suggests a co n cep tio n o f surrogacy— th e figurine to o k th e place o f an actual person in the to m b . W ives, children, and even pets a n d dw ellings m ay have follow ed the deceased to th e o th e r w orld in th e form o f an effigy. These th re e possibilities— incarnate m edium ship, guardianship, and surrogacy— reveal a pow erful and m agical u n ­ d erstan d in g o f rep resen tatio n . T h e rep resen ted form participates in the pow er o f th a t w hich it represents.

R itu a l Substitutes a n d the R em oval of R itu a l Pollution: Hitogata fro m the N a ra Period T h e c o n cern w ith ritual pu rity has been a rem arkable feature o f Japanese religiosity since th e earliest records describing the Japanese people. D eath , illness, elim ination o f bodily fluids, b lo o d , pus, ch ild b irth , and m en stru a ­ tio n w ere, and to som e ex ten t still are, considered to be p o llu ted states req u irin g purification rites. T h e state o f p o llu tio n was a difficult issue th a t n eed ed to be sym bolically expressed and ritually resolved. I t was in this c o n te x t th a t a n o th e r use o f effigies arose. T h e use o f som e so rt o f su b stitu te to rem ove p o llu tio n from persons, hom es, spaces, and even th e head o f the n atio n is a d o m in a n t th em e in

JiingyS from all over Japan. An effigy or body substitute was un d ersto o d to be capable o f attracting dangerous forces and even becom ing possessed by them . A ningyo could literally absorb pollution. W hen the ningyo was destroyed o r allowed to float away, it carried w ith it the evil spirit or p o llu ­ tio n from the person against whose body it had been ru b b ed o r in w hose h o n o r it was created. Such an u n d erstanding o f the ritual efficacy o f the represented hum an form depends on three ideas: a represented form par­ ticipates in the reality o f th a t which it represents and vice versa; inanim ate m atter can be possessed by spiritual forces; and pollution can be absorbed (and subsequently rem oved) th ro u g h contact w ith the p olluted source. T he study o f itinerant puppetry rites in later chapters show how this reli­ gious idea developed. O ne o f the m ost p ro n o u n ced uses o f h um an representation in Japan began in the N ara imperial co u rt and continu ed up th ro u g h the M uromachi period. A ccording to docum ents scattered over several hundred years, there was a m onthly practice using a kind o f ningyd called nctdemono, from the characters m eaning “ a thing one caresses o r pets,” or bitosata , the characters m eaning “in the shape o f a h u m an .” T he cere­ m ony was referred to as nanase o-haraiy or “purification at the seven shal­ low s.” 14 O n the first n ight o f each m o n th , according to the lunar calendar, a Taoist diviner ( onmyoji) w ould m ake a small doll, which w ould th en be sent to the im perial court. H ere, th e ladies o f the co u rt w ould adm ire, fondle, and dress it in fancy costum e. It w ould be presented to the em ­ peror, w ho w ould th en ru b it all over his body. T he doll w ould absorb all o f the em peror’s ritual pollution (kegare) and various and assorted sins and transgressions (waza-wai). N ext, the doll was placed in a special box and floated dow nriver to the sea, taking w ith it the assorted pollutions. T here were said to be seven locations from w hich the effigy could be sent to the sea, hence the nam e nanase o-hara-i.lh Yamagami Izu m o , in his article ‘“ N anase no harai’ n o genryu” (The O rigins o f the Seven Shallows Purification) provides a th o ro u g h evalua­ tion o f the extant sources th a t m en tio n the rite .16 H is reading o f the m ate ­ rials suggests an o th er interp retatio n w hich has a bearing o n o u r u n d e r­ standing o f the use o f an effigy in this rite. H e m aintains that a p art o f the em p ero r’s soul was actually externalized and entered the effigy. This “ externalized spirit” (jfaikon) was sealed in a box and carried to the sea­ shore, w here it was purified. T he ritual thus enacted a so rt o f spirit pacification th ro u g h the em peror’s detached spirit. This in terp retatio n , although n o t conclusive, suggests th a t rather than understanding the ningyo as only a body substitute, it should be un d ersto o d as a vessel for a detached spirit. Yamagami sees this rite as evidence for the belief in the detachability o f spirits d uring the ancient period in Japan.17 T he ability o f an effigy to serve b o th as a substitute and as a spirit vessel

is reflected in o th er examples from early Japan. Like the practice o f placing haniwa o f family m em bers in tom bs m entioned above, wives, children and siblings o f a deceased noble w ould place effigies o f themselves, called hito­ g a ta in the coffin o f the deceased as a way o f accom panying their beloved beyond the grave. These effigies w ere understood to contain part o f a living person’s spirit, detached from the person’s hum an form and able to travel w ith a dead loved one to the w orld beyond the grave.18 N agata Kokichi has suggested th a t hitogata were also used to curse peo p le.19 Since m ost hitogata were m ade o f paper, straw, or w ood, the survival o f ancient examples is m ost rare. In the Im perial Treasury in Nara, there is one ancient example from the N ara period which seems to be a hitogata. It is a simple w ooden stick-shaped person, w ith a head, legs, and a torso, ab o u t fifteen centim eters high, upon w hich is painted a crude face.20 It is believed th at this figure, and types like it, were used in magical curses.21 These uses suggest that effigies n o t only served as ritual substi­ tutes b u t also had the pow er to capture a living hum an spirit and take it to the afterw orld. A lthough we have few surviving examples o f hitogata, the practice seems to have been w idespread. References to o num erous to m ention from ancient Japanese literature tell o f people m aking effigies o f friends and loved ones and o f themselves.

A m agatsu an d H o k o (B oko) Beginning in H eian Japan and continuing until the present, various dolls o r effigies have been widely used as substitutes for fetuses, infants, and children to pro tect them from evil influences and disease. A reference in The Tale o f Genji notes in passing th at w hen the child o f G enji’s lover, Lady Akashi, is taken from her so th a t she can be raised in th e capital, her am agatsu (heavenly infants) are p u t into the carriage w ith her.22 Am agatsu were dolls or effigies used to pro tect children by serving as their ritual substitutes. Som etim es they were called hdko ( boko). They are also generically called o-san ningyo (birthing dolls). Sources m entioning these dolls appear in the H eian period b u t becom e m ore num erous during the M urom achi period. A text entitled “Ise ke hisho tanjo no ki” (Secret chronicles o f the births in the Ise family), records that at the birth o f an infant, the doll was dressed simply, taken to the shrine, offered on the altar, and then prayed to for the health and longevity o f the child. Hoko are simple stuffed dolls, shaped like an infant. A ccording to the ningyo scholar Kitam ura T etsuro, although this text refers to these dolls by an­ o th er nam e, the description is o f objects from the m ore general category o f am agatsu or hoko. H e further suggests th at the m ain difference is that

This am agatsu from M orioka-shi in Iw ate Prefecture is an example o f an effigy used to pro tect children by serving as a spirit vessel for their souls.

the am agatsu is a spirit substitute (katashiro) for a boy, while the latter is for a girl.23 This practice, th o u g h considered only o f historical interest to Kitam ura, seems to have survived until recently in Iw ate-ken. O n a research trip to M orioka in February 1989, I was shown tw o h5ko by a w om an w ho ran the inn w here I stayed. She said they had been in her family for generations and were passed dow n matrilineally. She also had the old clothes used to dress the dolls. Apparently, as soon as a w om an was p reg ­ nant, a doll was set up on the family altar, unclothed (like the u nborn child). W hen the child was born, the doll was dressed and taken to the shrine, w here it was presented to the tutelary deities o f the family (ujig a m t). I t was believed, she told m e, th at the ningyo w ould confuse super­ natural forces who m ight com e to take the child. For example, epidem ic spirits (ekibyoshin) w ould com e and possess the effigy, sparing the child an untim ely death.24

K o k e sh i as Effigy

O ne o f the best know n Japanese folk arts in the west is the kokeshi, the alm ost shapeless cylinders m ade o f w ood on w hich are painted faces, cos­ tum es, and various designs, m ade in the rugged and p oorer northeastern p art o f Japan (T ohoku D istrict). Kokeshi have becom e one o f the m ost

standard tourist items in Japan and are found in gift shops th ro u g h o u t the country. T he diverse styles o f T dhoku kokeshi can be classified into ten m ajor types, each w ith its own characteristics and shapes.25 T he m ost p ro ­ nounced aspect o f kokeshi, which range in size from a few centim eters to nearly a m eter, is their shapelessness. AU over the w orld, in agricultural societies w here famine was a real threat, to o large a family often m eant th at some family m em bers w ould n o t survive, whereas a well-spaced and manageable size ensured that a family w ould n o t only survive but do well. In Japan, family size was co n ­ trolled th ro u g h a practice called ma-biki (thinning or pulling o f seedlings). This was an infanticide carried o u t by a midwife, w ho “retu rn ed ” the child to the other world, perhaps to be reborn again at a later tim e.26 O ne Tokugawa period opponent o f this practice w riting in the early nineteenth century estim ated th at 60,000 to 70,000 infants were “retu rn ed ” in M utsu and Dewa Provinces alone, tw o areas w here the grow ing season is short and families tended to be exceedingly poor.27 It is possible that the practice o f carving these cylindrical objects d e­ rived from the need to som ehow concretize the grief and guilt families felt when faced w ith the choice betw een an unw anted child’s slow starvation and it’s quick death.28 Kokeshi, lightly decorated w ith painted flowers and serene and undeveloped facial features, may have served as ritual substi­ tutes for the dead child. The am orphous quality o f these figures is one o f their outstanding characteristics, suggesting the formlessness o f the u n ­ born or newly born. A nother interpretation o f kokeshi holds th a t they resem ble phalluses and th at they were perhaps used as fertility symbols to invoke the deities responsible for a good harvest. Today they are carved as art objects, and are bought and sold w ithout any religious connection.

MizuJco Jizo: Effigies of the A borted Fetus I f objects said to be “in the shape o f the h u m an ” are indeed used to co n ­ cretize inner em otions or spiritual states, it is likely th at effigies will be used in situations that give rise to difficult and often conflicting em otions. O ne o f the m ost extensive uses o f effigies in Japan today is the practice o f m izuko kuyo (rites for aborted o r miscarried children). W hile in the U nited States and o ther W estern countries a miscarried or intentionally aborted fetus is rarely the object o f burial and m ourning rites, in Japan there is a practice o f using small stone m izuko jizo (water child Jiz5, (“water child” referring to a child in utero) to represent the dead fetus.29 These small effigies o f the bodhisattva Jizo (Ksitigharba in Sanskrit) are purchased at tem ples after one has had a funerary service conducted for

an ab o rted fetus. T h e m izuko jizo is th en placed in a specific sp o t in the tem ple gro u n d s and serves as a m o u rn in g site for th e parents. M any tem ples th ro u g h o u t Japan have a small place o f w orship dedicated to m izuko jizo, and recently this service has developed as a business for m any tem ples— a practice th a t has been th e object o f o u tcry by B uddhist clergy w ho feel th a t to advertise for clients is to take advantage o f p eo p le’s em otions. T h e H ase K annon in Kam akura, a fam ous K annon tem ple in eastern Japan and th e fifth site on th e B anto Pilgrim age ro u te , is the m ost c o m ­ m only visited m izuko jizo site. H e re , the steps leading up to th e m ain w orship hall are covered w ith tho u san d s o f stone statues o f Jizo, th e aver­ age size being a b o u t fifteen centim eters. These small statues are decorated and dressed w ith baby clothes and bibs, and m ost o f th em have offerings placed in fro n t o f them : food, juice, snacks th a t children enjoy, pacifiers, rattles, or toys— pinw heels and cupie dolls are very com m on. H ere and th ere one sees a bib w ith a m o th e r’s w ords to her u n b o rn child w ritten o n it, frequently apologizing, o r asking for forgiveness. T he votive offerings (ema) hanging at th e Inari w orship hall in th e com plex also express such sentim ents. Som e o f the stone statues have small piles o f baby shoes and baby clothes in fro n t o f them . O n e can n o t go to H ase K annon w ith o u t being deeply m oved by th e h u m an grief concretized in the thousands o f stone m izuko jizo. This practice, seen all over Japan, gives us som e additional u n d e rsta n d ­ ing o f th e Japanese perception o f things “ in th e shape o f the h u m an :” First, we see th e c o n tin u atio n in to the p resent o f the p a tte rn o f am agatsu, nam ely th a t a doll o r p u p p e t can becom e the spirit vessel for a child, in this case an u n b o rn child. Second, m izuko jiz5 indicate th e role o f dolls, h o w ­ ever unconscious, in taking away p o llu tio n o r “sin,” in this case the tra n s­ gression o f having an a b o rtio n or a m iscarriage. Like th e ritual objects in the nana.se. bar at from th e ancient perio d , these jizo statues rem ove the pollution and are th en placed in a special zone outside th e profane w orld in w hich people dwell— in th e precincts o f th e tem ple. T h ird , these small statues provide a m o u rn in g site for Japanese w om en, w here they can grieve the child they have lost, a place w here an inner process can becom e externalized and ritually resolved over tim e. T he small stones becom e a so rt o f grave m arker to w hich th e w om en can re tu rn periodically to make offerings and recite prayers. W hen one m akes a visit to H ase K annon, one notices th a t som e o f the m izuko jiz5 are being carefully ten d e d by som e w om an o r family, while m any oth ers appear to have accom plished their task o f healing, and now their bibs and toys fade in th e w eather and fall apart, leaving only th e slowly eroding stone effigy standing in th e m idst o f h u n d red s o f thousands o f others just like it.

Stone an d Paper Bodies o f Bodhisattvas: H ealin g R ites a t the Togenuki Jizo In the tem ple o f th e T ogenuki Jizo (the “thorn -p u llin g Jizo ,” so nam ed because this bodhisattva pulls o u t the th o rn s o f hum an suffering) in the Sugam o district o f Tokyo, th ere is a large festival every spring around the end o f May. T he festival is in tended for healing, and m any ill people come on this day from as far away as H okkaido and Kyushu. Shugendo practitio ­ ners, people selling dried snake pow der (m ade on the spot by p u ttin g cured and dried snakes in a C uisinart), fortune-tellers, and palm readers set up b o o th s inside o r adjacent to the tem ple precincts. T he m ajor attrac­ tion o f the shrine is the Arai K annon, o r “w ashing K annon.” K annon (Avalokitesvara in Sanskrit) is th e bodhisattva w ho (according to th e Lotus Sutra) regards the cries o f the suffering w orld, and can perform miracles for those w ho call u p o n this nam e. A t Togenuki Jizo, a p etitioner takes a small brush and washes the part o f a stone statue o f K annon corresponding to the p art o f the body he o r she wishes to have healed. People stand in line for several hours to scrub the body o f the bodhisattva, w hose body is now featureless, scrubbed sm o o th by so m any petitioners and prayers. T h e operative reli­ gious idea in this practice is th a t the body o f the bodhisattva in effigy is th e body o f th e devotee. T he act o f w ashing th e effigy is an act o f pious devotion, perform ed while one recites the K annon chapter o f the Lotus Sutra. T he m erit o f this action can be transferred to the practitioner in the form o f a healing. A t the same tem ple people also swallow small paper Jizo, said to help in curing cancer and o th er illnesses. A fter reciting a n u m b er o f sutras, the ill person ingests a piece o f paper on which the figure o f Jiz5 is inscribed. T his practice is repeated daily for several weeks during a course o f tre a t­ m en t th at consists o f m editation, prayer, and recitation o f sutras. B oth o f these practices, Arai K annon and ingestion o f Jizo, indicate an u n d e r­ standing th at objects “in the shape o f the h u m an ” can em body pow erful transcendent forces th a t have a beneficial effect on the hum an body.

Im ita tive Magic: Sexual Puppets o f the Island of Sado T he island o f Sado, located in the Sea o f Japan, is ab o u t tw o hours by ship from the p o rt o f N iigata. U ntil the end o f the Tokugaw a period, Sado was used as a penal colony. M any im p o rtan t figures in Japanese history were exiled there, including the B uddhist figure N ichiren and the dram atist

40

CHAPTER 1

Kashira from the island ofSado, in tbe collection of Yamamoto Sbunosuke.

Zeami . The island is also famous for its mines, where prisoners were forced to work, sometimes to death . Another feature distinguishing this island is the prevalence of ritual and dramatic performances using puppetry. One of the most unusual features of Sa do puppetry is the prevalence of explicit genitalia on the puppets. The sexually explicit nature of Sa do puppetry comes in two forms : the anatomically correct male and female puppets, about which little is known, and the bawdy humor of the Noroma ningy6. 30 The former, no longer presented in performances, were originally used in a ritual in local shrines on Sado during the spring planting. According to one elderly man who was active in the local miyaza (shrine guild) when he was a child, the puppets were used in a ritual conducted by the local Shinto priest at night in the main inner sanctuary of the shrine. No one witnessed the rite, although the naked puppets were displayed to people gathered at the shrine for the festivities. I was able to photograph an old pair of puppets with genitalia when I visited the home of Yamamoto Shunosuke, the leading authority on Sado puppetry.31 "Noroma ningy6" designates four puppets used together to present short, humorous skits, usually derived from kyogen (skits performed in

IN THE SHAPE OF A PERSON

41

Male and female puppets from the island of Sa do, in the collection of Yamamoto Shunosuke.

Noh). These pieces are presented as light entertainment between long dramatic performances. Noroma ningyo have made Sado puppetry famous, even though they were intended as minor attractions in longer puppetry events. There are numerous plots used in Noroma pieces, but unfailingly at the end of each piece, the protagonist of the Noroma ningyo, Kinosuke, opens his clothes, extends his rather sizable penis, and urinates on the audience. The "urine" is water, and it is blown through a hole carved into his paulownia-wood penis by the puppeteer. This final scene usually has nothing intrinsically to do with the performance of a humorous kyogen to which it is attached, and sometimes it gets introduced by the character simply announcing near the end of the piece, "I need a piss." But it is this risque moment that has made the Noroma skits so popular. Concerning this aspect of the Noroma ningyo, Yamamoto Shiinosuke writes: "It is said that if showered with this urine, barren women will be blessed with children. It is also used to shower newlywed brides. This type of gesture is one example of a felici tous rite for the blessing of descendants and prosperous crops seen all over Japan.,,32 I saw a day-long performance of puppetry on Sado in July of 1991. It was presented at a senior citizens center, and everyone in the audience was

42

CHAPTER I

Four puppets used in Noroma ningyo. Notice that the main puppet is in his underwear. In many plays, he exposes himself and urinates on the audience through a large penis.

retired. In the course of the day, we saw two long joruri pieces, but four Noroma skits. Since the joruri pieces were all tragedies, the humor of the peeing Kinosuke gave everyone some comic relief, which perhaps left them more receptive to the profundity of the longer plays. Sexual effigies used in fertility rites are a common feature of Tohoku shrine practice, and such dolls and effigies can be seen all over northeastern Japan. The Sado puppets are most likely a Sado expression of a larger regional practice. Perhaps Noroma ningyo's pronounced use ofa "puppet with a penis" gag at the end ofa performance grew out of the larger ritual use of sex puppets. The Noroma ningyo may be the popular and "dramatic" expression of this ritual form of imitated sexual intercourse. It seems obvious that the exposure of genitals in a ritual setting would in some way indicate a concern with fertility. But what other meanings

This w o oden penis, a b o u t fo u r inches lo n g , is used in th e N o ro m a perfor­ m ance. It has a tu b e in it, th ro u g h w hich th e p u p p et sprays w ater ( “u ri­ n ates” ) o n the audience.

could be attach ed to such a gesture? I f we consider the m o st fam ous case o f “ genital flashing” in Japanese m ythology— the dance perform ed by A m e -n o -U z u m e -n o -M ik o to before the heavenly cave d o o r in o rd e r to draw th e goddess A m aterasu o u t o f h er concealm ent— we can tentatively suggest a n o th e r level o f m eaning to this act. A ccording to a Japanese m yth, reco u n ted in B ook O n e , C h a p te r 17 o f th e K o jiki133 the sun goddess concealed herself in a cave because h e r u n ­ ruly b ro th e r Susanoo defiled h er heavenly weaving cham ber by defecating on th e walls and tearing dow n th e dikes betw een th e rice paddies. W hen she tried to placate him th ro u g h gentle speech, he defiled her fu rth er by th ro w in g a flayed heavenly piebald colt on her th ro u g h an o p e n in g in the ceiling while she was weaving. In shock, she struck h e r genitals w ith the sh u ttle o f her loom and died. T h e re u p o n , she w en t in to a cave, and the Plain o f H ig h H eaven was p lu n g ed in to darkness. To lure h er o u t, elab o ­ rate preparations w ere m ade by th e o th e r deities, w hich involved the crea­ tio n o f ritual im plem ents to be used in w hat is considered to be a reflection o f a sham anic rite o f spirit pacification, a chinkonsai. Finally, th e deity A m e -n o -U z u m e -n o -M ik o to tu rn e d a w o o d en bucket over and began to dance. B ecom ing m ore and m ore frenzied, she finally p u sh ed dow n her w aistband to reveal her genitals. T h e o th er deities laughed at this. H e a r­ ing this lau g h ter outside th e cave door, A m aterasu peeked o u t, and was pulled back in to th e Plain o f H ig h H eaven by o th e r deities. This particular event has been extensively in te rp re te d , b u t for th e p u r­ poses o f this discussion, tw o in terp retatio n s seem particularly sem inal.

Chiri M ashiho has suggested that the exposure o f the genitals can be used to drive away evil spirits, and he has noted a similar ritual am ong the Ainu, called hoparata-.34 M atsum ura suggests th at this gesture is an act o f entertainm ent for deities, used in rites intended for divine rather than hum an audiences.35 So, while the exposure o f genitals certainly argues for a concern w ith fertility, it may also be a rem nant o f an earlier shamanic use o f ecstatic and erotic dance to lure and entertain deities and drive away evil spirits.36 O ne obvious reason why a puppet is used on Sado as opposed to a hum an being concerns modesty. People, even those w orking for shrines, can n o t go around flashing their genitals. B ut at an o th er level, the use o f a puppet makes it clear th a t the concern is w ith fertility in an abstract sense. T he w ooden penis o f the male doll and the long m em ber o f Kinosuke are rem inders th a t these objects are symbols o f som ething else. A penis on a real hum an being is a penis. O n a puppet it is the generalized sense o f fertility, bravado, and even transgression o f social boundaries and propriety.

B odies at the E dge, B odies In-B etw een: N ingyo in th e Awaji C ontext A t the start o f this chapter, I suggested th at there may be som ething in ­ tentionally inconvenient about using puppets rather than hum an actors, and th e preceding discussion has show n th at in Japan the use o f repre­ sented hum an beings reveals a complex underlying religious structure. T he survey o f objects in th e shape o f the hum an has shown th a t far from being merely decorative item s, ningyo (under many different aliases) have served im portant religious purposes in Japanese history. Representational objects having a hum an shape were used to concretize and localize o th er­ wise abstract dim ensions o f life, providing a focal p o in t for ritual actions that explore and resolve difficult spiritual issues. T he rem ainder o f this chapter examines how the puppet body was con­ structed, m anipulated, and understood in the Awaji case, a tradition that developed in the sixteenth century. As we shall see in m ore detail later, this tradition was divided into tw o distinct aspects. T he earliest use o f puppets, called shinji (sacred m atters), and consisted o f appeasem ent rites for the deity Ebisu and purification rites for visiting deities o f agriculture. O u t o f these ritual origins, there developed a dram atic tradition, ningyo joruri. Artists in this later tradition influenced and interacted w ith perform ing arts in urban centers, and developm ents parallel to those o f classical Bunraku can also be seen. W hile the perform ance th at first stirred my interest in Awaji puppetry came from the joruri p art o f the tradition, the questions

this perform ance generated about the use o f puppets are even m ore to the p oint in the study o f ritual puppetry. In this book, as I have already noted, I am leaving aside a discussion o f the dramatic joruri tradition and am focusing solely on ritual puppetry. This focus has tw o implications for a discussion o f puppet bodies and manipulation. First, ritual puppets were lighter, smaller, and simpler than their dramatic counterparts, and so in this work we will n o t have a detailed description o f the inner workings o f a puppet head, nor o f the subtle art o f puppet m anipulation which fasci­ nates audiences o f Japanese classical puppetry and tends to dom inate dis­ cussions o f this subject. Second, m anipulation o f ritual puppets was decid­ edly less developed, which means th at the im pact o f these rites depended on other, context-specific nuances o f ritual performance. But m uch o f w hat can be said ab o u t ritual puppets can also be said o f the early joruri puppets, before the developm ent o f the three-person m anipulation. H ere we will examine the im portant com ponents o f a ritual puppet body, the m ethods o f m anipulation, and the understanding o f puppets as beings once removed from the hum an realm.

The S u m o f Its Hurts; A Puppet Body The earliest puppets o f the Awaji tradition appear to have been small, handheld stick puppets in which the head was the most im portant feature. The puppeteer presented his puppets from a box hung around his neck. This style o f carrying puppets was called kubi-kuke (hung from the neck), and the box gave puppeteers the nickname hako-muwushi. The box was open at the back and served as a sort o f stage. This m ethod o f m anipula­ tion was also seen am ong the Nishinomiya kugutsu, discussed in chapter 3. These boxes were a natural choice for an itinerant perform er, as they could serve the dual purpose o f carrying case and portable stage. T he p u p ­ pets themselves were held by inserting o n e’s hands into the pant legs o f the puppet, a m ethod o f m anipulation called sushikomi ningyo. Because these puppets were light and simple, it was possible to carry several around as one moved from place to place, and concerns about staging and special effects were to come later. By the beginning o f the eighteenth century, Awaji puppeteers had made larger puppets for their rituals, which led to more specialization in their arts and also to m ore detailed attention to the puppet body. Puppe­ teers then began to carry their puppets in boxes on their backs, which they w ould unpack before performances. U npacking puppets and making them come alive as they moved o u t o f a “sleeping position” was made a p art o f the ritual process. The m ethod o f manipulation changed slightly from the earlier sushikomi ningyo\ the puppets were held through a hole

46

CHAPTER 1

Umazume Masaru holds examples of small stick puppets. These small puppets were widely used by itinerant performers because they were lightweight and easy to carry and manipulate .

in the back of their costume. Furthermore, due to their larger size and weight, the only way they could be manipulated gracefully was to hold them at waist level. The puppeteers performed in full view of the audiences, holding the puppets in front of them. This added a new dimension to the performances. Puppets, nearly half the size of the puppeteer and held in front, appeared to be almost second selves for the puppeteer. The puppeteer held them like shields between himself and the world as he ritually controlled dangerous deities and absorbed noxious forces and pollution . These larger puppets consisted mostly of a head attached to a long stick, with a fabric body draped about it. The most important part of a puppet was its kashira, or head. Early kashira were made of paulownia or cypress wood, and features were painted onto the face. The puppets used in Sanbaso rituals, however, had moving eyes and mouths, features that were used to indicate that the puppet was going into trance and was about to be entered by a deity. Carving kashira became an art related to the practice of puppetry, and techniques for making subtle kashira were handed down from master to disciple. The most famous puppet-head carver (ningyo-shi) of the Awaji tradition was a man named Tengu Hisa, who lived in the middle of the nineteenth century. Today, kashira carved by this artist are rare and are worth hundreds of thousands of dollars. Puppet-head carvers were understood to be liberating spirits from the wood they carved. They

spoke o f them selves as midwives to the process, and o f kashira as being either “ b orn w ell” o r “n o t b orn w ell.” A w ell-born kashira was one w ith a spirit in it.37 R itual puppets have hands and feet, and objects they need to carry or hold are inserted into small holes in the hands o f the puppets. T he illusion o f a body is constructed by draping an o u ter garm ent over the assem bled body parts, w hich th en m ust be held gracefully by a skilled pu p p eteer.38 A p u ppeteer is considered a failure if the parts appear un co n n ected w hen he o r she m anipulates them (som etim es in con junction w ith o th er p u p p e ­ teers). I f the pu p p eteer is skilled, this collection o f bodily parts is tran s­ form ed in to a living, m oving, and even ecstatic being. W hen a p u p p e t is p u t dow n in a perform ance, it m ust be set dow n face first w ith its sleeves folded in front o f it in the p ro p er “sleeping” position. Body parts o f puppets com e alive th ro u g h m anipulation. C om ing alive is the key analogical event giving these p u p p et rituals their force. This transform ation o f assorted body parts into a body w ith a being inside it belies the close relationship betw een puppetry, fertility, and life and death. This relationship seems to be an issue o f ritual concern. Som etim es life does n o t happen w hen it should, and the w orld— w hether seeds or hum an w om bs or fish from the sea— rem ain nonproductive, unyielding, and dead. W ith every gesture, th e pu p p et indicates th e o th er side o f life, for as soon as the pu p p eteer puts it dow n in its p ro p er sleeping position, the p u p p et is again no m ore th an the sum o f its parts: a dead body. Display o f the inert p u p p e t body before and after the m anipulation is a p a rt o f the perform ance.

Puppets: Beings Once Removed As we have seen, th e use o f puppets rather than hum an actors raises a nu m b er o f interesting questions. In chapters to com e we shall deepen o u r un d erstanding o f this ritual preference for p uppets in the Awaji tradition. T he m ost im p o rta n t feature o f the rites u n d er discussion in this volum e is th a t sacred beings, Ebisu, the deity o f the rice fields, or any n u m b er o f assorted kami are being sum m oned and m ade m anifest in the hum an com m unity th ro u g h a ritual context. H erein lies a problem o f representa­ tion. In h ere n t in the conception o f sacred (and dangerous) beings undergirding this ritual tradition is the idea th a t the sacred can be m ade present in the hum an realm , and yet no m ode o f m anifestation can fully convey its sacredness. In som e sham anic contexts, a sham an may sum m on a sacred force to enter his o r her body. This activity creates a coincidence o f o p p o ­ sites, since the sham an’s body becom es at once hum an and divine. H ere, the p u p p e t can stand betw een th e puppeteer and the sacred forces and

serve as the receptacle for the sum m oned deity. T he p u p p et th en is the physical object w hich enables ( yosasu ) the m anifestation o f the sacred force. In these rituals, puppets are often referred to as the yorishiro (pos­ session vessel ) y Jyoshintai (deity body), o r tam ashiro (spirit vessel) o f the deity— all these term s referring to physical objects th at becom e tem porary abodes for the deity— the m ost im p o rtan t ritual use o f puppets. T h e status o f the pu p p et in the ritual events exam ined in subsequent chapters is th a t o f a bridge betw een the tw o separate b u t in terd ep en d en t w orlds o f hum an and divine beings. T he puppet, however, belongs to neither. Rem oved from the realm o f hum an beings, the pu p p et occupies a w orld o f its own. A lthough the shape o f the pu p p et is m odeled o n the hum an body, the c o n ten t o f its spirit w hen fully possessed is supplied by sacred forces. W hile there is clearly a religious awareness o f the pow er o f ningyo in the Awaji choice to use puppets rather than hum an actors, there is also an o th er reading o f th e ritual form . T he use o f puppets intensifies an im ­ p o rta n t social and religious m etaphor o f pow er, control, and in terd ep en ­ dence. T he puppeteer controls, even invents his pu p p et as he puts it t o ­ gether, and yet he is dep en d en t on it for his livelihood. T he forces that enter the puppet are greater than the puppeteer, and because o f the ability o f the pu p p et body to attract and contain these forces, the hum an co m ­ m unity is able to survive and avoid calamities. T he puppeteer and the p u ppet, w orking to g eth er in a carefully scripted ritual space, are able to control those spiritual forces. This triangular relationship o f shifting pow er relations reveals a subtle awareness o f the connection betw een hum an and divine beings. In the following chapters, we shall see how the ritual use o f puppets has posed and resolved these relationships.

2 Kadozuke: T he O utsider at the Gates

th ird day o f the first m o n th o f the year. In a small village facing the Inland Sea, an elderly m an makes his way th ro u g h the narrow streets carrying on his shoulder a small bag and o n his back a box w rapped in a black cloth. H e wears sh o rt trousers, a haori belted at the waist, and straw sandals. T he bag on his shoulder holds a small h an d d ru m and a flute, and in th e box on his back are puppets, each alm ost a m eter in len g th , w ith their costum es and im plem ents. As he walks th ro u g h the village, he stops at th e gate o f each house and blows a few sharp notes o n his flute to announce his arrival. H e is expected this tim e o f year. H e is the D okum bo-m aw ashi, the ritual pu p p eteer w ho has com e to perform a p u ­ rification ritual and u tte r felicitous w ords and blessings for th e new year. Shortly, a h ouseholder appears at the gate, greets the puppeteer, and grants him perm ission to perform . A t som e hom es the pu p p eteer is invited n o fu rth er th an the gate and m ust set up his perform ance space while still half in the street. A t others, he is invited into the courtyard o r the e n try ­ way. Som etim es, his ho st escorts him up in to the Zashiki1 the m ain room o f the house, w here the family kam idana (Shinto altar) is located. A t this house, the householder invites him up inside and then rushes ahead to inform the o thers in his family th at th e D okum bo-m aw ashi has arrived. T h e pu p p eteer pauses at the entryw ay o f house and bows to the h o u se­ holder. H e th en rem oves his sandals and steps up in to the zashiki. T he host indicates th e location o f th e kam idana and th e p u ppeteer seats him ­ self in fro n t o f it. T he host leaves the room briefly and the puppeteer spreads a cloth on the tatam i before the altar, sets dow n his packages, and carefully unpacks his musical instrum ents. H e places the box containing the puppets in fro n t o f him , w here he can easily rem ove them one at a tim e. M om entarily, the ho st reappears w ith a small tray containing a dish w ith polished rice and a cup o f sake. These he places in fro n t o f the p u p ­ peteer. T he pu p p eteer takes from his bag som e w hite paper and a pair o f scissors and cuts the paper in to zigzag strips, called Jjohei1 w hich serve as m arkers for a sacralized space. H e arranges these strips o n the tray w ith the rice and sake, lifts the tray as an offering to the deities in the kam idana, and th en sets it in fro n t o f his small perform ance space. N ow , seated carefully on the floor w ith his puppets before him , he claps his hands and bows his head to the floor in prayer before the deities I t is t h e

in the kam idana. T h en , he picks up his flute and plays a sh o rt, shrill m el­ ody to invite the deities to w atch his perform ance. H e th en takes up a drum and beats it rhythm ically for a few m om ents to set the cadence for his perform ance. W ith m eticulous care, the p uppeteer perform s his sh o rt ritual, m anipu­ lating three puppets in succession. H ad his host let him no fu rth er than the gate, he w ould have been quicker and finished in a few m inutes. B ut since he is in the m ain room o f the family hom e, he can take his tim e. To show his gratitude for th e h o n o r o f being invited inside, he gives a m ore elaborate perform ance. H is opening notes on th e flute and th e d ru m ca­ dence have attracted the dangerous spirit forces in th e house, and shortly, w hen he begins to m anipulate his puppets, these forces will be so fasci­ nated by the perform ance th at they will enter the puppets. T he puppets perform a solem n and then frenzied dance, and the pu p p eteer recites a nuanced chant full ofim ages oflongevity, tim e, and nature. T he spirits are soothed by the poetry and dancing. W hen the puppeteer finishes his ritual, he packs up his puppets and instrum ents and leaves th e house, taking w ith him the spirit forces th at have entered his puppets. As he departs, th e householder presents him w ith a token paym ent o n a tray— a few coins o r notes w rapped in paper perhaps, or a small bag o f rice. A t the doorw ay to the house and again at the gate, the puppeteer bows to th e householder and moves on to the next house. As he makes his rounds th ro u g h the village, he will n o t m eet any com ­ petitors, for ritual puppeteers have already agreed to w ork in different neighborhoods, lest they drive one an o th er o u t o f business o r annoy their patrons. Tow ard the end o f the day, a householder will offer the p u p p e­ teer a meal and lodgings for th e n ight in exchange for the perform ance. If no such offer is forthcom ing, he can always sleep in the local shrine or camp at the edge o f the village o r on the bank o f a nearby river. This routine continues every day until the close o f O -shogatsu, the N ew Year period from the 1st to the 15th o f the first m o n th . T hen the p uppeteer will head back to the Sanjo district on the island o f Awaji, w here he lives for the p art o f the year w hen he is n o t perform ing. Perhaps he will present a different kind o f pu p p et perform ance in local shrines or along the fishing docks on his way hom e. Perhaps a w edding o r a new house will need blessing along the way. M aybe next year w hen he sets forth again, he will take o n a p artn er to play the musical instrum ents and help carry his gear— some puppeteers do this. A partn er w ould provide a little com pany on his travels. B ut th en he w ould have to share his earnings, so perhaps it is better to stay on his own. A solitary D bkum bo-m aw ashi works harder b u t can keep all his earnings. Back in the village, the N ew Year period winds dow n. The D bkum bo-

mawashi has com e and gone. H is appearance, announced by the shrill flute call from the street, is as m uch p a rt o f N ew Year as the pine branches decorating th e gate and th e p o u n d e d rice cakes eaten for the first fifteen days o f the year. B ut like these m arkers o f the season, the pu p p eteer is far from everyone’s m inds for the rest o f the year. T h e year draws to a close th en and it is O-Shdgatsu again. T h e n he w ould be noticed only if he failed to appear. I t just w o u ld n ’t be N ew Year w ith o u t this visit. This vignette, based o n oral histories, old films, and w ritten sources, d e­ scribes the seasonal visit o f a ritual puppeteer, a com m on feature o f fishing and farm ing villages su rro u n d in g the Inland Sea o r scattered th ro u g h o u t central and w estern Japan from as early as the sixteenth century until ab o u t 1947 o r 1 9 4 8 .1 C o u rt diaries from the late sixteenth century n o te in passing the arrival o f these visitors at N ew Year, and people I in te r­ viewed o n Awaji rem em bered the regular visits o f the seasonal puppeteer in the early tw en tieth century. W hile itinerant p u p p e t perform ers’ styles differed, the events described above could just as easily have happened in the late sixteenth cen tu ry as in 1938. These w andering puppeteers, know n by a n u m b er o f different nam es— kugutsu, Dokum bo-m awashi, deko-mawashi, hako-mawashi, Sanbaso-mawashi, Ebisu-m awashi, Ebisukaki, or simply ningyo-mawashi2— perform ed an essential ritual function. T hey m ediated the boundaries betw een the distinct b u t som etim es over­ lapping w orlds o f sacred forces and hum an beings, ord er and chaos, life and death, and fertility and infertility. T heir ritual perform ance served to usher in the new year, purify dwellings for an o th er season, and revitalize sacred forces in com m unities. O n occasion their rituals were used to bring rain, drive away noxious insects, prevent epidem ics, and ensure safe travel. They perform ed before audiences in all levels o f society, from the em ­ peror, lords, and ladies at c o u rt to m erchants and farm ers. A study o f ritual p u p p etry in Japan reveals the boundaries o f the cosm ological convictions o f Japanese society, the structures o f social hierarchy and ritual purity, and the understandings o f the pow er o f ritual to generate m eaning, cope w ith disorder, and m ediate difference. An essential elem ent o f p u ppeteers’ ritual function was th eir status as outsiders to com m unities w here they perform ed. T h eir itinerancy and otherness were as m uch a part o f the pow er and fascination they generated as the eerie puppets they m ade dance, and as the potentially dangerous and otherw orldly beings they sum m oned to the hum an realm. A lthough ritual p uppetry has a long and com plicated history in Japa­ nese religion, one feature stands out: puppeteers and puppets shared a w orld set apart from the everyday realm . As perform ers sharing this te rri­ tory, puppeteers and puppets enjoyed a unique freedom o f m ovem ent in society and th ro u g h o u t the cosmos. They carried the responsibility for

generating o rd er o u t o f chaos and purity o u t o f defilem ent, and conse­ quently suffered from shifting assessments o f their status and w orth. In this chapter, we explore the m eanings o f the am biguous relationship o f the puppeteers to their audiences and hosts by exam ining tw o in ter­ related issues: the role o f the stranger or outsider in Japanese religious perform ances and the dom inant ritual purity system o f medieval and early m odern Japan, w hich led to the symbolic exclusion o f these ritual per­ form ers and the form ation o f outcast strata o f society, contained w ithin special areas called Sanjo districts.ffw ill show how puppeteers as outsiders w ere inhabitants o f a symbolic space reserved for ritual specialists w ho m aintained social boundaries^Precisely because th e perform ers were u n ­ know n to their audience and came from the potentially disordered w orld beyond village boundaries, they were regarded as effective ritual p ractitio­ ners. T heir otherness m eant th at they em bodied the pow er o f the u n ­ know n. T heir rituals m ade th a t pow er available. T he unknow n from w hich they came also had a location in the Japanese ritual map: special precincts attached to shrines and along the edges o f tow ns. T he genera­ tio n and m anipulation o f this symbolic authority over the o th er realm was essential for their rituals to work. This specific discussion o f Japanese ritual puppeteers as outsiders can be best understood by situating it w ithin a series o f larger intellectual co n ­ cerns. Theoretical discussions o f the O th e r as b o th a fundam ental cate­ gory in the history o f religions and as a highly nuanced cultural construct have revealed the tensions and dynamics in ritual p uppeteers’s perform ­ ances from late medieval and early m odern Japan. Follow ing a general discussion o f these dynamics, I exam ine one specific elem ent, th at o f the stranger in Japanese religions. I investigate how this concept o f the stra n g e r/o u tsid er is encoded in Japanese cosm ology, solidified in the Jap ­ anese ritual purity codes and symbolic geographies, and presented in a series o f ritual practices and understandings pertaining to the role o f a n ­ cestors in Japanese religions. I th en tu rn my atten tio n to one specific type o f ritual p uppetry— kadozuke, “rites at the g ate.” T o appreciate the itin ­ erant context o f kadozuke rites, I discuss tw o im p o rtan t itinerant perfor­ m ance traditions, Senzu munzcti and M atsu-bayashi, b o th from late H eian and medieval Japan, which can be regarded as im p o rtan t prototypes o f the later kadozuke context. In kadozuke perform ances, tw o levels o f otherness shaped how people perceived puppeteers: the puppeteers’ geographical and symbolic rela­ tionship to the people for w hom they perform ed and the p u ppets’ status as beings rem oved from the hum an realm . T he itinerant context o f these rites, the relationship betw een ho st and perform er, the perception o f p u p ­ pets as special beings in betw een tw o w orlds, and the fluid idea o f the O th e r collectively im part m eanings to these ritual actions.

T he O ther: A G eneral Issue in th e Study o f R elig io n , C ultures, and Id en tity T he category o f the O th e r in all its guises— the sacred, the unconscious, the foreigner, the outsider— has becom e a central focus o f scholarship and discourse in the tw en tieth century. At once posited to be an ontological category o f experience, a necessary device o f all know ing and percep tio n 3 and a highly constructed aspect o f identity form ation, it is possible to argue th at because o f the idea o f an o th er we have such disciplines as anthropology (d ep en d en t o n the culturally o th er to supply its data) or history o f religions (d ep en d en t on the claim th at people live their lives in som e so rt o f relationship to som e o th er o rd er o f reality). M ichael Theunissen notes the u b iquitous quality o f the category o f otherness in the opening lines o f his study o f alterity, The Other: F ew issues have exercised as p o w erfu l a h o ld ov er th e th o u g h t o f th is c e n tu ry as th a t o f “ th e o th e r.” I t is difficult to th in k o f a se c o n d th e m e , even o n e th a t m ig h t b e o f m o re su b sta n tia l significance, th a t has p ro v o k e d as w id esp re ad an in te re st as th is o n e; it is difficult to th in k o f a se co n d th e m e th a t so sh arp ly m arks o f f th e p re se n t— a d m itte d ly a p re se n t g ro w in g o u t o f th e n in e te e n th c e n tu ry an d re a c h in g b ack in to it— fro m its histo rical ro o ts in th e tra d itio n . T o b e su re, th e p ro b le m o f th e O th e r has b e e n th o u g h t th r o u g h in fo rm e r tim es a n d has at tim es been a c c o rd e d a p ro m in e n t place in eth ics a n d a n th ro p o lo g y , in legal an d political philosophy. B u t th e p ro b le m o f th e O th e r has ce rtain ly n ev er p e n e tra te d as deeply as to d a y in to th e fo u n d a tio n s o f p h ilo so p h ica l th o u g h t. I t is n o lo n g e r th e sim ple o b je c t o f a specific discipline b u t has alread y b ec o m e th e to p ic o f first p hilosophy. T h e q u e s tio n o f th e O th e r c a n n o t b e se p arate d fro m th e m o st p rim o rd ia l q u e s tio n s raised by m o d e rn th o u g h t.4

T he idea o f the o th er has been a m ajor them e in the history o f religions, anthropology, and literary criticism. W hile it is beyond the scope o f this w ork to survey the history o f this theoretical concept in nineteenth- and tw entieth-century anthropology, it is possible to delineate three general approaches to the idea o f the O ther: the perspectives o f religiosity, a n th ro ­ pology, and critical studies o f hegem onic discourse. These three ap­ proaches reveal a com plexity o f attitudes and appropriations o f this term w ithin academ ia. O n the one hand, scholars such as R u d o lf O tto and M ircea Eliade have considered the category o f a wholly o th er, distinct order o f being and m eaning, set apart from hum an life as “the sacred” (the “n u m in o u s” in O tto ’s nom enclature), to be a necessary co m p o n en t o f a religious life. F o r O tto and Eliade, the perception o f this otherness is re­ m oved from and even prior to the developm ent o f constructions o f “g o o d ” and “ evil.” This focus has m any lim itations, b u t, it rem ains th at

b o th O tto and Eliade w ere recognizing a fundam ental fact o f the data in religious studies: people talk a b o u t, organize their lives a ro u n d , and seek to m aintain relationships w ith w hat they regard as an other ord er o f m ean­ ing. W hile the historian o f religions is n o t in a position to validate or invalidate the existence o f this o th e r o rd er o f m eaning, the conviction th at it exists is a central co m ponent o f religious experience in m any religious traditions, and for this reason the historian o f religions considers it a datum to be studied and u n d e rsto o d .5 F o r the purposes o f this discussion, w hat is significant ab o u t th e category o f th e sacred (or th e num inous) is th e way it is described as an other o rd er o f reality. This very otherness becom es the focus o f religious life and the center for o rien tatio n o f m ean­ ing and action. F or scholars in the history o f religions follow ing the trad i­ tion o f O tto and Eliade, then, constructions o f otherness participate in a religious structure. Early studies o f “prim itives and savages” often n o ted the “ fear o f strangers” as a co m ponent o f the “ savage m in d .” 6 T he field o f a n th ro p o l­ ogy has w idened in the last century, however, and studies o f societies closer to hom e reveal th a t alterity (and its signification, p ro d u ctio n , and m anipulation) is a co m p o n en t o f a w ider range o f hum an activity. S trang­ ers and O thers are as m uch a p art o f the generation o f social m eanings in upstate N ew York as they are in Papua N ew G uinea. F u rth e rm o re , schol­ ars have explored how the ethnographic O th e r reflects a need w ithin w est­ ern academic anthropology to develop th ro u g h a continually renew ed contact w ith actual O thers in the field. W ithin contem porary m ainstream anthropological discourse and later literary critical studies, the O th e r as a construct and idea has been studied as a central feature o f tw o interrelated (th o u g h often com peting) types o f ideology: identity and oppression. In b o th o f these constructs, som e o u t­ sider becom es a screen against w hich a conception o f group or self-iden­ tity is constructed. T he outsider m ay be som eone living in o n e ’s m idst, as the case o f Jews, gypsies, gays and lesbians, and the disabled in the T hird Reich; or an o th er “o u t th e re ,” as in the case o f M uslims as they were im agined and invented in E uropean religious and cultural studies in the latter n in eteenth cen tu ry ;7 o r an im agined idea o f alterity infecting society, as in the case o f the Tokugaw a nativists’ invocation o f Chinese elem ents in Japanese religion, literature, and aesthetics.8 Recent scholarship on the issue o f otherness has focused on the m anip­ ulation o f the category o f the O th e r in the co nstruction o f identities. These studies reveal th a t w ho O thers are in any society varies dramatically, depending on how the issues o f identity are form ulated. For exam ple, if the concern is w ith constructing an identity o f nationality, otherness will be determ ined by o n e ’s ethnicity and roots. I f the concern is w ith a gender identity, sexual preference, behavior, and gender assignm ent determ ine

o n e ’s otherness. I f the concern is w ith co nstru cting an idea o f m ainstream identity (“ the folk,” “ the m iddle class,” etc.), rendering certain groups o f people exotic serves to reinforce th e im age o f the m ainstream .9 T h e only constant seems to be the endless need for an O th e r against w hich an id en ­ tity can be constructed. T h e social co n stru ctio n o f otherness has every­ th ing to do w ith social rules and the generation o f pow er relationships, order, and m eaning in society. O thers are often symbolically constituted by images o f disorder, chaos, danger, and th e exotic. O ne o f th e m ost hau n tin g , m enacing, and frequently cited images in studies o f otherness is F o u cau lt’s specter o f the ship o f fools, the N arrenschiff, an actual p h en o m en o n in medieval E urope em bellished in litera­ tu re and the p opular im agination o f the tim e. A N arrenschiff was a ship th at traveled from p o rt to p o rt, carrying the insane or otherw ise m entally disabled w ho had been expelled from tow ns and villages. “ They did exist, these boats th a t conveyed their insane cargo from tow n to tow n. M adm en then led an easy w andering existence. T he tow ns drove them outside their limits; they w ere allowed to w ander in th e open countryside, w hen n o t en tru sted to a group o f m erchants and pilgrim s. . . .F re q u e n tly th e y w e re handed over to boatm en. . . . I t is possible th a t these ships o f fools, w hich hau n ted the im agination o f the entire early Renaissance, were pilgrim age boats, highly symbolic cargoes o f m adm en in search o f their reason.” 10 F o u cau lt’s im age o f th e insane floating a b o u t the p o rts o f E urope and the im aginations o f the Renaissance has becom e th e w atershed case for a study o f otherness d u ring th e past th irty years. H is key p o in t is em bodied in one w ord from this passage: “ reaso n .” A lthough o f course the lack o f a fixed abode, the travel at sea, and the foreigner status o f these ships3 cargo co n trib u ted to how they were im agined, w hat was im p o rta n t (at least as F oucault im agines it) was th a t their very m adness, real or o th e r­ wise, was a perfect p o in t o f reference against w hich a E urope seeking to give birth to th e E n lightenm ent (a celebration o f reason par excellence) could define itself. T h e operative dyad in this case was betw een reason and m adness, and the social co n stru ctio n o f m adness allowed culture to m ain­ tain its allegiance to reason. In her book P u rity a n d D anger, M ary D ouglas explores an o th er level o f otherness by presenting a theoretical fram ew ork for p o llution and tabo o. D ouglas’s study marks an im p o rtan t early step in appreciating the m anipulation o f otherness in the creation o f social o rd er and the sym bol­ ism o f p o llution often attached to m arginalized people. W hile defilem ent is n o t necessarily an attrib u te o f the O ther, it is n o t unco m m o n for per­ sons w ho are m arginal, placeless, w ith o u t status, o r in any way am biguous to be considered potentially defiling. D ouglas notes the com m on te n ­ dency to view m arginal people as potentially dangerous. T his, she writes, is due to the fact th a t a m arginal person in a ritual process is u n d erstood

to be in a dangerous situation, and this danger is com m unicable to others. W hile her cases are draw n from ritual processes w here a person is only tem porarily m arginal as p art o f the m ove to a new status, her examples well describe the im agination o f people w ho are perm anently m arginal.11 In her discussion o f ritual purity systems D ouglas focuses on the hum an body as a pow erful m etaphor for society. I f the body is a locus o f ritual action, this m eans th a t the do m in an t m etap h o r o f society is readily avail­ able to everyone. In this study, I will show how b o th the puppeteer and th e pup p et are im plicated in the ritual purity system, and the p u p p et serves as a body substitute for th e puppeteer, absorbing spirit forces and em ­ bodying disorder and ambiguity. T he equation o f the body and society, th en , is an example o f a total ideology, one w ith the pow er to be all inclu­ sive, while at the same tim e rem aining abstract enough to resist attem pts to challenge it. D ouglas writes, T he idea o f society is a pow erful im age. It is p oten t in its ow n right to control or to stir m en to action. This im age has form; it has external boundaries, mar­ gins, internal structure. Its outlines contain pow er to reward conform ity and repulse attack. T here is energy in its margins and unstructured areas. For sym ­ bols o f society any hum an experience o f structures, m argins, or boundaries is ready to hand. . . . T he b ody is a m od el w hich can stand for any b oun ded sys­ tem . . . . T h e functions o fits different parts and their relation afford a source o f sym bols for other com plex structures. W e cannot possibly interpret rituals c o n ­ cerning excreta, breast m ilk, saliva and th e rest unless w e are prepared to see in the body a sym bol o f society, and to see the pow ers and dangers credited to social structure reproduced in small on the hum an body.12

D ouglas notes th at closed societies (o f w hich Tokugaw a Japan is a p re ­ m ier example) are, like the hum an body, concerned w ith balance. O n e o f the prim ary goals o f a social system is th e m aintenance o f stability, and this requires m inute a tten tio n to those areas that could negatively affect this balance, places w here an o th er or otherness could m anifest itself. These areas she term s types o f contradictions and she m aintains th at the way these concerns get expressed w ithin society are usually encoded u p o n the body and its orifices. She identifies four threats to balance and stasis: ( I ) threats to external boundaries, such as foreign invasion, etc., (2) threats to internal boundaries, such as social revolt, (3) danger in the m argins o f society’s internal lines, and (4) the danger o f internal co n tra ­ diction, such as am biguously classified item s w ithin the system— people w ho fall into m ore than one social category o r into a place th a t serves contradictory purposes.13 D ouglas’s discussion is useful o n a num ber oflevels, m any ofw h ich will becom e apparent later in o u r discussion o f ritual puppeteers and the purity system they helped to m aintain. I m ention her w ork here because any view

o f a b o u n d ed system implies an O th e r against which th e system is bounded. Just as diseases and contagion (real o r im agined) are th o u g h t to violate the body from the outside and th en w ork w ithin it, so to o are outsiders and m arginal people feared as sources o f som e kind o f th re a t to the integrity o f social categories and order. T his, in sh o rt, is the stigm a suffered by strangers in cultures aro u n d the w orld. F u rth e rm o re , th ere arc the forces o f the darker side o f consciousness, frequently form ulated in m ythology as a b o rted , deform ed, or half-developed deities, w hich inhabit the boundaries, paradoxes, and interstices o f this b o u n d ed social system— and occasionally have to be reckoned w ith. In Japan, the handling o f such forces becom es the work o f specialists in impurity.

The World out There Comes in Here: The Dangerous Stranger and the M anipulation of Power W hat happens to a society’s sense o f order, boundaries, and purity w hen the unknow n w orld o u t there com es in here, in to society, violating its boundaries and sense o f itself? This is w hat a stranger— in any n u m b er o f guises— does. N u m erous examples in ethnographic literature illustrate how the unknow n person w ho arrives at a village o r ho m e from som e undesignated and unform ed place (the road, the w ilderness, the forests, the sea) is treated according to a special set o f rules o f conduct. T he reader can readily supply examples from folk tales, religious texts, and c o n te m p o ­ rary literature o f this pow erful m otif. A lthough the social identity o f this person may vary— the hom eless, the saint, th e leper, the beggar, the in ­ sane, the healer, the w itch, the aged— the range o f responses to the o u t­ sider are few: ( I ) the outsider is treated w ith utm o st caution and respect, and banquets, cerem onies, and e n tertain m en t are frequently presented for his o r h er benefit; (2) initially the outsider is treated in this m anner b u t after a b rief period o f tim e is expelled o r killed; (3) the outsider is treated in this m anner b u t after a period o f tim e is assim ilated as a “resi­ d e n t o th e r” and occupies an am biguous status w ithin the confines o f a given w orld o f m eaning; (4) the outsider is treated well and is assimilated into th e com m unity as a resident outsider, b u t w hen problem s arise, this new person becom es the focus o f negative atten tio n , prejudice, vio­ lence, and even m urder, ritual o r otherw ise; ( 5 ) the outsider o r stranger is either expelled o r killed (o r perhaps first to rtu re d and then killed) im m edi­ ately upon appearing at the village or hom e. O f n o te here is how a partic­ ular person, o r type o f person— an outsider— becom es a symbol for an entire order o f m eaning. T he stranger com es to sym bolize all th at is u n ­ know n, unseen, h idden, m ysterious, and potentially capable o f bestow ing blessings or calamity. An itinerant puppeteer, for exam ple, w hose role is

to unleash spiritual pow ers for revitalizing and cleansing the hom e and handle those forces th at are potentially dangerous to society, com es to sym bolize these frightening aspects o f sacredness and p o llu tio n , and is regarded w ith the very dread and abjection th a t p o llu tio n engenders. Ritual pu p p etry is an exam ple o f a m u lti-tex tu red alterity. R itual p u p ­ peteers, outsiders to th e com m unities in w hich they p erfo rm ed , m an ip u ­ lated beings once rem oved from the hum an realm — p u p p ets— to handle forces from outside the h u m an realm . A m orphous deities such as Ebisu and th e equally enigm atic ya m a no knrni (m o u n ta in deities) traverse th e cosm os at will and have at tim es adverse affects o n th e hum an condition in the form o f bad harvests, disasters, and even epidem ics. Ritual handling o f puppets expresses n o t only alterity b u t also th e c o n ­ tro l o f power. Just as m ainstream society m ust co n tro l and m anage the arrivals and departures o f puppeteers, so to o did th e puppeteers m anage the arrival and d ep artu re o f o th e r beings— particularly the deity Ebisu w h o , as we shall see, is the outsider deity par excellence: unform ed, am ­ biguous, unseen, unpredictable and potentially dangerous. I n sh o rt, p u p ­ peteers did w ith Ebisu and th e o th e r deities they contacted w hat society did w ith puppeteers— channeled th e forces represented by this alterity for their ow n benefit. C o n tro llin g the outsid er and m anipulating th e sym ­ bolic significance o f this status (be it occupied by a person o r a deity) was n o t only necessary b u t ritually desirable. T he outsider can be used ritually as th e representative o f th a t w hich is difficult to represent symbolically. H o w does one bring th e unknow n to life and m ake it present? Rituals involving outsiders are d ep e n d en t on the very alterity o f the perform ers; they arise o u t o f a need to p erio d i­ cally recreate and renew th e cosm os by inviting th e unkn o w n in to the realm o f th e know n for purposes o f purification, revitalization, and p ro ­ tection. Conversely, these rituals also d e p e n d o n th e ability to expel the outsider, so th a t a distinction betw een th e know n an d unk n o w n can be m aintained. AU o f these studies p o in t to a com m on apprehension: O utsiders possess a great deal o f sym bolic pow er. I t is n o t e n o u g h to keep these outsiders perpetually at bay, alth o u g h people m ay try to do so. R ather, it is neces­ sary to periodically m anipulate and even invoke their pow er, for a variety o f ends. In the case we are studying here, otherness is en coded and m an ip ­ ulated th ro u g h itin eran t ritual to achieve symbolic au thority over an entire o rd er o f m eaning. O utsiders have at once th e pow er b o th to destroy a closed system o f m eaning and to revitalize and purify th a t system. M aybe th e outsider will do w hat is desired, o r m aybe the outsider will refuse to cooperate. This dual potentiality creates yet an o th er level o f pow er, b o rn o f the fear o f n o t know ing how an exchange will com e o u t. H o w th e m any layers o f this symbolic pow er is handled makes th e difference betw een

these tw o very different outcom es o f co n tact w ith the w orld beyond the know n. T here is an im plied tyranny in the textures o f the unknow n. Now , we tu rn to a discussion o f the form ulation o f this problem in Japanese history by looking at the do m in an t categories o f o rd er th a t ritual puppeteers m aintained. As we shall see, society recognized the need for a ritually and semantically separate g roup o f people w ho could be responsi­ ble for taking care o f the w orld o f the unknow n— p ollution, chaos, death, and perhaps m ost im p o rtan t, the sacred. These people were isolated from the rest o f society physically and symbolically into a m ultitiered pariah gro u p , and yet it was their very o th e r ness th a t m ade it possible for m ain ­ stream society to m aintain its vision o f ord er and purity. B y th e M urom achi period (1 3 3 3 -1 5 7 3 ), m any o f th e rituals conducted by itinerant perform ers were those perform ed to rem ove p o llution and defilem ent from hom es and even entire villages. W hat ideological system underlay this practice o f itinerant perform ance? H e re we m ust look at a do m in an t ideological system in Japan d u ring the medieval period, the cor­ pus o f m eanings, attitudes and practices relating to purity and p o llu tio n .14 T his com plex, body-based ritual purity system, w hich has gone th ro u g h great transform ations th ro u g h o u t Japanese history, was expressed in an early and m ythical form in the redacted Japanese creation m yths, recorded in the eighth century the K ojiki and Nihongi. W hile this ritual purity sys­ tem was articulated w ith reference to the physical body, its excreta, and gender, these referents m ust be read as a com plex symbolic language d e ­ scribing the relationships betw een different levels o f the cosm os. These m ythical references to the body and pollution indicate th at a concern w ith ritual purity has a long history. In fact, m ost scholars o f Japanese society w ould agree th at the concern w ith ritual purity continues to be a strong preoccupation even today. Yokoi Kiyoshi, a scholar o f m edieval Japanese history, writes, “T he concept o f being contam inated by ritual pollution was n o t lim ited to people living in ancient or medieval Japan, b u t is som e­ th in g w hich, deeply ro o te d in ou r consciousness, continues to m ove us [even to d ay ].” 15 After an overview o f this system and its transform ations in Japanese history, we exam ine how puppeteers and o th er ritual perfo rm ­ ers w ere im plicated in these ideas o f the pure and th e polluted.

Pure and Im pure: A S h in to D yad It can be argued th a t th e m ost fundam ental and long-abiding distinction in Shinto practice is th a t betw een purity and impurity. A great deal o f S hinto practice is devoted to periodically ridding space, tim e, and bodies o f the pollution th a t tim e and living in the w orld generate. Perhaps the m ost p ro n o u n ced example o f this process o f ritual purification is the peri-

odic rebuilding o f the cerem onial center at Ise. Every tw enty years (until the recent controversy over lum ber) this large cerem onial center was co m ­ pletely to rn dow n, b u rn ed , and reconstructed. T h e significance o f this act was th at w ith o u t a periodic and total destruction o f th e site, th e center could n o t be spirituallyregenerated. Tim e itself in this case has generated the pollution th at requires the destruction o f the sacred architecture. T he new structure, built on the alternate site next to the previous stru ctu re, is u n d erstood to have been purified and revitalized. In this gigantic and costly exam ple, we see an im p o rtan t principle o f the Japanese ritual purity system. Ritual action o f a purifying natu re is u n d ersto o d to produce spiri­ tual revitalization and rebirth. T h e very process o f purification is a highly generative a c t.16 W hile the m ost co herent form ulation o f th e Japanese ritual purity sys­ tem can be found in texts considered today to be S hinto, m ost o f these texts are draw ing on a wider, pre-B uddhist religious ethos. As O hnukiTierney points o u t, “T he Japanese concepts o f purity and im purity are usually labeled as Shinto concepts. H ow ever, since the early historical rec­ ords included oral traditions, it appears th at Shinto simply gave an official stam p to already well-established values.” 17 W hat exactly counts as pure and polluted, and w hat w ords are used to d en o te these orders o f meaning? T w o term s are m ost com m only ttsed: hare, m eaning bright, clear, or pure, and kegare, m eaning pollu ted , d e ­ filed, or filthy. W hile these tw o term s and the states o f being they rep re­ sent stand in clear juxtaposition to one another, we m ust resist the te m p ­ tation to view purity as sacred and im purity as profane. As N am ihira Em iko and others have rightly po in ted o u t, and as o u r exam ination o f Japanese sources will show, b o th states are highly sacred, extrem ely p o w ­ erful, and potentially dangerous. In contrast to the sacredness o f both purity and pollution, there stands a th ird category, ke, w hich can be best u n d ersto o d simply as neither pure nor p olluted b u t m erely profane. Events th at create pollution are often potentially dangerous situations involving passage from one stage o f physical developm ent to another, such as childbirth, m enstruation, recovery from illness, and o f course death, funerary rites, and acts o f m ourning. C ontrolling the m ovem ent betw een states o f purity and pollution becom es a m atter o f extrem e im ­ portance in the Shinto ritual system. H are can be m ost easily defined as th at w hich enhances life and is crea­ tive. A num ber o f items in Japan are un d ersto o d to be inherently hare, and these reveal som ething o f th e substance o f this ord er o f m eaning: new rice seedlings, clear sunny days, festival days, sites w here a sacred presence has m anifested itself, and waterfalls are a few o f the m ost pressing examples. Ritual purity, however, is n o t a static or perm anent state. O ne moves into

and o u t o f states o f pollution in varying degrees th ro u g h o u t the day, the week, the m o n th , the year, and o n e ’s lifetime. Kegare can best be u n d ersto o d as th a t w hich undoes life and leads to death and d estruction. I t is a principle in the cosm os th a t works against creation. In the un d erstan d in g o f m edieval Japan, the m ost obvious exam ­ ples o f sources o f p o llution were actual d eath and decom position o f the body, b u t a n u m b er o f o th e r states and objects were also considered p o l­ lu ting and polluted: th e death or birth o f dom estic anim als, m iscarriage, ingestion o f m eat, reburial (kaiso), injuries th a t break the flesh, pregnancy, m en stru atio n , handling and extinguishing o f fire, and directly to u ch in g a w om an. W ithin this list, there was a hierarchy, and also a set n u m b er o f days d u ring which a p olluted person was excluded from appearing in g e n ­ eral public.18 H are exists in a constant dialectical relationship w ith its o p ­ posite, kegare, and this m ovem ent constitutes one o f the m ost im p o rta n t acts o f spiritual developm ent w ithin this system o f m ean in g .19 F u rth e r division o f kinds o f p o llution into tw o categories indicates the com plexity o f this system. Red po llu tio n (akafujo) generally referred to m en struation and childbirth; it also extended to all lettin g o f b lood, in­ cluding w ounds. Black pollution (kurofujo) referred to death and in­ cluded b o th the physical body o f the deceased, th e family o f m ourners, and o f course all acts involving m urder, killing, o r death o f hum an beings o r animals. In betw een all these categories, contact w ith bodily excreta— blood, sem en, m ucous or secretions from any orifice, pus, urine, feces, and vom it— was un d ersto o d to be polluting. B oth broad categories o f pollu tio n , red and black, carried a lim ited pow er o f contagion (called shokue, literally, “to u ch in g p o llu tio n ” ).20 C om ing in to contact w ith any o f these substances or states (d eath , m ou rn in g , childbirth, or m enstruation), w hether directly or indirectly, was u n d ersto o d to be p o lluting and re ­ quired ritual purification. Fear o f contagious p o llution gave rise to the practice o f separating people directly affected by p o llution from society— either tem porarily o r perm anently. Yokoi notes th at understandings o f p o llution in m edieval Japan were organized in to a system o f prim ary, secondary, and tertiary pollutions. Prim ary p o llution was the original place o r person w ho was p olluted (for exam ple, a corpse or a m enstruating w om an). Secondary pollution was com ing in contact w ith th e original pollution (to u ch in g a corpse or a m en struating w om an), and tertiary pollution was com ing in contact w ith a second-level p o llution (m eeting som eone w ho has just com e from a funeral, som eone w ho had to u ch ed a corpse, e tc .).21 O rdinarily, any individual will com e in co ntact w ith these pollution in the process o f being alive. Daily emissions, sexual relations, m enstruation and childbirth, and deaths in the family are com m on occurrences, so a

ritual purity system m ust provide ways for a person to be cleansed o f a tem porary pollution. A ccording to this system , there w ere basically four ways a person could be cleansed, and each had its appropriate application. T he different m ethods o f purification varied regionally. T he earliest m ethod o f ritual purification, w idespread in ancient Japan b u t used today only in certain areas, and then only in the cases o f m en stru atio n , child­ birth , or death in the im m ediate family, is the im m ersion in either a ru n ­ ning body o f w ater or the ocean. This practice is recorded in the earliest records describing the Japanese by Chinese w riters, dating from the fo u rth century C. E. T h e Gishi W ajiden describes the following: “W hen there is a death, they m ourn for ten days, during w hich period they do n o t eat m eat. T he chief m ourners wail and w eep, and the others sing, dance, and drink liquor. After the burial, th e w hole family goes in to th e w ater to bathe, like the Chinese sack-cloth ablutions.”22 T he m ost com m on m ethod is daily bathing, and this cleanses one from the pollution o f co ntact w ith the w orld outside the hom e (the contagion o f com ing into contact w ith persons or places o f undeterm ined purity), and the pollution o f daily bodily emissions. This kind o f bathing takes place in a tub o f pure w ater (either at hom e or at a public b ath ), which one enters only after having com pletely washed the body. Yokoi suggests th at this m ethod o f im m ersion probably grew from th e practice o f im m ersion in natural flowing rivers, and th a t tu b bathing as a m eth o d o f purification is rather late, perhaps dating from the medieval period w hen the intensifi­ cation o f the concern w ith ritual purity required m aking it m ore c o n ­ venient.23 T he th ird m eth o d o f purification uses salt, and this m eth o d is m ost com m only used to purify som eone w ho has com e in to contact w ith a death outside the family, usually th ro u g h atten d in g a funeral. U p o n retu rn in g hom e, a person w ho has atten d ed a funeral is m et at th e g a te ­ way o f th e house and salt is dusted over his or her shoulders (and som e­ times hair). Finally, th e use o f fire (and som etim es m erely sm oke) to purify objects th at have com e in contact w ith a m ajor pollution is n o t uncom m on. These four m ethods— im m ersion in natural bodies o f w ater, bathing in a tu b , salt, and fire— are w idespread even today in Japanese ritual purity practices. T he concern w ith this fundam ental distinction betw een purity and p o l­ lution is apparent in the Japanese creation m yths, recorded in the eighthcentury K ojiki and N ihongi texts. T he first book o f the K ojikir, consisting o f forty-six chapters, is widely considered to be a highly redacted narrative which incorporates various m yth cycles, including those from people living in the Inland Sea o f Japan aro u n d th e beginning o f th e historical period—people usually lum ped under the general category o f am azoku (literally “sea people” ). M any o f the m yths reflect a stro n g m aritim e influ­ ence. H ere we follow the redacted m y th ’s treatm en t o f essential physio-

logical events: sexual foreplay and union, illness, vom iting, defecation and urination, death, w eeping in grief, m urder and bloodletting, and bathing. These events, we shall see, all create substances o f am biguous status th at m ust som ehow be appropriated in to the cosmic system. W hat happens to the bodily excreta o f divine beings? H o w do these m ythical substances becom e symbols for larger m eanings in society? H o w does this creation m yth define these categories and paradigm s for hum an beings? T he first six chapters o f this m ythical narrative describe the com ing into existence o f five heavenly deities called the Separate Heavenly D eities and the Seven G enerations o f the Ages o f the Gods. T he Separate H eavenly D eities are w ith o u t form . T he text uses the w ords “ they hid their bodies” to express their invisibility and form lessness, b u t, as we shall see, this also denotes their com plete purity, for having a body becom es the source o f all p o llu tio n .24 In these first few chapters, we are to ld th a t tw o o f th e deities w ho are p a rt o f the Seven G enerations o f the Ages o f the G ods, Izanagi and Izanam i, are o rdered to consolidate a shifting and unform ed land. To do this, they low er a jew eled spear into the unform ed prim ordial brine and create an island (Onojyorojima, literally “ the self-curdling island” )25 from th e substance th a t drips o ff th e end o f it. T hey th en descend to this island and erect a heavenly pillar and a palace. H aving found a place on w hich to stand, th e discussion moves im m edi­ ately to bodies. Izanagi asks his wife, “H o w is your body form ed?” and she replies, “ My body, form ed th o u g h it be form ed, has one place w hich is form ed insufficiently,” to w hich Izanagi responds, “M y body, form ed th o u g h it be form ed, has one place w hich is form ed to excess. T herefore I w ould like to take th at place in my body w hich is form ed to excess and insert it in to th a t place in your body w hich is form ed insufficiently and [thus] give b irth to the land. H o w w ould this b e .” W hile possibly one o f th e least rom antic and m ost technical allusions to foreplay in w orld m ythology, the end result o f this exchange does n o t go w ell, for after the initial m ovem ents aroun d the heavenly pillar, the w om an speaks first before they have sexual relations. Because o f this, their creation is a leech, an am orphous blob, w hich even at the age o f three can n o t walk. (H ereafter in m ythology this child is referred to as H iru k o , the L eech C hild.) T he deities, realizing th a t som ething has gone w rong, abandon th e failed offspring in a reed b o at o n to the ocean and try again. T hey fail again w hen they create the island o f Awa, literally “ foam ,” n o t the stu ff o f w hich firm land is m ad e.26 Som e scholars have read these failed creations as m ythical expressions o f miscarriages (spontaneous abortions), abandoned to the w aters. W hile the physiological experience o f a m iscarriage m ay have inspired the fo rm u ­ lation o f the sym bolism , it w ould seem th at the idea o f a failed creation is a larger idea, o f w hich m iscarriage (along w ith failed crops, bad fishing,

and disorder) is but one dram atic example. T he m o tif o f the failed crea­ tio n is n o t uncom m on in w orld m ythology and does n ot always get inter­ preted as a direct reference to hum an beings. In this case, the tale o f the Leech Child gets resurrected at various tim es in Japanese history as a sym ­ bol o f the outcast. In the following chapter, we shall see th a t the legless Leech Child is identified w ith Ebisu and becom es a deity em bodying dis­ order, destruction, and epidem ics. H e m ust be en tertain ed and appeased by puppeteers lest he cause terrible calamities in th e hum an realm. T he creation m yth continues th ro u g h the successful creation o f n u m er­ ous islands by Izanagi and Izanam i in chapter 6, and in chapter 7 we see the beginnings o f the outlines o f a ritual purity system em erging. In this chapter, Izanam i gives birth to a deity called H i no Kaga H iko no Kami, the fire deity.27 “Because Izanam i no M ikoto bore this child,” the m yth relates, “her genitals were burned and she lay dow n sick. In her vom it, there came into existence th e deity Kana-yama hiko no kami; next Kana yama him e no kami. N ext, in her feces there came in to existence the deity Pani yasu hiko no kami; next Pani yasu bime no kami. N ext, in her urine there came into existence the deity M itu-pa-no-m e-no-kam i; next, Waku m usubi no kami. The child o f this deity is Toyo uke bime no kami. Thus at last, Izanam i no kami, because she had borne the fire-deity, divinely passed away.”28 This piece o f Japanese m ythology has been interpreted in a n u m b er o f ways, w ith m ost interpretations viewing the m yth as a narrational explana­ tio n o f natural phenom ena o f n atu re.29 W hat is outstanding in this m yth is the intense focus on a num ber o f physiological processes: vom iting, defecation, u rination, p arturition, and in the end, death. AU o f these b o d ­ ily events result in a physical product. W hat is the status o f the physical excreta (and finally the corpse) o f a goddess? In this case, these body products are highly creative: deities literally spring from them . O nce they were p art o f a divine body, but now they have been expelled. T he status o f these physical substances is am biguous and therefore they are regarded as highly polluted. T he next stage in the saga o f this creation m yth concerns the death o f Izanam i. T he m yth tells us: “A t this tim e Izanagi no M ikoto said, ‘Alas, I have given my beloved spouse in exchange for a m ere child!’ T hen he crawled around her head and around her feet, weeping. A t this tim e, in his tears there came into existence the deity w ho dwells at the foot o f the trees in the foothills o f M o u n t Kagu, nam ed Naki Sapa M e n o Kam i.” 30 In this section o f the m yth, another type o f bodily excreta, tears, is understood to be p o te n t enough to create deities. T he next section o f the m yth describes how divine blood likewise e n ­ genders deities. In his grief and rage that his wife has died in childbirth, the father Izanagi unsheathes his sword and cuts o ff the head o f his son,

the fire deity. B lood from various parts o f his sw ord— the tip, th e guard, and the hilt— all tu rn in to deities, eight alto g eth er.31 E ig h t m ore deities spring from the body o f the slain child, from his head, chest, belly, geni­ tals, and left and rig h t hands and feet.32 U p to this p o in t in the m yth, p roducts o f the body, far from being m erely profane, are regarded as pow erful and creative. This is a p o in t to be kept in m ind as we see the later m anifestations o f a m ore abstract ritual purity system. T h e w orld o f p o llution is far from profane. I t is highly g e n ­ erative w hen properly handled. T he climax o f this creation m yth is in the journey o f th e grieving Izanagi to m eet his wife in the land o f the d e a d .33 A lthough Izanam i asks her h u sb an d n o t to look u p o n her form , Izanagi cannot contain himself; he breaks one o f his com bs and lights it so he can look at her. W hat he sees is this: “At this tim e, m aggots w ere squirm ing and roaring in th e corpse o f Izanam i no M ikoto. In her head was G reat-T h u n d er; In her breast was F ire-T hunder, In h er belly was Black T h u n d er; In her genitals was CrackT h u n d er; In her left hand was Y oung-T hunder; In h e r rig h t hand was E a rth -T h u n d er; In her left foot was S ound in g -T h u n d er; In her rig h t foot was R eclining-T hunder.” 34 W ith this vision o f his wife’s ro ttin g corpse, Izanagi flees from the land o f the dead, pursued by his furious wife and th e hags o f the realm. A t the pass dividing the land o f the living and the dead, the tw o deities finally face each o th e r and announce th eir divorce. Izanam i vows to strangle a th o u ­ sand people per day, and Izanagi vows to build five th o u san d partu ritio n huts, o r places for b irth, p er day. After this ordeal, Izanagi is in need o f purification. H e says, “ I have been to a m ost unpleasant land, a horrible, unclean land. T herefore I shall purify m yself.” 35 H e th en purifies him self in a river. D eities arise from his p olluted articles o f clothing and various parts o f his body. W ashing his left eye, he creates the sun goddess A m aterasu; w ashing his rig h t eye, he cre­ ates th e m oon, Tsuku Yomi. F rom his nose com es the tu rb u le n t Susanoo. In all, fo u rteen deities are b o rn .36 This elaborate creation o f the Japanese p a n th e o n o u t o f the body p r o d ­ ucts o f deities’ illness, childbirth, d eath, grief, m urder, decom position, and w ashing could n o t be m ore em phatic: the products o f the body (and we can argue, w hat th e body symbolizes) are generative sacred substances. This m ythical trad ition casts the later do m in an t medieval Japanese ritual purity system in a different light. I t is n o t possible to argue th at this ritual purity system is a m isguided code to ensure hygiene. As D u m o n t points o u t, “ H ygiene is often invoked to justify ideas ab o u t impurity. In reality, even th o u g h the n o tio n may be found to contain hygienic associations, these can n o t account for it, as it is a religious n o tio n .” 37 U sing the ancient Israelite case as a m odel from w hich to interpret

the concern bordering on obsession w ith the body in ritual purity systems in o th er parts o f the w orld, M ary D ouglas suggests th a t a concern and obsession w ith the excreta o f the body reflects a deeper analogical p ro ­ cess. She m aintains th a t in cases w here there is a deep concern over the excreta o f the body, there is also a concern w ith guarding the b o u n d ­ aries o f the social group. T he body is an intim ate sym bol o f society itself, capable o f being generalized to the experience o f every living person w ithin the group. Yokoi presents a parallel case arguing from Japanese m aterials. H e m aintains th at the rigid ritual purity codes o f medieval Japan were really ab o u t social control o f the population and m aintenance o f social class distinctions. H e cites a fifteenth-century do cu m en t indicating th e p ro b ­ lem which arose w hen an outcast visited a hom e o n business, such as bringing w ood or delivering stone. T he outcast in this case was a kaw ara mono, o r “person o f the riverbed,” a term th at the text extends to people w ho lived in the Sanjo districts attached to shrines since b o th were d om i­ ciles o f outcasts. T heir very presence was defiling to anyone w ho came in contact w ith them , and one needed to consider this in o n e ’s daily life. Such a system guaranteed a rigid social hierarchy. 38 W hen the ruling class is concerned w ith its hegem ony and w ith the clarity and purity o f its identity, and w hen m oreover it wishes to enforce at all costs its right to rule, th e m argins o f th a t society m ust be carefully w atched for threats. Likewise, the body, w ith its m any orifices and am big­ uous products (tears, feces, blood) m ust be carefully guarded. A bodybased ritual purity system extends the symbolism o f social control to every single person in th e society, at once m aking the symbolism real, everyday, and im m ediate. I t is also n o t hard to im agine how the m aintenance o f a system such as this can serve as a suitable pseudoconcern for people, for it is all pervasive and dem anding. T he p ro p er handling o f these products is o f the greatest ritual as well as practical concern. P u t m ore directly, how we relate to w hat goes into and comes o u t o f th e body determ ines the purity o f th at body. You are w hat you eat and to u ch and how you treat w hat you expel, just as your village (or nation) is w hat it allows in and assimilates and w hat it forces out.

T h e P er so n ifica tio n o f D e file m e n t a n d th e S y ste m a tiza tio n o f D isc r im in a tio n W ith a ritual purity system such as the one described above, a n u m b er o f logistical problem s arise for society. H o w can the products o f pollution, b o th physical and spiritual, be safely contained and rem oved from society w ith o u t infecting everyone around w ith their potency? It is the job o f

KADOZUKE: O U T S ID E R AT T H E GATES

67

culture to m ediate experience for people and to safeguard order and classi­ fications. Pure and p o lluted, sacred and profane, inside and outside, this and th at, m ust be allowed to interact only according to a prescribed set o f rules. T he job o f handling those aspects o f society w hich by definition will be p o lluting becom es th e dom ain o f specialists, w hom Louis D u m o n t calls “the specialists in im purity.” 39 W hat is the status o f those w hose job it is to com e around and purify the neighborhood? W hat is the role o f those people w ho m ust perm anently traffic in the realm o f the m etap h o ri­ cal equivalents o f bodily excreta? Any analysis o f Japanese ritual purity systems m ust begin by recognizing th e im pact such a system has on those requ ired to m aintain it. A t w hat cost is this system m aintained?40

B u rak u m in : N om enclature a n d Unconscious Associations

In the last several decades, scholars in the U n ited States and E urope have becom e aware o f the civil rights struggle o f Japan’s historically discrim i­ n ated class, called today b u ra ku m in (m eaning simply “settlem ent o r vil­ lage p eople,” a label ad o p ted by the people them selves to replace dero g a­ to ry labels).41 In central and w estern Japan, w here m ost o f these people live, being a person from a village designated as the dom ain o f outcasts (,bisabetsu buraku) is still a highly charged issue. References to their stru g ­ gle for equal rights are on public buses and street signs th ro u g h o u t the Kansai area, calling for an end to social discrim ination.42 It is im p o rta n t to see puppeteers b o th as included in a larger group o f social outcasts and as d istinct from som e o f the associations o f this group. F rom the perspective o f m ainstream Japanese society, contact w ith anyone considered polluted was itself potentially polluting. From w ithin the o u t­ cast com m unities, however, one had to distinguish betw een different types o f specialists o f impurity. T here was n o t only a rigorous hierarchy b u t also great variation in specialization. T here were those w ho cleaned latrines and tem ple grounds, those w ho prepared the dead for crem ation, and those w ho specialized in the care o f the sick and the delivery o f chil­ dren. T here were also sacred specialists w ho rem oved invisible pollu tio n — spiritual transgressions and the pollution b ro u g h t ab o u t by the passage o f tim e; and there w ere ritual specialists w ho contacted the dead and deliv­ ered their oracles and messages to the hum an com m unity. W hat all o f these cases have in com m on is th a t they deal w ith the m argins o f the h u m a n condition·, death and sickness, body orifices, or m argins betw een the hum an and sacred realms. T hey cross boundaries th a t others do no t, and com e back to tell ab o u t it. T he rituals we will look at in the follow ing chapters involved neg o tia­ tions w ith the forces capable o f ensuring safety, ab u n d an t harvests and

good fishing catches, and the health o f the community. N egotiating with these forces meant crossing a line out o f the everyday hum an condition into a liminal plane where two realms meet. The task o f conducting these negotiations fell to certain sacred specialists who used beings once re­ moved from themselves—puppets— as receptacles for the forces they summoned. The burakum in o f Japan are historically from several different groups. O n Awaji, many are the descendants o f puppeteers.43 To understand some o f the forces creating their outcast status, we can look at a larger dynamic in Japanese religious history. From the H eian period, the office o f sacred specialist divided into two distinct and to some degree com peti­ tive professions. One was the kannushi (sacred specialists attached to shrines). The other was the waza bito (performers), who presented nonecclesiastical sacred performances, o f which puppetry was one type. This group included onmyoji diviners, nembutsu dancers, performers o f banza.ira.ku (felicitations and blessings), and kugutsu-mawashi (puppeteers) to name a few. They tended, from about the late Heian period on, to congregate in areas connected to major Sanjo shrine complexes, discussed below. The waza bito came to be regarded as outcasts for what seems to be two overlapping reasons. First, they were sacred specialists, and their power to make divinations, perform exorcisms and purifications, and bring back spirits o f the dead—in short, to “cross the bridge out o f this w orld”—gave them an aura o f sacredness and power that was frightening. They performed a necessary function in society, but they were kept at a distance. Second, it appears that they were partially or totally itinerant. These two features o f their lifestyle turned them into the powerful O ther from another plane—although often just another village. As Yoshida has pointed out, “The attribution o f mystical evil qualities to newcomers may be a reflection o f the Japanese villagers’ traditional fears and suspi­ cions o f outsiders and strangers. . . . Newcomers and strangers . . . are regarded as dangerous for some reason. In this we see the underlying pattern o f ambiguity.”44 A third, related issue in the case o f puppeteers concerns the puppets themselves. This aspect o f the perception o f the otherness o f puppeteers has been overlooked by scholars. It is n o t just that puppeteers come from outside, nonagricultural spaces that they are other. As we saw in chapter I, puppets challenge the boundaries o f what it means to be human. Pup­ pets, which exist “in the shape o f the hum an,” are n o t simply metaphors for the human but actually comprise a world o f their own, a parallel world bridging the domains o f the human and the divine. The puppet, as an intersection o f these two worlds, is powerful and frightening, eliciting both fear and fascination. They have the capacity to draw sacred forces

and can become vessels (yorishiro) in which the sacred comes to dwell. Furtherm ore, they can embody the souls o f hum an beings, both living and dead. Something about puppets is grotesque. They draw their great­ est power when they are not merely mimetic devices depicting human activity (a task at which they invariably fail) but when they occupy their own parallel world. Since this parallel world is powerful and frightening, the people who contact it are to be kept at a distance. Awaii ritual puppeteers were indeed specialists in impurity, as we shall see in subsequent chapters. But the impurity they dealt with was not that o f the physical, tangible world but rather that generated through contact with the realm o f ambiguous deities, particularly the deity Ebisu. W hat certain specialists o f impurity were to corpses, puppeteers were to the spir­ itual world. They were integral players in the “spiritual cleanup” crew. Their role in society, often highly negative, was regarded within the same system o f meaning. They took away waza-wai (assorted pollutions) and made sure that a m ost liminal deity, Ebisu, was kept under control.45 The deities handled by puppeteers and the bodily products discussed above had this in common: both were ambiguous entities within the larger structure o f the cosmos.

T h e L o ca liza tio n o f P o llu tio n : R iv erb ed s an d Sanjo D istr ic ts in M ed iev a l Japan The ideology o f the ritual purity system defined certain groups o f people as permanently ritually polluted. The status o f these outcasts changed over time. By the Tokugawa period, when most o f the rituals discussed later were presented, their place in society was rigidly fixed by laws o f domicile, dress, and occupation. In medieval Japan, when social movement was not so rigorously restricted, the place o f outcasts was maintained through a symbolic mapping o f the landscape into restricted zones where they could congregate and live. We turn now to a discussion o f two o f these special geographical zones where pollution became localized: riverbeds and Sanjo districts.

Kawara: When R ivers R u n Low The first o f these special zones in medieval and Tokugawa Japan was the kawara (riverbed or riverbank). Japanese rivers tend to run fast and full for certain periods o f the year, particularly during the rainy season, and so the channels o f major rivers have two levels. The deeper level remains wet

or at least m uddy w hen th e river is n o t high. T he higher level is a floodplain th a t the river covers w hen it is flowing at its fullest in the spring and early sum m er, at least in unusually w et years. A lthough this area u su ­ ally rem ains dry, is n o t a predictable area for agriculture, since any serious rainy spell can send the river up in to its higher level. Since this kawara land could n o t reliably be p u t to productive use, they were n o t taxed. They were legally as well as symbolically outside the Japanese authority system. M arginal groups o f people m ade their tem porary hom es on this land and were able to live a life som ew hat outside the law. T he occupations o f people living in these riverbed districts reveals an interesting localization o f pollution. Gravediggers, road cleaners, haulers o f night soil, com b m akers,46 ritual puppeteers, and midwives all m ade their hom es in kawara, and hence were referred to as kawaram ono (river­ bed people). M ore d erogatory was the term kaw ara kojiki (riverbed beggars). T h e m ost fam ous o f these riverbeds is th at o f the Kam ogawa in K yoto, particularly the area called Shijp Gawara (T he F o u rth Street Riverbed). A large n um ber o f people living in this area were engaged in the business o f perform ing. In fact, the popular perform ing art Kabuki is said to have been started by a shrine a tten d an t from the Izu m o shrine nam ed O kuni, w ho is said to have perform ed a dance at Shijo Gawara in 1603. This area later becam e a m ajor center for Kabuki, w ith m any actual theater buildings lining the riverbanks, a rem inder o f this relationship betw een specialized geography and the developm ent o f perform ance in Japan. Shijo Gawara was also the location o f the first form al p u p p e t theater, started by a puppeteer w ho had left the N ishinom iya shrine, a center for p uppetry to be discussed in the follow ing chapter. A second p o in t to note in a discussion o f kawara is how people w ho were regarded as perm anently polluted w ere geographically isolated into the m argins o f society. A ccording to a Japanese-D utch dictionary p u b ­ lished in Nagasaki in 1603, the following definition can be found under the heading “ kawara no m o n o ” : “Also called kawaya. People responsible for rem oving the hides o ff dead beasts and m anaging lepers.”47 This defi­ nition is interesting because it suggests th at by the early seventeenth century, the nom enclature o f the riverbed was already rigidly fixed on the idea o f these zones as places o f physical pollution— the abodes o f tanners and lepers— even th o u g h the people w ho lived there w ere doing a w ider variety o f jobs. Riverbed districts were m arginal in alm ost every respect— legally, g eo ­ graphically, agriculturally, and socially. People living there could literally slip th ro u g h the cracks o f society. T heyfloated on their boats th ro u g h o u t Japanese society tying up at a riverbank and staying on land for extended periods o f tim e only w hen the rivers g o t too dry. It is n o t hard to im agine

how such a p h en o m en o n could becom e em bellished in the popular im ag­ ination. O ne is rem inded o f the im age o f F o ucault’s N arrenschiffen. B ut w hereas the E uropean dyad was reason and m adness, here it was the p o l­ lu ted and the pure, the settled and the itinerant.

Sanjo D istricts: Setting the M argins T he symbolic geography o f Japan located the m ediation o f disorder, transgression, and p o llution elsewhere: in Sanjo districts. Sanjo districts offer an o th er vantage p o in t for understandin g the shifting significance o f ritual puppeteers from the late H eian period th ro u g h late Tokugawa. T hree m ajor shrines w here ritual p uppetry has been pro m in en t— th e Usa H achim an shrine, the N ishinom iya Ebisu shrine, and the cerem onial cen ­ ter o f Awaji puppetry, O m ido H achim an D aibosatsu— are adjacent to or located w ithin districts called Sanjo. These districts, and the people w ho lived in th em , w ere part o f the im agination o f otherness, expressed in this case n o t in the lives o f people b u t in geography. D uring m edieval Japanese history, m any ritual puppeteers were predom inantly affiliated form ally or loosely w ith these special districts and cerem onial centers. W hat is im p o rtan t in studying these districts is n o t just the activities o f the people w ho lived there and their roles w ithin Japanese religious life. R ather, the very existence o f special districts o f divination and pollution suggests th at they were im p o rta n t in the Japanese im agination o f lan d ­ scape. M eaning and value were literally transposed o n to the geography o f Japan. C erem onial centers o rdered the practice, ideology, and politics o f Japanese ecclesiastical religious life, and districts at the m argins o f these centers form ed a locus for all th a t did n o t fit w ithin the realm o f the co m ­ m onplace. W hat Eliade has called “b oundary situations” o f the hum an condition— death, birth , extrem e sickness, epidem ics, delivery o f oracles, and longing to make contact w ith spirits— these aspects o f life were c o n ­ tained w ithin certain physical spaces, so th a t the m eanings these situations e ngendered could be contained, controlled, and channeled. W hat a study o f Sanjo districts discloses is a medieval sym bolization o f the cosm os and a m eans for con ten d in g w ith disorder, am biguity, and ru p tu res in planes o f m ean in g .48 W hile m ost studies o f cosm os in the history o f religions focus on the problem o f finding a center (a la Eliade), w hat Sanjo districts tell us is th a t an essential aspect o f a cosm ic o rd er is its m argins. This, in large p a rt, is w hat th e gradual ostracization o f puppeteers in Japanese his­ to ry is all ab o u t. T he Sanjo cases provide a clear expression o f this task o f w orld building and spatial organization in m edieval Japan. We shall dis­ cuss below th e general history o f these districts, and in subsequent chap­ ters look specifically at those Sanjo districts fam ous for puppeteers.

Sanjo: H istory an d M eaning

W hat exactly w ere Sanjo districts and w hat can be know n a b o u t them ? T he exact con n o tatio n o f the term is difficult to determ ine, because the term has been w ritten using different characters th ro u g h o u t Japanese his­ tory. T he variety o f character com pounds used to w rite “Sanjo” provoke som e interesting speculations as to how these districts were im agined and signified th ro u g h o u t Japanese history: “scattered place” (as opposed to central place or real place),49 “place o f divination,” “ birth in g place,” “ m o u n ta in to p ,” and “th ird district” are all characters th a t have been used to write Sanjo. M ost scholars attem p t to find the m eaning o f these sites by insisting th a t one reading o f these characters is correct and developing an interp retatio n along these lines. B ut even the same Sanjo district can be w ritten differently at different tim es. This variety and confusion tells us som ething n o t only a b o u t the range o f m eanings th at these places had b ut also a b o u t the confusion and even th e hysteria th a t signifying the unsignifiable has generated in Japanese society over tim e. Sanjo districts are decidedly a m edieval p h en o m en o n , and their exis­ tence reveals a great deal ab o u t the medieval vision o f place. N agahara Keiji presents a list o f textual references to Sanjo districts and persons, including tw enty tw o references dating from 1045 to 1 2 4 6 .50 T he ques­ tion o f the nature and existence o f Sanjo districts before M urom achi is com plex, b u t the earliest d o cum ent using this term writes the nam e w ith the characters “scattered place.” T here were at least tw o lines o f developm ent for these places called Sanjo. M ost o f them originated as special com pounds for ritual p erfo rm ­ ers (musicians, diviners and deity pacifiers, etc.) attached to m ajor cere­ m onial centers. D uring the medieval period, such people were the d o m i­ nan t residents o f Sanjo districts. In this context, it is im p o rtan t to n o te th at in prem odern Japan, perform ance was prim arily intended for divine entertainm ent. Frequently, perform ances had a sham anic elem ent, and beings from o th er realms w ere regularly contacted th ro u g h the m edium o f perform ance. R eferring to Sanjo districts and the perform ers w ho live in them , Hayashiya has n o ted th a t ritual perform ers attached to T ennoji (located in present-day Osaka) lived in a Sanjo district attached to the shrine; and he suggests th at these perform ers had a large im pact on ritual perform ance in o th er areas o f Japan. As the dem and for ritual perform ers specializing in sham anic activities (delivering oracles, curing, calling back the dead, in ­ voking divine beings, etc.) began to decline and residents o f these districts suffered the inevitable fall from status, they dispersed th ro u g h o u t Japan, and specialists originating in the T ennoji Sanjo district could be found as

far away as T o h o k u . Hayashiya cites a selection from ChUyuki (dated 1114) w hich discusses the perform ance o f dances at Byodoin in U ji (near present-day K yoto). It seems th at at the end o f this perform ance, w hen paym ent and praise w ere being d istributed to the perform ers, those from the Tennoji Sanjo district were denied equal rew ard.51 T he passage clearly shows th a t even from this early date, Sanjo residents were discrim inated against. A lthough residents o f Sanjo districts had clear religious duties at the shrines and tem ples to w hich they w ere attached, otherw ise they were a pariah group. T h ro u g h o u t th e M urom achi period, Sanjo districts appear to have been fluid spaces o f p ollution, and so references around Japan describe these sites differently. W hile they may have originated as the dom iciles o f perform ers and diviners, by the Tokugaw a p eriod they w ere often associ­ ated w ith blood pollution. A look at perceptions o f these sites d u ring the Tokugaw a period reveals the results o f Sanjo districts acting as a catch-all for various m em bers o f outcast groups for several h u n d red years. A ccording to Kida Teikichi, the m ain inhabitants o f Sanjo districts were diviners who p u t sacred chants (tonaegoto) to song and dance and w ho, on the side, “co u ld n ’t resist” becom ing perform ers as w ell.52 A text dated 1734 lists a n um ber o f districts called Sanjo th ro u g h o u t the co u n try and indicates th a t these districts w ere places w here m en struating w om en w ould go. A ccording to this list, by the m id -eighteenth century there were twenty-five Sanjo districts th ro u g h o u t the country. As the list is rela­ tively late, m any o f the sites were at the tim e o f com pilation n o th in g m ore than place-nam es. O f th e tw enty five, three were prim arily th e domiciles o f T aoist diviners ( onmydji), tw o o f gate chanters (shomon hoshi), tw o o f puppeteers, one o f a female sham an ( m iko), and one o f ritual purifiers (harae-m ono); four were place-nam es only, tw o belonged to farm ers, and in over ten the origins o f the site were unclear since descendants o f the original district were n o longer living th ere .53 These districts called Sanjo shared several basic characteristics: ( I ) they were affiliated w ith or attached to m ajor cerem onial centers or m anors; (2) they were restricted zones, and access was discouraged lest one com e in contact w ith pollution; (3) they were frequently places w here w om en w ould give b irth and spend th e po stp artu m p eriod o f pollution; (4) they were occupied by Taoist diviners, perform ers, and o th er specialists w ho m ade contact w ith spirits and deities; (5) the dead were frequently p re ­ pared for crem ation in these places; (6) lepers, anim al skinners and leather w orkers, and people engaged in m enial w ork at tem ples and shrines w ould reside in these districts; (7) they are m ost prevalent in the Kansai area, w here the im perial family resided for m ost o f Japanese history, and are frequently th o u g h n o t always in close proxim ity to an imperial to m b ; and (8) they were n o t taxed. Itin eran t perform ers w ho traveled aro u n d Japan

could com e and go freely th ro u g h these districts, and Sanjo districts may have becom e com m on stopping o ff places for m arginalized people d u ring the medieval period. T he fluid m ovem ent in to and o u t o f these districts is a sharp contrast to the lack o f social m obility outcasts suffer u n d er the Tokugawa shogunate governm ent. T hese places o f exorcism , divination, magical healing, childbirth, p u p ­ p etry perform ances, ecstatic dancing, and tre a tm e n t and burial o f the dead co nstituted b o th a physical and im aginary landscape o f otherness. It is unlikely th a t the average Japanese w ould have ventured in to a Sanjo area except on business o f a highly pressing spiritual nature. These districts participated in the pow er and dangerous potential o f the sacred, w hich the itinerant sacred specialists w ho issued from Sanjo em bodied and to o k on the road. By the late m edieval period, traveling papers always indicated a person’s nam e and type o f business. Because perform ers from these dis­ tricts often had th e nam e “Sanjo” (or “Sanjo h oshi” ) affixed to their nam e (as in “ the D okum bo from the Sanjo on Awaji” ), a perform er w ho resided in such a district carried a signification o f otherness w ith him w hen he traveled. Sanjo was n o t simply an address, it was a m ode o f being.

U n u su a l B ein gs: O th ers in Japanese C o sm o lo g ie s T h e outsider at once encodes symbolism o f the dead, the polluted, the banished, the unseen. A large part o f the Japanese ritual calendar is o rg a ­ nized around the idea o f beings from o th er realms visiting the hum an realm to bring blessings and fecundity. In h ere n t in the o u tsid er’s entry in to the established ord er is the conviction th at the everyday w orld will be in som e way revitalized by contact w ith outside m odalities. T he sym boli­ cally controlled m ovem ent o f the outside, which co n stitu ted a fundam en­ tal ritual com plex in Japanese nonecclesiastical religion, is predicated on an understanding o f the pow er o f outsiders. Japanese scholars refer to it as marebito shinko (belief in m arebito).54 M arebito is the w ord given to the sacred person w ho visits o n special occasions, and is usually w ritten w ith the tw o-character com pound m eaning “ the rare p e rso n ,” im plying the phenom enal nature o f the visit. T he category o f m arebito has been m ost widely discussed by the Japa­ nese folklorist and p o e t O riguchi S hinobu (1 8 8 7 -1 9 5 3 ), w ho is credited w ith creating the term to describe a wide ranging p h en o m en o n in Japa­ nese religious practice.55 T he basic prem ise o f marebito is th a t at certain tim es o f the year, and in particular the period betw een the harvest and the new year, the entire cosm os is in a state o f negative capability and p o te n ti­ ality. T h a t is to say, w hether or n o t the cosmos will spring to life in the new year is undeterm ined. T h e entire realm o f hum an existance is depen-

dent upon the forces from the other realms—sacred beings and an­ cestors—to regenerate the cosmos, revitalize life, release fish for the com ­ ing year’s catch, and germinate rice seeds. D uring this period o f time, therefore, the visit o f these sacred beings is ritually actualized. Certain beings come not just into a village but into each house, where they recite words o f blessing and perform magical rites to purify the hom e and revi­ talize the sacred forces upon which existence depends. These beings are understood, ritually if no t literally, to be forces from other realms, and they are treated with special care while they are present. Dressed for travel, they are given food, money, and lodgings. Banquets are held in their honor, and they and are welcomed into homes and village places. Their arrival and departure are highly marked m om ents in these ritual visits. At the end o f their visit, just as at the beginning, they are carefully escorted back to the margins o f the village, so that they can return to the world from which they came and life in the hum an realm can return to its bal­ ance. Too much sacredness would threaten to change the very nature o f the hum an realm, which requires a carefully guarded economy o f sacred forces mixed with the activities o f daily life—fishing, agriculture, sericul­ ture, tending children, and living and dying. Inherent in the pattern o f m arebito is the possibility that a stranger or beggar could be a visiting ancestor or kami in disguise. The traveler partic­ ipates in this symbolic structure, which equates the unknow n with the potential to generate renewal (or threaten destruction). Consequently, notes Origuchi, Japanese attitudes toward beggars and travelers can be seen as a veiled expression o f this underlying religious sensibility.56 One o f the only studies in English o f the practice o f the visiting stranger at New Year in Japan is Yoshiko Yamamoto’s thorough work, The Nam ahajje.57 This excellent ethnographic study is based on fieldwork by the author on the Oga Peninsula, the small finger o f land jutting o u t into the Japan Sea in northern Japan, where the practice o f New Year visitors (young men from the village dressed as “other beings” in straw coats and hats) was, until the postwar period, a dom inant ritual event. These visi­ tors, called Namahage, w ent from door to door on New Year’s Eve and threatened and terrorized young wives and children so that they would be obedient in the com ing year. D uring their visit, they made frightening noises, and broke rules o f propriety by searching the house for their hid­ den victims, pinching and chastising them when they were discovered. The purpose o f the visit was n o t to entertain but to terrify: to safeguard order and hierarchy, in gender and age, during the com ing year. In return for their visit, Namahage were given sake and food, becoming increasingly more intoxicated with each household visit. Yamamoto’s study rightly points o u t that belief in these visitors had little to do with the ritual event. M ost people (even some children) knew

these people actually were m em bers o f the com m unity dressed as “o u tsid ­ ers.” N evertheless, people found th e visits terrifying. W hat her study makes clear is th at a visit from the outside— even if only in a constructed and symbolic way— is an im p o rta n t elem ent in the transition from the old year to the new. As Y am am oto’s case m aterial points o u t, th e N am ahage w ere p art o f a larger practice o f the “ N ew Year’s visitor.” A m ap in her book shows th at in the prew ar period, examples o f visiting outsiders com ing aro u n d at N ew Year, like the D okum bo-m aw ashi described at th e start o f this chap­ ter, were quite w idespread, particularly in the n o rth east and in the area around the Inland Sea.58 T he visitors were either friendly o r terrifying. B ut in all cases the visits were connected w ith the transition from the old year to the new and th e visits were at a fixed tim e, usually the lunar new year (the sixteenth o f th e first m o n th ); the visitors were always m arked in som e way, th ro u g h costum e and dress, as O thers; their arrival was announced by noise-m aking o r music; and their d epartu re was effected th ro u g h a transfer o f gifts o f goods (sake, rice, m oney, etc.). A nother aspect o f the m arebito m atrix can be seen in the Japanese prac­ tice o f the festival o f the dead (o -b o n ), held every sum m er d uring the seventh o r eighth m o n th (depending 011 w hat p a rt o f Japan one is in). W hile ostensibly co nducted w ithin the dom ain o f B uddhist practice, the view o f ancestors is clearly influenced by o th er traditions in Japanese reli­ gious life, including the m atrix o f m arebito shinko. D uring this rite, family m em bers go to the cem etery— th a t physical space symbolically serving as a link betw een the w orlds o f the living and the d ead— and bring the spirits o f their ancestors back to the house w ith them for several days o f festivi­ ties. Rites p u nctuate th e arrival and departu re o f the spirits. D uring this interval, the family m em bers can receive the benefits o f the presence o f these beings while at the same tim e serve their ancestors to stren g th en their spiritual power. T h e interdependence o f these tw o realm s is m ost apparent during th e rites o f o-b o n . A t the end o f the festival, an elaborate dance is held. I participated in th e o -b o n rites in a Japanese family living in th e village o f Fukura on Awaji in 1988, and this year was the first o -b o n after the death o f the gran d m o th er in this family. T h e afternoon o f the first day o f the festival a n u m b er o f us w ent to the graveyard and built a ritual fire on the family to m b sto n e, all the while reciting the H e a rt Sutra. We then invited the gran d m o th er to com e hom e w ith us, and to bring all the o th er ancestors w ith her. W hen we retu rn ed hom e, the father o f the house jo k ­ ingly asked, “ D id you bring G randm a back w ith you?” F or the days th a t the festival lasted, we set a place at the table for G rand­ m other, and each evening spent o u r tim e looking at old p h o to album s and

discussing her personality while m em bers o f th e family recalled events from her life. This event im pressed upon me the shallowness and sim plic­ ity o f insisting o n “b e lie f’ in events as the prim ary category to assess their religious validity and potency. D id m em bers o f this family “ believe” th at the g ran d m o th er’s spirit was back w ith them ? Perhaps, and perhaps not. But her m em ory was being ritually revitalized th ro u g h this intim ate and coded annual practice, and a sense o f her presence was generated o u t o f o ur daily activities designed to focus on her life. O n the final n ig h t o f the festival, a dance was held o n the beach near the ocean, and the spirits o f the ancestors w ere sent o u t on small boats w ith candles in them back to the realm from w hich they came. Arrival and departu re are the m ost elaborate m om ents in this “visit” from an ­ o th e r realm. In b o th o f these cases o f m arebito shinko, a n u m b er o f key features are w o rth underscoring. First, alth o u g h it is possible for beings from o th er realms to visit u n an nounced, the ritual system controls the tim ing o f these visits, so th a t they com e d u ring the period o f the year w hen the presence o f sacred forces is required. Second, these visits from sacred beings are tem porary, and th e ritual system ensures th a t the visits com e to an end. C on tro llin g the tim ing and the d u ratio n o f sacred visits ensures th a t a balance and econom y is m aintained betw een the tw o orders o f m eaning. This is accom plished th ro u g h the elaborate sym bolization o f arrival and dep artu re, and the careful m arking o f the visitor as outsider (w hether th ro u g h strange clothing, noises, musical instrum ents, or ritual proces­ sions). T h ird , sacred beings are un d ersto o d to bring blessings and have the magical pow er to revitalize the com m unities and hom es they enter. These blessings are m ade possible th ro u g h elaborate ritual events in which song, dance, and m im e ritually actualize the bestow ing o f blessings and good fortune. I t is clear th a t the ritual context o f itinerant perform ance appropriates this system o f m eaning. We now tu rn to a discussion o f this context.

K adozuke: R ites at th e Gate Itin e ra n t perform ers have been a com m on feature th ro u g h o u t the history o f Japanese perform ing arts. A num ber o f w ords are used to d en o te itin er­ ant perform ance: hyohaku no g c i (drifting arts), yugyo no get (itinerant arts), and kadozuke nogei (arts attached to the gate). AU three term s situ ­ ate a wide range o f ritual perform ances in the context o f the relationship betw een perform ers and the people for w hom they perform . Q uite simply, the perform ers are outsiders, com ing from the road to perform in the

shrines, gateways, and even the zashiki o f private hom es and m anors, at street crossroads, and even in the gardens and room s o f the Im perial Pal­ ace before the imperial family. Perform ances presented by itinerant perform ers were one o f the m ain form s o f en tertain m en t in M urom achi Japan. T he m ost com m on term used to describe perform ances presented in this way is kadozuke.59 T he w ord literally m eans “ attached to the g a te ” o r “ at th e g ate,” b u t it also refers to rites, skits, acrobatics, and dram atic perform ances at shrines, crossroads, and m arketplaces o r “on the ro a d .” An im p o rta n t g roup o f such itinerant perform ers were puppeteers, w hose drifting lifestyle was a source o f fascination, fear, ritual renew al, and even political expediency for people in m edieval and early m odern Japan.

Locus an d Meaning: The Significance o f Itin era n t Performance As n o ted above, the visit o f strangers at certain tim es o f the year was asso­ ciated w ith the arrival o f sacred forces from beyond the hum an realm. Because o f the way th at the w orld “o u t th e re ” beyond the settled and defined param eters o f village life was perceived, strangers w ho arrived from unknow n places were clearly associated w ith the pow er o f the u n ­ know n. A dd to this general perception the abilities o f these perform ers to enact all kinds o f fanciful theatrical transform ations, and it is n o t hard to im agine how itinerant perform ers could be perceived as magical and p o w ­ erful. Jugglers, conto rtio n ists, sw ord swallowers, storytellers, and ventril­ oquists w ere am ong the magical perform ers w hose originally religious art form s w ere slowly popularized outside o f their strict ritual settings. T he term kadozuke arises from a com m on u n d erstanding o f th e gate as a bou n d ary betw een tw o orders o f m eaning. Gates figure p rom inently in Japanese dom estic and ecclesiastical architecture. Large Japanese m anors have always had large covered gates as entryways in to the inner area sur­ ro u n d in g the house itself, and this p a rt o f the architectural design o f a hom e (which was seen in a less elaborate form ) h ad, and continues to have, special symbolic significance. I t is a barrier betw een the w orld in ­ side the hom e and the w orld outside. T he distinction betw een “inside” and “o u tsid e” as an im p o rta n t elem ent o f Japanese ord er has been widely discussed.60 Gates o f hom es, tem ples, o r cities as symbolic spaces have figured prom inently in depictions o f im p o rta n t shifts in Japanese society. F or ex­ am ple, many readers will be familiar w ith Kurosawa Akira’s fam ous film Rashomon (Rasho G ate). T he entire eerie narrative o f the film is fram ed by the large gate o f R asho, as tw o speakers m eet u n d e r th e shelter it provides

and discuss the strange tim es they are experiencing. In using the gate as the settin g for this film, the d irector is appealing to a com m on Japanese perception o f gates as lim inal places. T h e character “g a te ” in kadozuke refers n o t to a particular gate but rath er to a m etaphorical quality o f space. N eith er inside the house nor part o f the street itself, the gate is the n o -m a n ’s-land to w hich all m eanings can be attached and o n to which all projections o f potentiality and danger can be projected. It is tem p tin g to assume th a t the term kadozuke arises b e­ cause perform ers presented their rites m oving from gate to gate, and m ost interpreters o f the term suggest this, b u t m ost likely th e term refers to the dom icile o f itinerant perform ers d u rin g the medieval period, w hen they frequently congregated and even lived u n d e r the large gates o f tem ple complexes. T hese gates were m ore than m ere doors to tem ples. T hey had huge roofs and were architecturally im posing, forcing w orshippers to a tem ple to consider th e issue o f passage from one realm to another. In Pure Land B uddhism th e tem ple gate sym bolized the doorw ay to A m ida’s paradise itself. They w ere also shelters for people, a place to escape the elem ents for beggars, lepers, and o th er lim inal p eo p le.61 T he gate was a feature o f the urban im agination o f space th a t em bodied otherness by being neith er here n o r th ere, n eith er inside n o r outside. T he history o f kadozuke perform ance reveals a subtle m anipulation o f this quality o f space to generate and enhance no tio n s o f sacredness. Tw o early perform ance form s, Senzu (or senshu)62 m a n za i and M atsu-bayashi, can be considered prototypical and paradigm atic exam ples o f kadozuke, and in fact it can be argued th at m uch o f the success and popularity o f kadozuke perform ers arose from th e popularity o f these earlier perfor­ m ance form s, which prepared the way for a w ider use o f this perform ance context. In b o th form s, the arrival o f an outsider at the gate to recite magical incantations and bless the hom e is a central feature. A review o f the tw o will give us an understanding o f how the kadozuke context and sensibility was form ed.

Senzu m an zai Senzu m a n zu i, literally “a th ousand autum ns and ten thousand years,”63 refers to a perform ance presented by itinerant perform ers, popular from the late H eian period th ro u g h early Tokugaw a, in which the perform ers w ould dance and intone felicitous m agical incantations for blessing and longevity. T he nam e is taken from th e refrains sung by the perform ers during their rite: “Senzu m anzai” o r “m anzairaku, m anzairaku, m anzairaku!”64 T he perform ers were referred to as Senzu m a n za i hoshi. T he

term hoshi, w hich literally m eans priest, was com m only used in m edieval Japan to d en o te religious specialists w ho m ay o r m ay n o t have had form al ecclesiastical affiliations.65 A dictionary from th e m id th irteen cen tu ry e n ­ titled Myojyoki defines senzu m anzai hoshi as follows (note th e residence o f these perform ers): iiSenzu m a n za i is a celebration o f th e first day o f the year seen these days o n O -S hogatsu. Beggar priests from Sanjo w earing the clothes o f m ountain people and h olding in th eir hands a small pine, m ake their rou n d s, reciting various w ords o f blessing, and recording things d o w n .” 66 O th e r sources from H eian th ro u g h M urom achi reflect this Kam akura definition, and com m on elem ents o f th e Senzu m anzai hoshi w ere as fol­ lows: A t N ew Year perform ers dressed as priests w ould proceed from gate to gate reciting magical blessings for longevity and prosperity and p e r­ form ing ritual dances. They carried in their hands a small pine branch. These perform ers were u n d e rsto o d to bring sacred forces w ith th em , and the branch in their hands was u n d e rsto o d to be the torimono, a receptacle for the sacred seen in m any Japanese sham anic practices.67 T heir perfo r­ m ance, a com bination o f song, dance, and magical incantation, was u n ­ derstood to bring th e blessings o f the sacred to everyone present. Al­ th o u g h they were dressed in ecclesiastical attire, they were in fact n o t p ro p er m onks o r priests, b u t m any received food and lodgings from par­ ticular tem ples, m uch like in d en tu red servants. F requently they were resi­ dents o f the Sanjo districts. O n occasion, these hoshi w ould perform certain roles in the tem ples and perform divinations and healings in tow ns and villages. In this way, notes M isum i H a ru o , these Senzu m anzai hoshi were p art ecclesiastical and p art nonecclesiastical religious specialists o f very low ran k .68 T he standard th eory as to the origin o f this perform ance form is th a t it developed o u t o f Japanese toka, a type o f dance th at involved stam ping and tripping the feet u p o n th e g ro u n d while singing and dancing. T he tripping and stam ping, while stylistic, also suggests the practice o f driving o u t evil spirits and dem ons. T he dance form is said to have com e from T ’ang C hina, and was popular at c o u rt, w here it was used a t N ew Year for blessing rites. O riginally toka were p erform ed in the Chinese style, o n the first day o f the new year at n ig h t u n d er the light o f the m oon. T he practice began in a b o u t the m iddle o f the seventh century as a N ew Year ritual in the im pe­ rial co u rt, presented every year by naturalized Chinese in Japan to pray for good things in the year. Toka w ere perform ed o n the fo u rteen th and fif­ te e n th o f the new year by m en and on the sixteenth by w om en perform ers. Beginnings o f the songs included T ’ang poetry, b u t th e others parts often used saibara, a popular if at tim es bawdy song style in Japan d uring the

N ara period. Som e o f the references may have had sexual overtones, b u t invariably the songs w ould conclude w ith references to ten thousand years, a set expression im plying longevity. T he w ords “Senzu m anzai” w ere p art o f the chants and poem s in these toka perform ances. T he oldest record describing t5ka is in the th irtie th book o f the Nihonshoki. In this chapter o f th e text, w hich deals w ith events d u ring the reign o f Em press Jito (r. 6 9 0 -6 9 7 ) we find several references to perform ances o f arareba-shiri. '9 perform ed by m en on the sixteenth day o f th e new year. At the end o f each stanza o f these perform ances, th e w ords “ m an nen arare” w ere repeated, m eaning roughly “m ay you live for 1 0 ,0 0 0 years!” 70 Gradually, toka perform ances becam e familiar to a larger g roup o f p e o ­ ple in th e capital as well. T h e popularity o f this ritual form caused n o small am o u n t o f agitation, and things got so o u t o f hand th a t th e singing o f toka in public was p ro h ib ited by a decree.71 D ue to the popularity o f these toka form s, the expression “Senzu m anzai” becam e very w idespread as a felici­ tous phrase. H en ce, m any scholars have argued th a t the later practice o f Senzu m anzai is derived from these toka origins.72 By the early H eian period, it appears th a t itin eran t perform ers n o t c o n ­ nected to th e co u rt were engaged in activities n o w recognizable as Senzu m anzai itinerant perform ances. These perform ers w ould travel from gate to gate reciting blessings and songs for the benefit o f w ealthy land owners. A n u m b er o f features o f this co n tex t for Senzu m anzai perform ances a t­ tract o u r atten tio n : First, there is the itin eran t nature o f the perform er. In his study o f Senzu m anzai, M orita discusses the probable origins o f N ew Year rites presented by “o th ers” w ho com e from outside the village or family unit. H is discussion is rem iniscent o f the m arebito shinko in te rp re ­ tatio n o f th e role o f outsiders. “A m ong Japanese p eo p le,” he notes, “it is believed th a t at certain prescribed tim es, divine spirits com e to visit, par­ ticularly at th e start o f the year, bringing blessings for th e new spring to each household. T his belief th a t deities w ho bestow blessings o n people com e to visit exerts a great deal o f influence even today. A person called a shinnin— divine person) dressed in strange clothes (to indicate th a t he or she is a deity) visits the gate o f each house and recites m agical w ords as blessings. . . . Senshumnnza,i is a p erform ing art w ith its origins in this practice o f a sacred person visiting each g a te .” 73 Second, the perform ers were m arked as outsiders by their attire and th e fact th at they held a pine branch in th eir hands. In subsequent chapters, I show how this m arking o f th e outsider and the transform ations in m eaning o f this person becom e m ajor com ponents o f these perform ances. T h ird , we no te their loose ties to ecclesiastical centers and their som ew hat am biguous status as “u n o r­ d ain ed ” o r n o t fully recognized priests, althou gh they are called hoshi. In this way, p art o f the role o f these perform ers was to m ediate a n u m b er o f

dichotom ies on society, including the tension betw een ecclesiastical ver­ sus nonecclesiastical religious authority. F o u rth , the relatively low status o f these perform ers as “religious specialists for h ire” suggests th a t while their services were necessary d u rin g certain seasonal events, they were rel­ egated to the m argins o f everday life for large parts o f the year— b u t it is n o t u n co m m o n for sacred specialists to be irrelevant for long periods. M ediating betw een different orders o f m eaning is necessarily highly irreg ­ ular em ploym ent. Fifth, we note periodic involvem ent o f these perform ers in m agic and divination, as well as occasional healing rites. Last, as the standard definition o f a Senzu m anzai in the m id-K am akura period points o u t, these perform ers were often from Sanjo districts. This origin was com m onplace eno u g h to make it in to the dictionary definition. In sum m ary, a special set o f agreed u p o n codes and symbols m arked the Senzu m anzai as set apart from th e rest o f society. N o t only did th eir attire tie them to cerem onial centers, b u t the fact th at they carried branches w ith them indicated th at they w ere possessed by, o r at least m essengers for, deities. As we shall see, these features o f the Senzu m anzai are all com m on to the later kadozuke context and shaped peo p le’s u n d e rsta n d ­ ing o f it.

M atsu-bay ash i M atsu-bayashi, a term w hich can be literally translated as “pine players,” “ pine bearers,” or “ pine noisy (or m errym aking),” refers to a practice com m on in Japan during th e M urom achi period in w hich strangers from m ountain villages w ould travel to the hom es o f w ealthy landow ners b e ­ fore N ew Year, carrying pine branches and reciting m agical blessings for the com ing new year. A ccording to M isum i H a ru o , they provided a ser­ vice by presenting people w ith the pine branches used to decorate entryways, b u t the branches were also to rim o n o , or receptacles he sacred. M isum i refers to the still com m on practice in Japan o f decorating o n e ’s hom e w ith a small pine branch for N ew Year, n o t dissimilar to the way w reaths decorate hom es aro u n d the holiday season in the U n ited States. FFe argues th a t this custom in Japan, now seen as m erely decorative, probably has its ro ots in a religious practice o f m arking o n e ’s hom e after the m ysterious visitor had arrived d u ring N ew Year.74 These M atsubayashi w ould leave a pine branch at the gate to w ard o ff evil forces and show th a t the house has been blessed and purified for the com ing year. In h ere n t in this visit, however, was an additional dim ension n o t readily apparent in the Senzu m anzai case. By m erely visiting the house, these M atsu-bayashi were also carrying away the pollution th at had been gener­ ated in the previous year. In exchange for their services, they were given

food and m oney. T h e n they were sent on their way, lest the p o llution they attracted should rem ain w ithin the area. Zeam i is said to have m en tio n ed M atsu-bayashi. In his so n ’s record o f his com m ents, Zesbi R okujd Igo Surugaku D a n g i (1 4 3 0 ), we find this observation: N ow adays there are no families specialized in M atsubayashi. W hat is perform ed d uring the G ion festival is probably an exam ple o f it. N everth eless, in the first m onth o f the secon d year o f Eikyo, as there was n o family at all to perform the M atsubayashi o f the palace, they asked Zeshi a few things. T h e o p en in g m elody is a congratulatory so n g and sh ould be plain. It sh ould be like this: “In the pine tress the w ind dropped and the clouds are m otio n less on M o u n t Inari, dropped and the clouds are m otion less on M o u n t Inari, the flow er dress o f this reign that b ecom es m ore and m ore prosperous . . . o h , spring is sp len d id .” B u t this tim e it was a little lo n g .7’

This b rief reference affirms w hat we know a b o u t M atsu-bayashi. I t also dates th e practice, show ing th a t it was already in decline by 1430, and indicates th a t the singing o f felicitous w ords was a central p art o f the per­ form ance, m uch like the Senzu m anzai perform ers. It is a com m on u n d erstanding in Japan th a t the deities responsible for generating rice seedlings ( ta no kam i) com e from the tops o f m ountains in th e springtim e and retu rn to th e m ountains in the fall after th e harvest. These perform ers, dressed in strange attire and clearly com ing from the m ountains, w ere seen as the bearers o f these sacred forces. T hey served the dual purpose o f purifying the village and revitalizing the forces o f life. T he case o f M atsu-bayashi resem bles Senzu m anzai o n a n u m b er o f counts. T h e perform ers are outsiders, itin eran t, visit at N ew Year, recite magical blessings, and are perceived as having a sacred status. F u rth e r­ m ore, th e role o f the itinerant p erform er as p a rt o f a divine distribution system is apparent. W hen the M atsu-bayashi came to o n e ’s hom e, he was literally delivering the sacred to o n e ’s d o o r along w ith a recognized insig­ nia for the sacred, a pine branch. In the next chapter, I show th a t this was also a com m on activity o f ritual puppeteers from the cerem onial center called N ishinom iya for they, to o , distrib u ted protective talism ans and placards for this m ajor shrine fam ous for puppeteers. By th e end o f the M urom achi period, the practices o f M atsu-bayashi and Senzu m anzai w ere being absorbed in to the broader context o f kadozuke rites. T h e popular perform ance th a t developed in this loosely stru ctu red settin g were, like those elsewhere in th e w orld, highly syn­ cretic. B ut th e basic religious stru ctu re— th e visiting outsider b ringing sa­ cred forces to the hum an com m unity— becam e a part o f the perception, h a lf conscious at tim es, o f kadozuke perform ers. In h e re n t in the Japanese popular conceptions o f the sacred is the idea th at sacredness, in its

raw, unchanneled state, is undifferentiated and potentially dangerous. As kadozuke perform ers becam e less form ally tied to religious centers and Sanjo districts, and as their duties becam e less prescribed, this n eg a­ tive potential was projected o n to the perform ers them selves, and this led in no small part to the heavy discrim ination they suffered during the Tokugaw a period.

K adozuke: A General Pattern o f A r r iv a l a n d D eparture N o t all kadozuke perform ances are o f an overtly ritual o r religious nature. Som e are m erely hum orous skits and sh o rt scenes from fam ous ballads. N evertheless, the context o f gate-to-gate perform ance absorbed m uch o f the highly religious prototypical perform ance traditions. W hile the list is som ew hat red u n d a n t, here are w hat can be considered the m ajor features o f the kadozuke c o n te x t.76 First, kadozuke perform ances were presented by itinerant artists. Second, because they were n o t m em bers o f the co m ­ m unities in w hich they were perform ing and w ere n o t situated in a familiar social setting, their nature was a mystery, and th u s a n u m b er o f attributes o f radical negativity were projected upon them . T h ird , as outsiders, these perform ers were regarded w ith a certain degree o f ambivalence and even fear o r loathing. T he itinerant perform er represented an unassim ilable d i­ m ension o f life— the realm beyond ordered space and m eaning— and was hence potentially dangerous. Expressions o f this can be seen in the claims th at these people were p art hum an and p a rt animal. T he fundam ental h u m an /a n im a l distinction cannot be broken in the norm al o rd er o f things, b u t people from the m argins were th o u g h t to transgress it r o u ­ tinely.77 F o u rth , the am biguity o f itinerant perform ers led to their being perceived as having special magical powers to cure illness, bring o r ward o ff epidem ics, cause or prevent calamities, and so on. Fifth, because itiner­ ant perform ers aroused apprehension, their arrivals and departures were carefully m ediated. Arrivals were tim ed and announced, and departures w ere n o ted w ith exchange o f gifts. Usually, the ho st in the village w here the artist perform ed w ould give the p erform er either food, m oney, o r tem porary lodgings. Far from being m erely paym ent for services re n ­ dered, this exchange o f m oney o r goods m ust be un d ersto o d as a ritual exchange betw een tw o realms o f existence: the settled realm o f everyday, profane life and the realm o f the sacred, represented by the itin eran t ritual artist. T hose living in the settled and established w orld o f everyday m ean­ ings (villages and tow ns) had need o f the services o f the ritual perform ers b u t were also afraid o f them . T he paym ent o f goods was in p a rt a gift to send the perform ers away. Sixth, elaborate systems o f m arking the per­ form ers as outsiders developed th ro u g h tim e, and the style by w hich these

outsiders an nounced th eir arrivals becam e highly stylized. F o r exam ple, a ritual p u p p eteer becam e identifiable by the m anner in w hich he carried puppets in a box and the way he announced his arrival w ith a flute o r a call. O th e r types o f perform ers had a series o f codified signs m arking them as yosomono (outsiders). T his process o f m arking becam e incorporated in to the signs and symbols o f the ritual perform ances, as I show in subse­ q u e n t chapters. Last, the arrival o f these perform ers was usually seasonal in nature. T he period betw een th e harvest and th e new year, w hen the fu tu re o f the cos­ m os was in question, was the m ost com m on tim e for their visits. This was a m arginal tim e w hen the new year was n o t yet fully form ed and the ord er o f th e cosm os was being recreated. H ence these days w ere regarded as pow erful and sacred. T he rites presented by these visitors served to h ar­ ness this sacredness and channel it for th e benefit o f th e com m unity or hom e, thereby revitalizing life forces, instilling g o o d health and pro sp er­ ity, prom ising large fishing catches, and so fo rth. In addition to the N ew Year period, there were also appearances at o th e r m om ents d u rin g the ritual calendar. F or exam ple, ritual puppeteers presented th e Shiki Sanbaso rite at w eddings and boat launchings, or w hen rice seedlings were transplanted to the paddies in the spring. W hat all these m om ents share— N ew Year, w eddings, boat launchings, planting tim e— is the quality o f transform ation and transition. T he kadozuke perform ance context challenges our usual assum ptions ab o u t ritual. As G erd Baum ann has p o in ted o u t, “We ten d to take it as a given, o n the w hole, th a t rituals are symbolic perform ances w hich unite the m em bers o f a category o f people in a shared pursuit th at speaks of, and to , their basic values, or th a t creates o r confirm s w orlds o f m eanings shared by all o f them alike.” 78 As B aum ann argues, o u r un d erstan d in g o f ritual carries on the trad itio n o f D urkheim , in w hich rituals assume h o m o ­ geneity. Leach echoes this n o tio n o f hom ogeneity: “T h e perform ers and the listeners are th e same people. We engage in rituals in o rd er to transm it collective messages to ourselves.” 79 B aum ann’s p o in t raises an interesting problem in our view o f Japanese rituals. T he standard in terp retatio n o f Japanese ritual is likewise predicated o n the n o tio n o f hom ogeneity: spe­ cifically, th e co n stru cted n o tio n o f a seamless people, undivided and living as “one family” w ith the em peror as father. As th e case o f kadozuke clearly show s, however, for a significant period o f Japanese religious history, and at a very popular level, a central ritual context in Japanese religions was driven by a rigidly m aintained distinction betw een the p erform er and the host. They were seen n o t simply as socially different; they w ere ontologically distinct. T he kadozuke p erform er (Senzu m anzai, M atsu -bayashi, D okum bS-m aw ashi, etc.), precisely because he was an outsider, was able to m ake th e ritual co n tex t w ork. T he do m in an t ideological system was

dependent on the existence o f this other not only as an entity against which to define itself, but as the symbolic nexus o f power capable o f keep­ ing the system viable, pure, and whole. Nowhere else could the ideologi­ cal function o f the outsider be more apparent. The signification o f the puppeteer as visitor, outsider, and other is more than just one aspect o f this ritual context. I t is the essential sign. H ow is this otherness generated and m aintained in the ritual? As we shall see in later chapters, the levels o f otherness go deeper than the social context and status o f the performer. The beings represented in the rites and the puppets themselves layer this alterity in a complex and multivalent fashion.

T h e D em ise o f th e K adozu ke C on text The ritual context described in this chapter was widespread throughout Japan until it began to gradually decline at the start o f the Meiji period. By the late 1940s, kadozuke rituals were almost completely defunct.80 As I shall show later, some o f these itinerant rituals were highly developed both aesthetically and dramatically. I was often told during my fieldwork that the “visits” o f ritual puppeteers were an im portant com ponent o f the festivities o f the New Year period. In fact, so com m on was the appearance o f Sanbaso-mawashi and Ebisu-kaki during this time that one inform ant noted, “ It just isn’t New Year’s anymore w ithout Sanbaso.”81 N everthe­ less, immediately following the war, there was a campaign throughout Japan to obliterate itinerant performance, including puppetry rites. O ne reason these practices were discredited was th at people regarded them as a kind o f begging, an activity which for a num ber o f reasons came to mean many things in the postwar period. Because the daily life o f m ost people was terribly difficult for a num ber o f years after the war, any activity that looked like asking for handouts was immediately discouraged and even violently p ut down. I was told the story o f one elderly man on Awaji who was beaten one New Year after the war when, at the request o f some people near his home, he made the rounds o f the village to present the Sanbaso ritual. According to my source, he was told he was just looking for a free handout. T hat was the last time he performed. The exchange o f goods (food, money, lodgings) was an im portant com ponent o f the ritual action being accomplished. A highly negative attitude tow ard the exchange o f goods at the end o f these rituals came up in my fieldwork during 1984. While on Awaji conducting the preliminary segment o f my fieldwork, I had the opportunity to interview an elderly man shortly before his death that same year. H e had been famous shamisen player o f the Awaji theater. In the course o f the interview, he men-

tio n ed th a t in the prew ar period, he had w orked for a while w ith a Sanbaso-m aw ashi as a m usician. As he was telling me this story, he simply said in passing, “T okidoki, m onom orai wo shita.” (Som etim es we did monom orai). M onom orai m eans “b eg g in g .” A t the tim e, it did n o t strike m e as at all o d d th a t a perform er should be given som ething in exchange for his perform ance, even if it was done from d o o r to door. T he old m an w ho to ld m e this saw this as perfectly norm al. L ater th a t evening, h o w ­ ever, my ho st on Awaji (w ho had been present at th e interview ) was deeply tro u b led by the fact th a t this m an had said he did m onom orai. “This makes Awaji p u p p e try look terrib le,” he insisted. This anecdote reflects a certain attitu d e tow ard itinerant perform ers, even in the village they called hom e. In the postw ar period, w hen all religious activity had to be rational­ ized, ritual exchange o f goods for services was considered to be simply begging, and therefore it was necessary to wipe it o u t. T here was an o th er serious concern (underscored by attitudes and p oli­ cies o f th e occupation forces) th at c o n trib u te d to the dem ise o f the kadozuke context. M any people regarded such ritual practices as supersti­ tious and prim itive. S um m oning spirits and appeasing deities was n o t com patible w ith a newer, m ore rational view o f religion as ethical behav­ ior. This general attitu d e tow ard religious life in Japan was expressed at b o th a popular and scholarly level. T he highly influential w ork o f N akam ura H ajim e, Ways o f T h inking o f Eastern P e o p lesf m etanarrative o f history if ever there was one!), w ritten in Japanese in 1948—49 and p u b ­ lished in E nglish in 1964, can be seen as a reflection o f the Zeitgeist am ong scholars. Referring to the entire m agical tendencies expressed in sham anic practices in Japan, he writes th a t “such sham anistic o r m agical tendencies will, w ith th e diffusion o f scientific know ledge, disappear sooner o r later. A fuller investigation is required concerning the problem o f the post-w ar g row th o f heretical religions o f this s o rt.” 82 This co m m en t o f Professor N akam ura’s is indicative o f th e c o n stru c ­ tio n o f a new, postw ar vision o f a rationalized religious w orld in Japan. C h ieftarg ets for elim ination o n the grounds th a t they w ere “ irrational and barbaric” were any rites w ith sham anic overtones. Professor N akam ura was expressing a cultural tendency o f th e tim es, b u t th e success o f his w ork as a w hole, w hich argues th ro u g h o u t for a pu rer B uddhism than is found in actual practice in Japan, suggests th a t b o th the denial and suppression o f nonecclesiastical form s o f religious expression was a pow erful force in postw ar Japan. T he co nstruction o f a new identity in th e postw ar years dem anded a denial o f all religious behaviors th a t did n o t agree w ith Japan’s new and reform ed im age as a rational democracy. This general negative attitu d e tow ard ritual practices in the postw ar period is n o ted by R o b ert J. Sm ith in his in tro d u c tio n to Y am am oto’s study o f th e N am ahage festival:

It seems that during and after the war village officials waged a campaign against the masked visitors o f N ew Year’s Eve, urging the people to abandon this supposedly barbaric and childish custom that had no place in wartime Japan or in the new Japan o f the post-surrender period. N o doubt the ceremony was denounced as a remnant o f the feudal past— an epithet routinely directed at any Japanese practice, belief, or custom that anyone wanted to discredit. Although the campaign against the observance was successful, and the villagers were per­ suaded to give up their festival, it was clear that they remembered it fondly as an occasion when everyone had a good time and no harm was done to anyone.83

In sh o rt, the renew ed “p ro te sta n tiz a tio n ” o f religious life in postw ar Japan m eant th at the intentions, contexts, and styles o f kadozuke rituals w ere no longer deem ed w orthy o f the effort needed to m aintain them . T here is an irony in these cam paigns to wipe o u t seasonal rituals w ith sham anic overtones such as kadozuke p uppetry rites. Less th an forty years later, the very activities w hich were discouraged and even flatly outlaw ed are now being reclaim ed as part o f an “authentic religious past” and “folk­ ways.” In the last chapter o f this book, I shall show how the Awaji p e o ­ ple’s decision to revive their past— and em brace their very otherness as evidence th a t they are “truly Japanese”— created an unusual m ixture o f nostalgia, guilt, and healing.

3 A C rippled Deity, a Priest, and a Puppet: K u g u tsu and Ebisu-kaki o f the N ishinom iya Shrine

A F ra g m e n te d H is to r y In this chapter, we tu rn to a discussion o f th e early puppeteers in Japan com m only called kugutsu, and exam ine the fragm ented evidence from the H eian period th ro u g h early Tokugaw a describing their activities. This term is the earliest in Japanese w hich specifically refers to m anipulated dolls, as opposed to static ritual objects such as the examples o f various kinds o f ningyo discussed in chapter I . Ideally, a history o f these ritual puppeteers could begin in the early p erio d o f state form ation in Japan and unravel as a rich narrative spanning nearly fifteen h u n d re d years, linlcing up the scattered references to k ugutsu in to one long story and show ing in exhaustive detail the ritual roles they played for the state, for villages and collective concerns, and for individuals. In such an idealized history, we w ould know p eo p le’s nam es and be able to d o cu m en t how they interacted, and we w ould have detailed descriptions o f the puppets and th eir ritual uses. Sadly, references to ritual puppeteers are usually m ade in passing and seem ingly refer to som ething so com m onplace th a t a description is seen as unnecessary. We have m any examples o f puppets b u t few records o f the people w ho m anipulated them or w hat they d id w ith th em w hen they “m ade th em d an ce.” Passing references, vignettes in p o e try and song, b rie f m entions in journals and travelogues— these tell us little. Like m any m arginalized groups in Japanese history, puppeteers have n o t always been recorded o r rem em bered. Ritual p u p p etry has suffered from being at once to o com m onplace and to o m arginal to gain the a tte n tio n o f h isto ­ riographers. H aving only fragm ented evidence in the history o f religions is n o t an unco m m o n predicam ent. T h e w ork o f an historian o f religions is n o t u n ­ like th a t o f an archaeologist trying to excavate th e past. Som etim es one finds a w hole p o t, b u t som etim es one recovers o n ly shards, which m ay possibly reveal a p attern o r a m o tif b u t are fractured nevertheless. Usually there is a m ixture o f b o th . B ut we m ust resist th e tendency to assume th at

th e fragm ented evidence accurately depicts a w orld. I am rem inded o f a carto o n I once saw. T h e caption was som ething like “A British schoolboy im agines ancient Greece and R om e,” and in the carto o n , people were walking ro u n d w ith incom plete torsos— the odd arm , leg, o r head m iss­ ing— co n ducting th eir business in incom plete buildings. I m ust adm it th at the evidence for a history o f ritual p u p p etry is frag­ m ented. W hat are we to do w ith the shattered bits o f clay th a t we have? We can only hope th at they may give hints o f a m uch m ore com plex reality o f w hich they are rem nants. As Jo n ath an Z. Sm ith w ro te, “T he historian in his w ork detects clues, sym ptom s, exemplars. H e provides us w ith hints th a t rem ain to o fragile to bear the b u rd en o f being so lu tio n s.” 1 So it is w ith m uch o f the evidence in the study o f ritual puppetry. T hey give hints and clues b u t do n o t yield a fluent narrative, perhaps because one never existed in the first place. Fragm ents may be able to w hisper stories to us, b u t are to o often silenced by the larger, m ore fluid, co n stru cted narratives o f m etahistories w hich give us the false im pression o f seamlessness. A study o f ritual p uppetry m ust try and make sense o u t o f these frag­ m ented pieces, fitting them end to end, trying to find w hich pieces make sense to g eth e r and are m ost likely to yield som e u n d erstanding o f p u p ­ petry in Japanese religious life. We m ust resist the tem p tatio n to make ritual p uppetry in Japan in to one single story, for th e evidence in chapter I dem onstrated th at there are num erous sources for th e ritual use o f effi­ gies and dolls in Japan, and m any o f these religious and magical ideas clearly continued to influence one an o th er th ro u g h o u t Japanese history. N ow here is the herm eneutical axiom th at “ all history is in te rp re ta tio n ” ever m ore clearly evidenced. N evertheless, the religious and perform ance idiom s o f ritual pu p p etry reveal a great deal a b o u t Japanese religious life and perform ance. In this chapter, I present and in terp ret som e fragm ents o f evidence, and I try to understand the forces th a t led to fragm entation. A lthough ritual p u p p etry was w idespread th ro u g h o u t Japanese reli­ gious history, the status o f puppeteers has fluctuated betw een respect and esteem and the perception th a t puppeteers are, because o f the nature o f the w ork they d o , h in in (non h u m an ). In docum ents relating to kugutsu, we can see evidence— som etim es faint and obscured— o f the signification o f puppeteers as outcasts. T he occupation o f puppeteers in religious rites for deities, especially Ebisu, has a direct bearing o n how they were per­ ceived and treated at different tim es in Japanese history. O u r discussion begins in the late N ara period (based on H eian and Kam akura period sources) w ith the earliest sources describing puppeteers. First we look at the role o f ritual puppeteers in a decisive battle (at least from a m ythical perspective) in early Japanese history, th e defeat o f the H ayato by the centralized governm ent in the early eighth century. N ext, O e Masaffisa’s often cited tw elfth-century source, K airaishi-ki, provides

us w ith evidence o f how puppeteers w ere im agined in Japanese literary discourses o f “ o th ern ess.” T h e process o f depicting these people as exotic and exaggerating th eir nom adic lifestyle served to define Japanese n otions o f “ the g o o d p eo p le” (rydm in) .2 N ext, we explore the evidence from the N ishinom iya Ebisu shrine, the center w here kugutsu groups began to congregate d u rin g the m edieval period.

M agicians in Battle: T h e U sa Shrine and th e H ayato R eb ellion o f 7 2 0 In the early eighth century, the centralized governm ent o f Japan was in ­ volved in an ongoing struggle to secure a stro n g h o ld on the m ajor island o f Kyushu. O ne o f the problem s in the gov ern m en t’s cam paign for h eg e­ m ony was a stro n g tribal resistance particularly am ong the H ayato tribes o f the O sum i and H yoga Districts. A ccording to sources from the Usa H achim an shrine in w hat is now O ita Prefecture, in 720 a m ajor m ilitary effort was launched to bring these tribes in to subm ission, and the Usa H achim an shrine played an im p o rtan t role in this cam paign. Sources from the shrine tell o f battles being led by “w arrior priests” from the shrine w ho, in the nam e o f H achim an, so u g h t the subjugation (and even annihilation) o f th e H a y a to . Shrine records tell us th at the H ayato were “ subjugated” th ro u g h a violent and p rotracted battle. Because o f its role in this im por­ tan t cam paign (along w ith the fam ous Usa H achim an oracles, w hich stabi­ lized the imperial line by dictating th a t only m em bers o f the imperial family could ascend the th ro n e), the U sa center began to rise in power. O f interest for o u r discussion is th e role o f m agician sham ans using p u ppets in the battle. A key source, from the Usa H achim an shrine Usa H achim angu Hdjo-e Engi, tells us th a t d u rin g the battle the H ayato se­ questered them selves in seven castles, and it was n o t possible to break th ro u g h their defenses u ntil ritual m agicians using puppeteers from the U sa shrine staged a p u p p et perform ance on the walls o f the castles. T he H ayato inside becam e so enam ored w ith the perform ances th at they let their guard dow n and w ere successfully attacked. In th e w ords o f the text, “T hey also m ade IittJe m ale p uppets dance. T he p u p p e t perform ance was so in teresting th a t the H ayato forgot a b o u t their hostilities and came o u t o f the castle to see the perform ance. They all surren d ered and were su b jugated.” 3 A ccording to this trad itio n , th en , ritual p uppetry played a p a rt in the subjugation o f the H a y a to . It is possible to in terp ret these Usa shrine m aterials as later glosses on an event th a t happened during Yorb (7 1 7 — 724). Given th a t such ritual m agic was com m onplace in early Japanese battle, however, it seems likely th at these texts reflect som e degree o f his-

torical fact. F or o u r purposes, w hat is m ost im p o rta n t is n o t w h eth er or n o t puppets were actually used, b u t rather the conviction, m entioned tim e and again in these sources, th a t ritual puppetry was pow erful e n o u g h to bring an entire tribe o f separatists unw ittingly and magically in to subm is­ sion. Ritual puppetry, far from being trivial or despised, was regarded as an extrem ely pow erful force. P resenting pu p p etry in this light was p art o f the em erging tradition o f ritual perform ance at Usa. T h e H achim an cult, although clearly a strategic ritual o u tp o st for the Yamato state, created its ow n set o f ideological problem s for the gov­ ern m en t w hen it was used to legitim ate violence. H achim an was by this tim e already a m ajor figure in am algam ated Shinto (or at this tim e, “p ro to -S h in to ” ) and B uddhist practice and doctrine, so the fact th a t the center was involved in an o u trig h t slaughter created a discrepancy b e­ tw een H achim an as a deity p ro tectin g the nation and H achim an as a bodhisattva. U sa sources claim th a t H achim an was the reincarnation o f E m peror O jin and th a t H achim an had been to n su red , becom e a b o d h i­ sattva, and was practicing austerities in the Kunisaki region o f Kyushu. P ut simply, th e question is how can a bodhisattva co ndone and even legit­ im ate slaughter? W hat happened after th e battle is o f even deeper interest for a study o f Japanese puppetry. Follow ing th e defeat o f the H ayato, Usa sources claim, an epidem ic raged in the area o f K yushu.4 A ttrib u tin g this epidem ic to the m alevolent spirits o f the defeated H ayato, priests o f the Usa H achim an shrine began an appeasem ent rite based on oracles and revelations from the deity H achim an himself. In this rite, puppeteers were again active. They ritually reenacted the battle and perform ed a sacred dance and stylized sum o m atch in w hich m ajor Japanese deities battled one another. This was in ten d ed as en tertain m en t for th e souls o f the defeated H ayato. A H d jo -e (literally, “ Rite for the release o f sentient beings” ) is prim arily a B uddhist rite, but in this context it shows signs o f being m ore o f an appeasem ent rite. In the rite, pu p p etry again plays a central role, b u t this tim e as a reenactm ent o f the original events. It appears th at the em peror, w hen invited to travel dow n to the region to get a b e tte r understanding o f w hat happened, heard how the puppeteers had played a role in the H ayato defeat. H e w anted th a t reenactm ent to be a p art o f the ritual appeasem ent. Its purpose, I w ould m aintain, was to legitim ate the centralized govern­ m en t and to consolidate its ritual relationship w ith th e H achim an center. A cting o u t th at alliance was at one level the w ork o f ritual puppeteers. Again, the Usa Hachima-ηβΰ Hojo-e Engr. In the next year, i n the reign ofShom u in the first year o f Jinki [7 2 4 C . E . ] , there was an oracle: “I, the god, as the retribution for killing many o f the Hayato, on separate years will do a H ojo-e. To lead the dead spirits to nirvana, and for

repentance o f o u r sins, we do the H o j5 -e .” So, follow ing this oracle, in the reign o f E m p ero r Shom u in T cm pyo 16 [7 4 5 ] on the fifteenth day o f the eighth m o n th , th e em p ero r gave the official tab let and for the first tim e the H ojo-e was perform ed. (Actually, it was first perform ed follow ing an oracle in the first year o f Jinki, tw enty one years before.)5 Even th o u g h th e great bodhisattva [H achim an] kills countless people, because he has an enlightened status, and does g o o d , there is a lo t o f m erit in his killing. T h e internal p ro o f is th at no ray is h idden, each gets b rig h ter an d brig h ter and there is no hiding the b rig h t light [presum ably a reference to H achim an again], As a result o f th a t oracle they began to perform Hojo-e in all the provinces. W hen the em p ero r journeyed to W am aham a from the tem porary palace to the floating palace, and he learned how the ritual perform ers had perform ed w hen they attacked the H ayato, he had them reenact those performances. So th a t is why in the rite, various dances— d ra g o n ’s head, b ird ’s head, lions, puppets, etc., are presented. T he perform ers go o u t to the fro n t o f the floating shrine and present music and dances. Everything is d one just as it was in the ancient tim es [i e., w hen th e hdshi perform ed to distract the H ay ato]. As for the H ayato, they are released as snails u n d er the floating w orship hall.

T he case o f the Usa H achim an H o jo -e provides tw o examples o f the ritual use o f puppets: First, it presents a case (possibly m ythical) in which ritual puppeteers use their magic to assist in battle itself. Such cases are n o t uncom m on in early Japanese history.6 It has been argued by a n um ber o f Japanese archaeologists th at ritual specialists w orked in battles w ith the Yamato m ilitary cam paigns in the early stages o f “state fo rm ation” in Japan.7 M agicians, including those using puppets, w ere com m only em ployed in battle to stren g th en w arriors o r to fascinate, co n tro l, and confuse the enemy.8 F itual perform ers also helped legitim ate the central governm ent by perform ing purification and land-calm ing rituals. This text suggests th a t the m agicians were S hugendo practitioners from the Kyushu center on M t. H iko, a center w ith strong ties to the Usa cere­ m onial complex. Puppeteers still participate in the H oj6-e rite held every four years in a revived “ traditional folk perform ing a rt” th a t has been resuscitated in the last decade. P uppeteers from tw o subsidiary shrines, Koyo and Kohyo, presented the p uppetry ritual until the tu rn o f the century. T here are no records indicating how early these tw o centers came to be affiliated w ith the Usa center as ritual centers. I t seems possible, however, th at their ritual specialization in p u p p etry may date from the N ara period around the tim e o f the H ayato Rebellion. This case is strong evidence th at puppeteers were held in high regard and w ere used as ritual magicians in m ilitary cam paigns. I t also indicates th at the role o f ritual puppeteers in appeasem ent rites was considered im-

portant enough to merit their inclusion in Japan’s oldest H ojo-e, the Usa Hachim angu H ojo-e. D uring this early period, puppeteers were equated with powers capable o f controlling enemies in battle, sacred forces, and even epidemic spirits. A num ber o f scholars, including Suzuka Chiyono, have looked to the Usa Hachim an case and the puppets o f the Koyb and Kohyo shrines as an origin o f the late H eian and medieval kugutsu groups. There is clearly a relationship between the spread o f Hachim an worship th ro u g h o u t the Inland Sea during the early H eian period and the proliferation o f religious rites with puppets as a medium. Ebisu rites, examined in detail in chapter 5, most likely have strong ties to the early Hachiman cult. Ebisu is a mari­ time deity, and it is possible that the worship o f this deity was incorpo­ rated into the Usa complex when confederated tribes formed the power base in the Buzen region in the protohistorical period. O ne o f the heredi­ tary lines in the Usa center, the Usa dan , was thought to have been origi­ nally made up o f maritime peoples, called by the generic term am a (sea people). Perhaps the attributes o f the deity we now know as Ebisu were absorbed from their religious tradition.9 O ur concern here, however, is n ot to unravel this impossible question o f origins. N o t only is such an inquiry impossible to prove, but determ ining the origins o f ritual puppe­ teers in medieval Japan does not go far toward accounting for their signifi­ cance and role in society. O ur question is o f an entirely different nature: we are concerned with how the valuation o f puppeteers in society changed over time, how puppeteers have had traditions o f meaning ascribed to them in Japanese history, how they have struggled to name traditions for themselves, what puppeteers are said to have done, and what the scanty historical evidence shows.

H eian Period Puppetry: Sources and M eanings As we noted in the opening chapter o f this work, the practice o f using hitogata, or body substitutes, to represent an individual for purposes o f purification was a widespread phenom enon during the early H eian p e­ riod. H ere, we will look at a later development in the history o f Japanese uses o f the represented hum an form, the kugutsu groups. These people manipulated effigies to create the illusion that they were coming alive, rather than regarding them as static ritual objects. Before looking at one o f the key texts thought to describe the early activities o f puppeteers, let us consider the term kugutsu and its possible derivations. The first appearance o f the w ord is in a gloss on a Chinese Buddhist text in the eighth century. Interpretations the w ord’s meaning

and significance for u n d erstanding the origins o f Japanese p uppetry are basically o f tw o types: the nativist theory, which insists on the uniqueness o f Japanese puppetry, and the foreign origin theory, w hich seeks to locate the origins o f Japanese pu p p etry a b ro a d .10 M any scholars suggest th at the characters used to w rite this w ord w ere attached to an indigenous w ord p ro n o u n ce d “k u g u tsu ,” and the key to un d erstan d in g th e history o f kugutsu can be found by looking at the derivation o f this pronunciation. R epresentative o f this approach is the w ork o f O riguchi S hinobu, w hose theories m aintain a nativist in terpretation. A ccording to O riguchi, kugutsu originally referred to a kind o f basket called kugu carried by the early sea people (a-muzoku) to collect seaweed and shellfish. Sacred specialists w ould carry their puppets in these same baskets w hen they presented their perform ances describing the events o f deities. L ater th e w ord for these baskets came to be attached to the p u p ­ pets. These puppets w ere, according to O riguchi, sacred vehicles o r resi­ dences for the deities (goshintai) o f these am azoku.11 This th eo ry seems to have no evidence to su p p o rt it, save the desire to construct a grand narra­ tive th a t w ould account for the am azoku as a “lost trib e .” 12 A contrasting interpretive strategy suggests foreign origins and argues th at the etym ology o f the w ord can be traced as far as G reece, suggesting diffusion n o t only for the practice o f p u p p etry b u t for th e very w ord kugutsu as well. T sunoda, perhaps the m ost widely used and least cited scholar w riting o n Japanese puppetry, has p o in ted o u t the sim ilarity b e ­ tw een the w ords for p u p p e t and pu p p eteer in C hina and E astern E urope. In C hina, the w ord for p u p p et is p ro n o u n ced k u ir u i, and the w ord for pu p p etry is kuorong, or kuotu. A m ong the “ gypsies” the w ord for pup p et is ku ki o r ku kli, pup p et th eater is kukiengero o r khelepen, and pu p p eteer is kukienrengero o r khelepaskero.13 C om pared w ith the Japanese w ord kugutsu, this seems evidence eno u g h to suggest som e Altaic contact in this w ord derivation. D onald Keene, probably basing his discussion on T sunoda, argues this: The earliest Japanese name for “puppet” was kugutsu, a word found in an eighth-century gloss on a Chinese Buddhist text. This mysterious name has intrigued scholars for centuries; it has variously been traced to a Chinese word for puppet, pronounced approximately kuai-luai-tzu in the same period, or to kuki or kukli, gypsy words which some claim were probably the origin o f both Chinese and Japanese terms. The Turkish kukla , and the late Greek koukla have also been cited as proof of the transmission o f the art o f puppetry from Asia M inor across the vast Central Asian regions to China, Korea, and eventually Japan. . . . The possibility o f foreign origins is intriguing, but the evidence is by no means conclusive.14

O ne small problem has to do w ith w hat part o f th e w ord m eans puppet. In som e sources, kugutsu is used to refer to p u p p et, while kugutsumuwashi (one w ho makes p uppets dance) is used to refer to puppeteer. In o th er sources, kugutsu m eans b o th p u p p et and puppeteer. Regardless o f the conclusions w hich could be draw n from inquiries o f this so rt, such philological research is n o t w ith o u t its problem s, for it is easy to invent linguistic evidence to su p p o rt a theory. Beyond raising the possibility o f a foreign influence on Japanese p u p p etry expressed in ety­ m ology, it seems futile to argue the p o in t further. Perhaps it is best to play it safe, along w ith Keene, and avoid a nativist quagm ire. H e writes w hat should be the last w ord on this m atter: “T he early history o f Japanese pu p p etry m ight be in terpreted entirely in term s o f a spontaneous, native developm ent, b u t probably continental influence was present even in the earliest stages.” 15 A lthough the term kugutsu (and the characters w ith w hich it is w ritten) referred to p uppetry th ro u g h o u t the H eian period, it frequently referred to traveling w om en w ho were entertainers and frequently p ro stitu tes.16 C onsequently, we can n o t assume th a t every tim e we see this term th a t it is a reference to puppeteers. I t does seem th a t m any o f these w om en were involved w ith itinerant perform ers o f som e so rt and were th e m ­ selves itinerant. Six verses ab o u t kugutsu dating from th e m id-tw elfth century (exactly w hen O e M asafusa was w riting his text) explored the lives o f these w om en. O nly his one text, ICuiruishi-ki, m entions puppets spe­ cifically, however. By th e late H eian period, perhaps because o f the popularity o f this text (to be discussed shortly), kugutsu becam e the com m on w ord for a w an ­ dering puppeteer. Just w hat these w andering perform ers did and just how they lived is greatly open to question. We have no reliable textual evidence describing the lives o r activities o f these m arginal people. Scholars have ten d ed to build an edifice based o n one eleventh-century text, to which we now turn.

O e n o M asafusa and th e Signification o f O therness T h e eleventh-century c o u rt scholar O e no M asafusa (1 0 4 1 -1 1 1 1 ) w rote a b rief description o f th e kugutsu groups o f his day, called Kuirctishi-ki (also p ro n o u n ced “ K ugutsu n o Ki,” alternate readings for the same char­ acters).17 T he d ocum ent, th o u g h t to be dated a b o u t 1070, is a m ere 320 characters in length, yet is perhaps the source th at Japanese scholars cite m ost often w hen describing th e history o f itinerant perform ance and p u p ­ petry in Japan. H ere is my translation o f the text:

K a ira ish i-k i (K u g u tsu n o K i)

(“A R ecord o f the K u g u tsu ”) T h e kug u tsu are those w ho have no fixed abodes, no p ro p er hom es. T hey pitch their ru g ten ts u n d er th e sky, follow ing w ater and grass, rather like the custom s o f th e n o rth e rn barbarians [o f C hina]. As for th e m en, they all use arrow s and horses to h u n t gam e. Som e juggle tw o swords o r seven o r balls or make peachw ood p uppets w restle. T hey can make it look as if th e p u p p ets w ere live hum an beings. T hese perform ances com e close to th o se o f th e C hinese transform ation artists.18 T hey can change sand and stones in to gold coins, and grasses and w o o d in to birds and beasts, (d azzlin g )19 p eople’s eyes. As for th e w om en, they make up th eir eyebrow s to appear sad and grievous. T hey w iggle their hips w hen they walk, flash devilish teeth w hen they smile, and use verm ilion p ow der on th eir cheeks. T hey act and sing o f licentious pleasures, and lure you w ith th eir sorcery and m agic. T h eir fathers, m others, and h u s­ bands know this and do n o t adm onish th e m , yet help them in their m eetings w ith travelers. T he travelers do n o t hate to spend a n ig h t in beautiful un io n w ith th em . I f they find th e w o m e n ’s charm s agreeable, they give th em lots o f m oney, clothes sewn o f brocade fabric, gold o rnam ental hairpins and lacquered boxes. T h ere is no one w ho w ould n o t accept this as w o rth cherishing. T hey d o n o t cultivate a single section o f land, n o r g ath er a single branch o f m ulberry.20 C onsequently, they have n o connections w ith the governm ent and are all landless people. T hey are strictly drifters. M oreover, they do n o t know w ho th e sovereign is. T h e em p ero r does n o t know o f them either, and they enjoy an entire life o f n o t being taxed. A t n ig h t, they w orship a lo t o f deities21 w ith d ru m m in g and dancing and a great deal o f boisterousness. T hey pray to a large-headed m ale doll for g o o d luck.22 In th e eastern provinces, th eir groups are very pow erful and daring around M ino, M ikawa, and T o to m i. In th e sou th o f the m ountains (in Sanyo) H arim a and along to the w est o f th e m ountains (in San’in) Tajima bands are next and th e Saikai bands are regarded as lowest. Som e o f th e know n w om en am o n g them are: K om i, N ichihyaku, San’zensai M anzai, K ogim i, M agogim i, etc. 23 T hey kick up the d u st w hen they sing, and all th e noise they m ake perm eates th e rafters. T he listeners soak the very tassels o f th eir hats an d are unable to calm them selves. Im ay5, Furukaw ayd,24 Ashigara, K ata-oroshi, Saibara, K urotoriko, rice planting songs,25 sacred songs,26 b o a tm e n ’s songs27 roadside songs,28 m ikata, folk songs,29 magical B uddhist chants30— I t isn’t possible to list all th e m ethods they use. T he kug u tsu are one o f th e things u n d e r heaven. H o w could one n o t be m oved by them?

Uses 0/K airaishi-ki by Japanese Scholars This text has been used in two ways. M ost scholars o f ritual perform ance and folk perform ing arts in b o th Japan and the W est use this d o cu m en t as a reliable historical record describing the puppeteers o f the H eian period (even th o u g h the text was using the term kugutsu in a m uch looser way, referring also to female perform ers). They allow for a “ co h eren t” m o m en t in the history o f ritual p uppetry th a t looks just like w hat O e no Masafusa w rote. T he skimpy descriptions o f the puppeteers in the tex t are read as ethnographic accounts, usually with little suspicion as to any o th e r agenda the au th o r may have had in m ind w hen preparing the piece or how this text may have participated in a w ider discourse. This optim istic and naive use o f this text is probably due to th e dearth o f o th er reliable sources, and m ost scholars tu rn to this sh o rt piece w ith relief after expressing their lam ent th at o th er sources are n o t available. O ften scholars m ention K airaishi-ki in passing as a source for m ore inform ation in discussions th a t shortchange early p uppetry altogether. Inoura and Kawatake, for ex­ am ple, simply m ention in their tw o pages on “T he O rigin and D evelop­ m ent o f P uppeteering” th a t “ accounts o f [puppeteers’] m anner o f living are to be found in a late eleventh-century d o cu m en t, K airaishi-ki or K ugutsu-m aw ashi no K i, w ritten by O e no M asafusa.” 31 Hayashiya Tatsusaburo, in his Chusei Geinoshi no K enkyu, is representative o f this approach to the text: “As perform ance artists o f the late ancient period, it is truly the kugutsu w ho show an extrem ely unique existence. A nd, fortunately, we have the w ork o f O e M asafusa, K airaishi-ki, concretely describing the life­ style o f these people.”32 H ayashiya’s next several pages discuss each section o f the sh o rt text and sum m arize the historical realities expressed, raising n o suspicions a b o u t the reliability o f the content. To see w here we w ould end up w ere we to accept the m ethodology o f Hayashiya and o th er folk perform ing arts scholars, let us consider a refer­ ence in the text th a t Hayashiya does n o t discuss: O e no M asafusa’s refer­ ence to “ the Chinese transform ation artists” (yulong m anyan zhixi). O e tells us th a t the kugutsu greatly resem ble these artists o f H an China. W hat did these artists actually do? A ccording to W u H u n g , a C hinese art h isto ­ rian, their perform ances “ began w ith one actor dressed as a lynx, dancing in a courtyard. W hen the lynx reached the fro n t o f th e palace it jum ped into a pool and transform ed itself into a flounder. C louds b urst from the m o u th o f the fish, obscuring the sun. By the tim e the clouds had dis­ persed, the fish had changed in to a dancing yellow dragon eighty feet long, w hose scales gleam ed and flashed m ore brightly th an the su n ­ lig h t.” 33 AU this, and in Japan, w ith peachw ood puppets! E xciting stuff, to be sure, b u t if we read this text literally, we w ould have to w onder why

such great theater is n o t m entioned elsewhere in Japanese records. This is but one example o f why Oe no Masafusa’s references cannot be taken at face value, however tem pting it may be to imagine feline-fish-dragon transm utations in ancient Japanese performances. Hayashiya’s interpreta­ tion m ust be attributed to the desperation and wishful thinking that a dearth o f sources may provoke in a researcher. On the other hand, we have those scholars, mostly Japanese, who raise objections to what are regarded n o t as ideological uses o f poetic license but rather as errors o f fact introduced by Oe no Masafusa. Scholars using this line o f interpretation point to two types o f errors. First, they contend that the author relied too heavily on fixed poetic forms and stock phrases from Chinese histories o f the period, in part to show his familiarity with these literary traditions and in part to play insider w ord games with his equally savvy audience (much the way we scholars today sprinkle our prose with French terms or references to current theories to show that we are both clever and part o f an in-group). Oe no Masafusa appears to have borrowed heavily from a m otif o f the “wandering nom ad” in Chinese ethnographic poetry. Some o f the lines in the text are lifted directly from popular Chinese chronicles and may n o t accurately describe eleventh-cen­ tury Japanese realities. As Tsunoda Ichiro has noted, at least one m otif in the text is taken directly from Chinese texts: the reference to “people w ithout fixed abodes following grass and water.” This exact same refer­ ence occurs a num ber o f times in popular Chinese texts—the Sui Shu and the Hou H an Shu—with which Oe was undoubtedly familiar.34 Yamaji Κ οζδ raises this objection: “In summary, the text describes a group o f itinerant performers. But, in this text, there is the influence o f Chinese poetry, and thus it cannot be th o u g h t o f as a simple description o f the realities o f the kugutsu as they really w ere.” 35 Donald Keene has also raised this concern, following such Japanese scholars as Tsunoda: “N o other record indicates the existence o f such n o ­ mads in Japan, and some scholars have therefore asserted that Oe Masafusa merely used stock phraseology borrowed from Chinese accounts o f foreign tribes to decorate his brief account o f the puppeteers. U ndoubtedly the choice o f words was influenced by Chinese examples, but we cannot disre­ gard O e’s general implication that the puppeteers led lives so unlike those o f the sedentary Japanese that they were taken for foreigners.”36 The second type o f error o f fact scholars point o u t concerns the rela­ tionship o f the kugutsu groups to governm ent forces. The text, the reader will recall, claims that kugutsu do not cultivate land or gather mulberry, that they have no connections with the government, and are untaxed landless drifters unaware o f w ho the sovereign is. Yamaji Κόζό argues against these lines o f the text, pointing o u t O e’s poetic license. “ Even if there were groups o f kugutsu who did n o t follow the existing tax laws,

those were the kugutsu o f a period prior to the late Heian (i.e., when the text was w ritten).”37 He backs this up with sources from the same period. Several docum ents from around the twelfth century show kugutsu groups in litigation over labor and tax disputes. In Fujiwara no M unetada’s docu­ m ent Chiiyiiki (which spans the period from 1032-1114) from Eikyw 2, (1114), there is a lawsuit dated the sixth day o f the fourth m onth in which a kugutsu-shi had taken horses and cotton from some conscripted shrine servants (kodoneri). Further, the diary Meijyetsukiy dated 1121, describes kugutsu being brought to trial for disturbing the peace. In both o f these cases, Yamaji points out, they requested what am ounts to some sort o f legal council or were within the jurisdiction o f the law. This pattern seemed to continue, since as late as 1249, on the twenty third o f the sev­ enth m onth, there was a lawsuit against M inam oto no Yoritomo by the kugutsu concerning conscripted labor or taxes for a ceremony. The case was brought before the Kamakura government. The kugutsu won the liti­ gation.38 While these exceptions to O e’s description do not commend the kugutsu as the most neighborly sort, they do raise a question about the text’s insistence on the legal liminality o f the kugutsu groups. It is not entirely possible to say that they had no access whatsoever to the legal system at the time, that they did not appeal to it when necessary, and that they were quite apart from established authority. This second way o f dealing with the text, namely, calling its flights o f fancy and “facts” into question, leads us to wonder if these scholars would suggest that other areas o f the text can be used as they stand. Can we trust descriptions o f the performances? Were their puppets really made o f peachwood? Are his discussions o f the coquetry o f the women reliable? Is it safe to assume his lists o f performances and places o f residence he assigns to these people are factual? These two strains o f scholarship on this text, one bordering on blind historicism and the other, arguing at the level o f factual error on two counts, leave us with two options: either we use the text as a reliable description o f the kugutsu or we pick from it the data that we want to use. Both seem to be based on an assumed translucence o f the text as a historical docum ent, something it probably was never intended to be.

A Third A lternative: Oe no Masafusa ys Signification of Otherness I propose a third way to read this text. While O e’s poem had to have enough basis in a recognizable reality to catch people’s attention, it is best read as a piece o f fanciful “exotica.” I suggest that this text created a dis­ course that cast the itinerant puppeteer as O ther in the Japanese imagi­ nation, and in so doing helped to define the meaning o f being mainstream

Japanese. T h e reader will recall how the ships o f fools o f medieval E urope w ere em bellished in literary reference. Such texts should be read n o t as reliable descriptions o f the tre a tm e n t o f the insane in E urope at the tim e b u t rath er as reflections o f how these m arginal people were appropriated by do m in an t discourses o f identity. O u r reading o f the kugutsu text should therefore n o t sim ply ask, “Is it factual?” b u t rather, “H o w did it participate in the m arginalization o f these people in subsequent Japanese history and w hat purposes did it serve?” AJI texts, in som e way or another, are highly constructed descriptions o f phenom ena, and one m ust bring the usual “herm eneutics o f suspicion” to bear on them . This case is no exception. T he specific o f this text reconfirm ou r view th a t it is n o t simply an ethnographic description o f a particular strain o f the Japanese p erfo rm ­ ing arts. W hat then is this text all about? This sh o rt 320-character text creates a fantastic w orld o f beautiful, se­ ductive w om en, w ho exert pow er over the m en w ho happen u p o n them , and virile, horse-riding, bow -and-arrow -w ielding m en, “ noble savages” w ho com m and the pow erful transform ation o f m atter using w ooden p u p ­ pets. Perhaps O e ’s kugutsu tell us m ore a b o u t w hat a noble c o u rt scholar (such as O e) was not (and perhaps w hat he secretly at tim es w ished he could be) than w hat a pu p p eteer was. This p a tte rn o f signifying otherness seems to be an im p o rta n t dynam ic in the m ovem ent tow ard the social discrim ination th at kugutsu groups suffered later. O e seems in te n t o n im pressing his eleventh-century readers w ith the follow ing features o f the k u g u tsu ’s otherness. They are n o t agrarian b ut resem ble hun ters and gatherers. T hey entertain them selves w ith exotic, sham anic types o f en tertain m en t using peachw ood puppets, w hich indi­ cate a connection w ith spiritual forces. B oth th e m en and the w om en, in different ways, are able to enact strange transform ations and weave webs o f m ystery and fantasy th a t entrap people w ho com e in to contact w ith th em — the m en th ro u g h perform ances w ith puppets, the w om en th ro u g h songs and overt sexuality. T h eir families d o n o t follow estab­ lished m oral codes o f behavior as regards the p roper activities o f daughters and wives. T hey are outside th e law and have no responsibilities to the governm ent. Lastly, they live in out-of-the-w ay places scattered along the w ater and land routes o f Japan. E verything ab o u t them suggests the m argins o f society. “W hat is special a b o u t peachw ood?” the reader may ask. Scholars o f ritual p u p p etry in Japan are always looking for peachw ood puppets. W hen I traveled in Japan d oing research and looking at lots o f different puppets, I was incessantly asked, “H ave you found any puppets m ade o u t o f peach­ w ood?” B ehind this question, it seems to m e, is the desire to find the validation o f O e n o M asafusa’s description, alm ost so th a t we can breath a sigh o f relief th at in fact this text can be used as evidence sans inter-

pretation o f its intentions. Peachwood, as we know, was a wood used in ancient China in purification rites; this use o f peachwood was in tro ­ duced into Japan in the Nara period and is recorded in Ε ηβί-shiki. Book 16, item 8 lists peachwood staffs as a requisite substance for the Na festival o f driving out noxious vapors from the country.39 The English translator o f this text, Felicia Bock, notes that “the bows and staffs o f peachwood demonstrates the Chinese belief in the efficacy o f the wood o f the peach tree for warding off evil.”401 have yet to come across a peachwood puppet in my research, and although this lack o f evidence may simply mean that no puppets from the Heian period have survived, it is also possible that they were never made o f peachwood in the first place. Issues such as choice o f wood are likely to remain fairly consistent am ong artisans, and we would expect to find some puppet heads made o f peachwood had this been a significant feature o f the puppets. It is possible that Oe no Masafusa was familiar with the association o f peachwood with divination and purification, and mentioned it to add mystery and otherness to his description.41 If this text was on one level part o f a genre o f fantastic literature, some­ thing like Ripley’s Believe It or N ot, what was its purpose? It calls to mind the stories out o f the urban legends genres, which tell o f fast women and wild men who usually meet terrifying ends. While this text spares the terri­ fying ends, it could be read on another level with attention to two dom i­ nant features o f the text. In one sense the text can serve as a mirror held up to polite Japanese society. It can be read somewhat loosely as a “what nice people (rydmin) d o n ’t d o ” sort o f text. I propose that a playful, sub­ text in mirror opposite o f the Kairaishi-ki m ight look something like this: T he ryom in are those w h o have fixed abodes, proper h om es. T h ey sleep under roofs, stay w here their crops and jobs are, and are d ecidedly unlike northern barbarians. As for the m en, they use sorobnn and brushes to w rite o u t accounts and litera­ ture. Their activities involve n o slight o f hand, n o tricks. T hey dazzle n o one with their activities. (T hey are honest, hardworking p eop le.) T he w om en d o not overly alter their appearance with makeup. T h ey walk in a polite m anner and cover their m ouths w hen they sm ile. T hey discuss polite m atters, and their fathers and m others are aware o f what they d o. T hey d o n ’t spend the n ight w ith strangers and never receive extravagant gifts from m en. This is the normal way things go. R yom in work the land and grow silkworms for cloth. They have con n ection s w ith the governm ent and are tied to the land. T hey stay put. T hey know w h o the sovereign is, and the em peror is aware o f their existence. T h ey pay their taxes. T hey w orship their deities in orderly rites. Their deities are decidedly

n o t phallic. T hey live in p o p u lated areas w ith arable land and avoid out-of-theway places and passageways. T hey d o n o t live in the m ountains. W ell-know n ryom in w om en d o n ’t drive m en to distraction and frenzy with singing and dancing. T hey use w ell-know n and suitable m ethods to attract th eir m en. R ydm in are one o f th e things u n d e r heaven. T hey may n o t m ove anyone, b u t aren ’t you glad y o u ’re n o t a kugutsu?

By offering this text with tongue in cheek, I am suggesting that the original be read as fantasy, based only in part on perceptions o f w hat itin­ erant performers, puppeteers, singers, and prostitutes may have been doing in the eleventh century. The Kairaishi-ki text serves to reinforce the status quo, and even create a reality, by carefully describing everything a literate person reading the text is not. As Keene has pointed out, there are no other texts from the period describing such nomadic groups in Japan. And we certainly do n o t have references to anything as exciting as peachwood puppets enacting the kinds o f transform ation presented by such Chinese artists as the yulong m anyan zhiyi. Based upon this short text, however, all subsequent scholarship on itinerant puppetry in Japan has m aintained the existence o f these distinct perform ing groups as a single tradition, moving about the margins o f Japanese society. Scholars have looked for confirmations o f O e ’s description, and when the sources have n o t surfaced, have stuck by their text. It would be going too far to say that Oe no Masafusa imagined the very existence o f these performers. Their existence is not a question o f debate. H aving confronted the dearth o f medieval sources myself {and been tem pted by the promising title o f O e’s text as well), however, it is clear to me that the really interesting area o f inquiry regarding this text is how it has com m anded such prestige as a docum ent. The text came to have ideo­ logical clout in later Japanese discourses o f alterity. The w ord kugutsu, rather than merely referring to a person who manipulated puppets, came to mean a person who fit the description in Oe no Masafusa’s text. Ritual puppeteers, who in fact perform ed a large num ber o f different functions and who most likely lived a variety o f lifestyles, came to be th o u g h t o f as a distinct group. They had been nam ed, and they had become a tribe. In short, in this text we see the beginnings o f the invention o f the O ther—in this case itinerant performers—in Japanese history. W hat at first appears to be a clever w ord game mixing observation o f prostitutes, descriptions from Chinese ethnohistories o f wandering nomads and transformation artists, and references playing on word titles from current works in Japa­ nese literature at the time, in the end becomes the launching o f a discourse about marginality.

K a ira ish i-k i an d the Bodies of M edieval Itin era n t Women

A nother im portant aspect o f this text is its reference to the bodies o f women, which takes up more than two thirds o f the piece. Women make up their eyebrows, wiggle their hips, flash their teeth, p u t rouge on their cheeks, and accept fancy clothes from men in exchange for sexual favors. AU this is hardly unique: the text m ight as well be describing a prostitute from Colorado Boulevard in Denver or Sannomiya in Kobe. The text of­ fers little inform ation about the women beyond these standard (and, it would seem, timeless) images o f “ bad girls.” The text is grammatically divided into two parallel (though unequal) parts, one about men and one about women. Unlike the men, who weave their magic with puppet b o d ­ ies, these women weave their magic with bodies that are their own—or perhaps created for them by O e no Masafusa. This process o f signification calls to mind an expression “signifying is worse than lying,” used by Charles H . Long in the introduction to his Significations: Signs, Symbols and Images in the Interpretation o f Religion. Long, an Afro-American who grew up in Little Rock, Arkansas, writes, From the colloquial and slang expressions o f my youth I learned som ething about the forms o f linguistic expression. Signifying is worse than lying because it obscures and obfuscates a discourse w ithout taking responsibility for so doing. This verbal misdirection parallels the real argument but gains its power o f meaning from the structure o f the discourse itself w ithout the signification being subjected to the rules o f the discourse. As a matter o f fact, the signifier may speak in agreement with a point o f view, while the tone o f voice creates doubt in the very act and words o f agreement. Or the signifier may simply add com m ents that move the conversation in another direction. Or the signifier will simply say a word or make a com m ent that has nothing to do with the context o f the discourse, but immediately the conversation must be formulated at an­ other level because o f that word or phrase. Signifying is a very clever language game, and one has to be adept in the verbal arts either to signify or to keep from being signified upon.42

Clearly, Ce no Masafusa was adept in the verbal arts and was “ signify­ ing upon” kugutsu for reasons that probably had as much to do with poetry as a game as anything else. His introduction o f Chinese phrases, his exaggeration o f the itinerancy o f these people, and even his plays on words and character compounds with references to Japanese poetry (such as Ryojin Hisho), suggest he was creating a clever game. M ost likely his con­ temporary readership knew and appreciated this. If we use a dialogical approach to the relationship between Oe no Masafusa, aristocratic society in Japan (the audience for his text), and the actual people he may have encountered who manipulated puppets and sang and danced for audi-

ences, we see a clear example o f signification for all subsequent itinerant puppeteers and female perform ers. W ho they are, w hat they do, and how they behave and relate to established au thority has been w rit in stone by a co u rt po et, w ith o u t any co n trib u tio n from the people them selves. T hese perform ers now have a nam e, and th e text clearly is w ritten gram m atically as a poetic definition (or, to use L o n g ’s expression, a signification) o p e n ­ ing w ith the w ords, “T he kugu tsu are those w ho . . . ” and closing with “T he kugutsu are one o f th e things u n d e r heaven.” By looking at this text in this m anner, we can account for a n u m b er o f puzzling facts o f the historical record. First o f all, why do we n o t see m ore references to bands o f kugutsu roam ing the countryside? I t w ould seem th at such groups w ould have left m ore th an the few passing references to puppeteers th at occur in the historical record. Second, why do we begin to see a radical shift in th e social status o f itin eran t puppeteers from the late H eian period on? T h e answers to these questions can be found, at least in part, n o t in gaps in the historical record b u t in the process o f signification in w hich this celebrated text clearly participates. O e created a reality, and subsequent to this text, all itinerant puppeteers were re ­ garded w ith a preset and defined conception: itinerant pu ppeteer = O e no M asafusa’s definitive description in his text. We find no record o f similar groups sim ply because they did n o t exist. C onsciously o r n o t, at the h eart o f his signifying process was a description o f w hat it m eant to belong to O e ’s class and to be Japanese. As these issues o f national identity becam e m ore p ro n o u n ced , so did the need to ostracize those w ho had already been labeled as being outside this circle o f m eaning. A t this p o in t in Japanese history, if we adm it th at at best O e ’s account describes a partial reality (if n o t a fantasy) o f the otherness o f the kugutsu, we do n o t see evidence th at these groups were regarded w ith disdain. O n the contrary, the text seems to suggest th a t they were the object o f fantas­ tic curiosity and pleasure, at least to poets and scholars like O e no Masafusa. Read as a literary gam e, it is a fun and clever piece. As history, it is som ething else entirely. Every tim e we search th ro u g h Japanese history asking for confirm ations o f this text, we have fallen prey to a slight o f hand— n o t th a t o f a pu p p eteer or a Chinese transform ation artist, b u t th at o f a scholar w ho tu rn s m en and w om en in to ink and paper, and ink and paper in to gold coins.

S u bseq uent References to K u jju tsu A fter C e no M asafusa’s text, we have a period o f nearly tw o h u n d red years before an o th er reference to kugutsu (as puppeteers) appears. T he midth irte e n th -c e n tu ry text C hiribukuro (Bag o f D ust) m erely confirm s th at kugutsu had ceased to be a source o f fascination and w ere no longer doing

performances. The text informs the reader that although once kugutsumawashi had done all kinds o f entertainm ent, their women now were no more than prostitutes and the men were employed killing animals.43 In the fifteenth century, a reference to tekugutsu (written with the char­ acters “hand kugutsu’’’) in the diary o f Sadafusa (em peror G o-H anazono’s father) mentions that tekugutsu came and presented Sarugaku, a popular performance style. The entry is dated the twenty-fifth day o f the third m onth o f 1416.44 C. J . D unn queries w hether this can be taken as a refer­ ence to puppets at all, as it may have been a confused reference to tesurugaku, meaning Sarugaku by amateurs. For nearly four hundred years after Oe no Masafusa’s text, we have negligible references to kugutsu. W hat happened to puppeteers during this time? Perhaps the Chiribukuro was right and their descendants were busy eking out other livings. Or perhaps they continued to manipulate puppets, but there were no records. We will never know. But we should not imagine their lives to be as fragmented as the evidence. N or can we immediately assume that any reference to a kugutsu in the late medieval period implies a continuity o f tradition with the late H eian phenom enon. For studies o f the early history o f puppetry as dramatic entertainm ent, one o f our best sources o f inform ation is the num erous period drawings o f puppet performances, often showing detailed views o f stages, methods o f manipulation, audience responses, and general ambiance.45 For itinerant puppetry, we have fewer drawings with less detail. But those we do have, scattered throughout various sources from the sixteenth through the eighteenth centuries, give us some information about how itinerant p u p ­ peteers worked, for whom and where they perform ed, what their puppets looked like, and how they manipulated them. Since these were low-rank­ ing performers, we do n o t know where they came from. They are u n ­ named in the pictures, but many, we can assume, came from the Nishinomiya shrine, a center for puppetry to be discussed in this chapter. The earliest drawing o f an itinerant perform er is from a scroll called Muchidu-ke kyiizo rukuchii rukugeti, depicting events in the outer parts of the Machida family estate from the first through fifth years o f Daiei (1521-25).46 The hazy picture shows a man holding a small box at waist level in which appear to be four or five small dolls. H e wears a straw hat and trousers to his knees, as well as a haori o f some sort. H e appears to be standing on the street. The next set o f references we have is a series o f four pictures in a scroll called Uesugi-ke kyiizo rakugai, probably drawn in Tensho 2 (1574).47 The first drawing shows four people near a gateway to a home. Since all are barefoot, we are to assume they are all performers, since going barefoot indicates low social rank. They appear to be passing through the gate onto a footbridge. The one in front has the lower half o f his (her?)

iissss

These drawings from a scroll describing activities at the U esugi estate (probably date 1574) show itin eran t puppeteers. N o te th e location o f th e perform ances at a gate and outside a hom e.

face covered w ith a veil and bears on his head a box w ith a small doll p ro tru d in g from it. T h e person in th e rear in the draw ing is possibly a m em ber o f the hou sehold show ing the perform ers to the street. W hile we learn little ab o u t the puppets in this picture, th e a tte n tio n to th e perfo r­ m ance context is striking, as the entire picture is fram ed by the gate. T he reader will recall the discussion o f gates as perform ance spaces in the p re­ ceding chapter. T he second draw ing from this collection underscores this description. I t is a side view o f tw o veiled persons w ith boxes on their heads, in which we see puppets. They are walking in single file on w hat appears to be a street. T he th ird draw ing, also from this collection and quite detailed, shows a crow d o f people w ith children in tow and on their backs, standing outside the entryway o f a hom e. A puppeteer w ith a box h u n g in fro n t of him is standing just inside the d o o r o f a hom e in the p art o f the house that later came to be called a genkcm. H e is perform ing for tw o seated people up in the house; his face is n o t veiled, suggesting th at he is from a different tradition o f puppetry than those in the previous pictures. T he box c o n ­ tains at least tw o small p u p p ets.48 T he style o f p u p p et m anipulation o f this draw ing was called kubi-kake (hanging from the neck), b u t the puppeteers were also som etim es called hako-ma-washi.^ A fo u rth draw ing from the U esugi m anor scroll also shows veiled perform ers holding w hat appear to be puppets on stands in front o f them . This draw ing clearly indicates th at they are perform ing at the gates o f houses, and it appears to be the N ew Year period because p o tte d pine trees are in fro n t o f the gates. Small children are w atching on the streets. T he next draw ing we have o f puppeteers is from a scroll at the M itsui-ji E nzokuin.50 T he scroll is th o u g h t to have been prepared to d o cum ent a festival com m em orating th e com pletion o f the rebuilding o f the tem ple E nzokuin in G enna 5 (1619). O f all the draw ings considered here, this one shows the m ost a tte n tio n to detail and appears to be th e w ork o f a highly skilled illustrator. I t shows tw o m en h o lding elaborate stagelike boxes in fro n t o f them , clearly w orking to g eth e r and m anipulating p u p ­ pets by p u ttin g their hands th ro u g h the back o f the box to hold their small dolls. O nlookers include a person carrying a tray o f fish, a child, and tw o samurai. Because the design o f the stages closely resem ble N o h stages o f th at tim e, it is th o u g h t th a t these hako-m awashi were actually doing m in ­ iature N o h perform ances using puppets, suggesting th at this may have been one o f the m any traditions o f hako-m awashi th a t were to be found during the sixteenth and early seventeenth centuries. References in texts at this tim e referring to “N o h ayatsuri” (“m anipulated N o h ,” ayatsuri being a w ord th at came to denote pu p p et and puppeteer, alm ost replacing the term kugutsu) were probably to this practice.51

A nother draw ing, from an u n d ated text called M atabee fu ry u r a k u z u , appears to be very similar to the previously discussed draw ing, except th at it shows puppeteers from the rear.52 From this draw ing we learn th a t these puppeteers w orked as a team , and b o th have boxes h u n g around their necks. O ne box is m ore elaborate than th e o th er, suggesting th at if the N o h th eo ry is correct, perhaps one p erform er m anipulated the m ain char­ acter and the o th er the m inor character. T he tw o puppeteers in this draw ­ ing are p erform ing before the w indow o f w hat appears to be a fu su m a (papered d o o r) shop, as the m aterials for spreading the paste are in view. W om en are leaning o u t o f a w indow next d o o r looking on at the p erfo rm ­ ers in the street. T he next draw ing from a scroll called Yoshikawa-ke kyiizd rakuchii rakujyai is also from th e early G enna era, and shows tw o perform ers (possibly even th e same tw o from the previous draw ing).33 Again they are w orking as a team , each holding a box and inserting his hands to m anipulate tiny puppets th ro u g h the back. They are standing on th e street at the entryw ay to a hom e, and w om en lean o u t o f the house to w atch. O n e o f them holds a baby, and a small child also looks on. O n e last draw ing, from a text called Konokoro-£usa, dated T enna 2 (1 6 8 2 ), shows a closc-up o f a kugutsu-m aw ashi.54 H e holds a tw o-tiered box in fro n t o f him and is clearly presenting an intricate perform ance in ­ volving a fox. T he puppets are at his eye level. AU o f these draw ings p o in t to a w idespread practice o f itinerant p u p ­ petry, w hich was apparently becom ing m ore and m ore popular by the sixteenth century. T he pictures make it clear th a t th e kadozuke context was com m on for the perform ance o f puppetry. A lthough we get some idea o f the kadozuke’s appearance and the perform ance context, it is hard to tell from the pictures w hat kinds o f perform ances they did and how they were understood. F o r this inform ation, we have to rely o n o th er kinds o f sources. To learn m ore ab o u t these puppeteers, we may examine references to a type o f p u p p etry th a t gained popularity at this tim e, Ebisu-kaki. By the m iddle o f th e sixteenth century, we begin to have m ore reliable textual trickles o f inform ation indicating the rise o f puppetry. Passing ref­ erences in co u rt diaries give us little inform ation b u t confirm th a t p u p ­ petry rites for th e deity Ebisu w ere becom ing com m onplace. A diary dated T enbun 24 (1 5 5 5 ), th e tw enty-fifth day o f the second m o n th simply states, “An Ebisu-kaki cam e.” 55 A n o th er diary from this tim e o f ladies at c o u rt called Oyudono no Ue no N ik k rt6 indicates th a t perform ers called Ebisu-mawashi o r Ebisu-kaki w ould com e from tim e to tim e to the palace. These b rie f entries begin in the second year o f Koji (1555) and continue th ro u g h T ensho 18 (1 5 9 0 ), indicating the seasonal appearance o f these Ebisu perform ers at c o u rt.57 T he diary notes th at d u ring the N ew Year

This drawing from Konokoro-gusa, dated 1682 (Tenna 2), shows the new level of intricacy in itinerant puppetry performance. period o f 1590, Ebisu-kaki were com ing nearly everyday.58 Sadly, we get little idea o f their perform ance. Let us tu rn therefore to the center from which these perform ers came, the Nishinom iya shrine, and determ ine their relationship to the deity Ebisu. Exam ining the nature and develop­ m ent o f Ebisu and other inform ation relating to puppeteers from Nishinomiya should provide some idea o f the intentions and c o n te n t o f these perform ances. Since it is from this center that Awaji puppetry eventually developed (and against this center that Awaii puppeteers define th em ­ selves), these Ebisu-kaki are an im portant source for understanding the rise o f ritual puppetry on Awaji as well.

A ffiliation as N onecclesiastical Sacred Specialists: Kugutsu at th e N ishinom iya Shrine N ishinom iya, located in the present-day city o f Nishinom iya in H yogo Prefecture betw een Osaka and Kobe, has been the center o f Ebisu worship from at least the Kamakura period. T he original shrine on this site, H iroda (or H irota) was included in a list o f shrines in M ontoku Jitsuroku dated

Kasho 3 (8 5 0 ), indicating there has been a shrine on this site since the n in th c e n tu ry /9 T he significance o f the place-nam e nishinomiya (literally “w estern shrine” ) is puzzling. It may simply have referred to the fact th at this shrine was situated on the road west o f the capital. O ne hypothesis suggests th at this shrine is the w estern c o u n te rp art o f Ise (the “ eastern shrine” ), and just as Ise is th e shrine o f the sun m aiden A m aterasu O -m ikam i (hirum e, literally “ sun m aiden” ), this shrine in N ishinom iya is the worship site for the Leech Child ( hiruko, translatable also as “sun la d ” ). I t has also been suggested th at perhaps this shrine was originally a w orship site for the ro u g h spirit m anifestation (a ra -m i-ta m a ) o f A m aterasu, while her nigim i-ta m a (blessing spirit) is w orshipped at Ise.60A fu rth e r interp retatio n o f the place nam e suggests th at this was originally the site o f an imperial to m b , and th a t female sham ans capable o f com m unicating w ith the dead w ere attached to th e original shrine on this site, the H iro d a shrine (which later becam e a center for w om en o f pleasure and their p u ppeteer c o n ­ so rts).61 Speculation on the m eaning o f nishinomiya, raises m any in terest­ ing theories, no n e o f which in the end can be proven.62 Som etim e after the tw elfth century (th o u g h no clear dates are available) a Sanjo district developed at the N ishinom iya site. In his definitive w ork on the early history o f the N ishinom iya kugutsu, Yoshii Taro (a hereditary priest from the N ishinom iya shrine) writes th a t this Sanjo district, a b it to the no rth w est o f the m ain N ishinom iya Ebisu shrine precincts, was the abode o f kugutsu from the early medieval period, and th a t the people who lived in this Sanjo district suffered from a low er status than others. T he area was considered a p o lluted district, and w om en w ould go there to give birth. N ishinom iya shrine records from as late as the G enroku era ( 1 6 8 8 1704) indicate there were th irty to forty households in the N ishinom iya Sanjo district which listed their sole occupations as puppeteers.63 I t was also in this district th a t we find the shrine dedicated to th e p atro n deity o f puppeteers, H yakudayu.64 In th e preceding chapter we discussed the symbolic significance o f these Sanjo districts as lim inal areas in th e geography o f m eaning in m edi­ eval Japan. T h e people living in this Sanjo district were involved in the w orship activities o f th e shrine and perform ed the appeasem ent rites o f the deity Ebisu. They used puppets as goshintai, o r spirit bodies, o f the deity in th eir rites.65 These Ebisu -kaki o f the co u rt diaries were actually conscripted sacred specialists living in the Sanjo district attached to the Settsu Nishinom iya Shrine (usually called N ishinom iya Ebisu Jinja).66 They were the p recu r­ sors o f the later dram atic tradition o f ningyo jo ruri, in w hich the art o f ballad recitation, sham isen playing, and pu p p et m anipulation were co m ­ bined to create a new dram atic form in the early m odern period. They

were also the group from which the puppeteers w ho form ed the Awaji puppetry tradition splintered. In the rest o f this chapter, we will focus on the significance o f the deity Ebisu as the liminal being par excellence in Japanese religions, and the relationship betw een this deity and the kugutsu o f the N ishinom iya center. An obvious question opens o u r dis­ cussion. In Japanese religions, there are myriad kami. Why, th en , do we n o t see puppeteers specializing in rites for every one o f the o th er deities? W hy n o t an A m aterasu-kaki or a Susano-kaki? W hy n o t a H a ch im a n -kaki o r an Inari-kakH To understand why there are special p uppetry rituals for the deity Ebisu, it is necessary to appreciate som ething o f this deity’s lim i­ nal and dangerous nature. Since detailed sources in English for a study o f Ebisu are n o t available, I present an overview o f the im p o rtan t features o f this deity before tu rning to a discussion o f the N ishinom iya kugutsu.

E bisu Worship: P ossib le D evelop m ent W hy did Ebisu need to be handled by puppeteers? W hat was it about this deity th at required this ritual treatm ent? W hat were ritual puppeteers try ­ ing to accom plish w hen they traveled the country presenting Ebisu rites? In chapter I , I noted th at puppets are beings once rem oved from the hum an realm. They are objects capable o f being possessed by sacred forces, b ut they rem ain separate from hum an beings. In the following discussion o f Ebisu, we will see th at Ebisu w orship is fundam entally about the issue o f containm ent. This dangerous deity, capable o f causing any n um ber o f terrible problem s in hum an society from disease to d ro u g h t, needs to be localized, channeled, appeased, and entertained. A pu p p et in this case serves as a control for spiritual contagion. A disease m etaphor is apt in this case, as we shall see. R ather than representing E bisu’s power w ith a hum an being, an interm ediary body betw een the sacred specialist and the deity provides a protective ritual space for the resolution o f E bisu’s problem atic and am orphous nature. T here is an intentionality in the representation o f the unrepresentable— and unpredictable— deity Ebisu w hen a puppet is used that w ould n o t be present were a hum an actor to “im personate” o r becom e possessed by Ebisu. Puppets then are n o t merely aesthetic choices in this case; they are ritually necessary. An im portant source o f inform ation for the study o f som e Japanese deities is the shinzo (shrine sculptures) and the o-fuda (iconographic d e ­ pictions on placards) used in worship. These sculptures and pictures have served a num ber o f ritual functions in Japanese w orship. O ften shinzo are spirit vessels ( mitamashiro or goshintai) and lodging places for the kami w ho is sum m oned during rites. Som etim es, a picture o f the shinzo is dis­ tributed as an am ulet from the shrine in the form o f an o-fuda. These

placards were frequently distributed to worshippers so th a t they could p u t it o n their ow n altar at hom e. They w ere un d ersto o d to participate in the sacredness o f the central im age o f the shrine, and they helped to distribute th e pow er o f th at center. A m ajor role o f the Ebisu-kaki setting o u t from N ishinom iya was to distribute these placards for w orship purposes th ro u g h o u t the countryside. T he p u p p et body became a sort o f traveling goshintai, a physical object in w hich the deity was actually present. In this way, ritual puppeteers taking th e sacred body o f Ebisu from the cerem onial center in to the periphery were literally spreading the presence o f the deity around the countryside, and the o-fuda distributed were a source o f revenue for the shrine, as were th e perform ances o f the puppeteers. T here are n o clear records indicating ju st how m uch revenue these puppeteers generated for the cerem onial center, b u t it is likely th at in addition to augm enting the prestige o f Nishinom iya as the center (which w ould have som e m onetary value), the itiner­ an t puppeteers attached to this shrine did actually engage in fundraising o f som e so rt.67 As the o-fuda had depictions o f Ebisu, we can use these plac­ ards as evidence for understanding how Ebisu was portrayed. D epictions o f Ebisu show rem arkable variety. T he earliest record o f any depiction o f Ebisu is a passing reference in a text dated the first year o f C hokan (1 1 6 3 ) at the Iw ashim izu H achim an shrine. It is n o t an actual picture b u t m erely a reference to the placard from this shrine: “As usual, Ebisu holds a fish,”68 suggesting th a t by this date th e association betw een Ebisu and fish was already well established. Perhaps this was always a cen ­ tral feature o f Ebisu w orship. From the m edieval period o n , Ebisu has usually been depicted as a dark-faced, bearded, obese old m an ho ld in g a sea bream u n d er one arm and either a fishing pole o r a sack o f m oney in the o th er.69 A lm ost invaria­ bly, th e figure is smiling— a deity o f the m o st docile and dom esticated nature. In keeping w ith this im age, Ebisu is usually described in encyclo­ pedias and dictionaries o f Japanese deities as “one o f the seven gods o f luck, w ho bestow s blessing o n fishing and business.” T he history, iconography, and ritual invocation o f this deity, however, p o in t to a figure o f m uch greater complexity. Frequently the depictions show Ebisu w ith a deform ed leg o r even a stum p o f a leg. H e is often obese, and so appears alm ost am orphous. W hile iconographic standards vary widely for E bisu, a com m on elem ent in these depictions concerns the unfo rm ed and m arginal, the ugly and th e dangerous. Obesity, drunken stu p o r in his expression, a deform ed leg— all these attributes p o in t to the m argins o f everyday life. A ccording to N am ihira Em iko, this quality o f deform ity is a key feature in u n d erstanding the significance o f the deity Ebisu. Ebisu is often d e ­ picted as lam e, deaf, one-eyed, herm aphroditic, or very ugly. AU o f these

114

CHAPTER 3

This placard from the Nishinomiya shrine shows the standard iconographic depiction of Ebisu .

qualities, she rightly notes, are attributes of the unformed potential space of the margins of society, where categories such as beauty, gender, health, and order are as yet undetermined .70 Furthermore, ugliness and deformity are considered polluting attributes in the Japanese ritual purity system, and so Ebisu becomes the deity presiding over this realm of chaos, pollution, and all the power and potential it contains. This liminal quality of Ebisu is evident in the many narratives describing the origins of Ebisu shrines, our next source of information for determining the emerging and conflicting nature of this deity.

The scholar Kato Genchi has suggested that there is even a sexual d i­ mension to Ebisu’s deformed leg. H e argues that the stump o f a leg sug­ gests a huge stylized penis protruding from the lower extremities o f the image, and th at this subtle suggestion invokes the powerful world o f sexu­ ality.71 M ore likely, the bad leg is meant to tie Ebisu to the Leech Child myth, to be discussed shortly. While there is no doubt Ebisu is often asso­ ciated with excess avaricious desires, sometimes a bad leg may be just a bad leg. Many o f the images o f Ebisu show him with a chin beard and dark skin. These qualities have been explained to me as p ro o f that Ebisu works at sea, since fishermen working o u t in the sun and open air often have very dark skin. But it seems m ore likely that these depictions o f Ebisu reinforce his foreignness. H e appears as a dark outsider, absorbing some o f the at­ tributes given to threatening M ongols and other menacing continental types in standard Japanese iconography. O n the surface, iconographic standards for Ebisu may appear to be straightforward. In reality the simple, smiling fat man with a fish emerges o u t o f a larger process o f iconographic transformation. W hat does this beneficent fayade obscure?

E n gi N arratives an d Ebisu Worship

A nother im portant source we have for discerning the development and nature o f a deity in Japanese religious history is a body o f narratives de­ scribing the origins o f sacred sites. Called engi, which means literally “or­ igin accounts,” these narratives usually describe some miraculous event which explains why a shrine or temple exists where it does, how it came to have the affiliations it does, and why certain kinds o f worship practices take place at the site. For every shrine or temple in Japan that still has a rec­ ord o f its origin narrative, there are num erous sites for which such records have been lost. It is not unlikely that they were often redacted and u p ­ dated throughout Japanese history and even developed anew at key points, providing definition for drifting traditions. (The Usa Hachiman materials cited earlier are a good example.) Frequently engi are recorded in written texts, but n o t infrequently, these texts have been lost or destroyed, and only oral versions or frag­ m ents survive. Often, the texts or narratives are extremely old, and so are reliable docum ents for reconstructing early perceptions o f deities and worship patterns. Sometimes, as we shall see in subsequent discussions of the Awaji myth tradition, the narratives appear to be the result o f a process o f carefully constructing a center and origin for a tradition. H ere, we will look at a num ber o f these narratives for Ebisu shrines th roughout western

Japan, including the N ishinom iya center. AU these narratives insist on the im portance o f ritually controlling the deity’s potentially dangerous pow er and locating Ebisu in a set worship space. A lthough w orship shrines dedicated to Ebisu th ro u g h o u t w estern Japan differ greatly from one another, m any share the com m on m o tif o f “th e drifting deity” : th e deity w orshipped at a shrine arrived a t its location by literally “floating in ” from the sea. A ccording to O gura M anabu, a study o f m any o f the tutelary deities in villages along the N o to Peninsula reveals th a t m any o f the deities are said to have floated ashore from the o th er w orld across the sea. In his fieldwork in the N o to Peninsula betw een 1947 and 1957, O gura collected sixty-eight occurrences o f the “drifted deity” trad itio n .72 In each case, the shrine’s goshintai, o r receptable o f the sacred, was a stone or some o th er object th a t had som ehow come ashore. In m any cases it was said to have arrived unexpectedly. O gura determ ined th a t there are generally tw o ways th e deities are un d ersto o d to have reached the shore: they are carried by the waves, or fisherm en find them in th eir nets (often in the form o f unusually shaped stones) and bring them ashore. In a n u m b er o f the examples he studied, som e o f the deities arrive in boats, and som e are carried ashore by an animal o r fish (an o c to ­ pus, shark, deer, tortoise). Ebisu figures prom inently as a deity w ho arrives in the hum an realm in this way. For example, th e Ebisu shrine in Walcamatsu-shi in northeastern Kyushu has a large stone as its receptacle for the sacred. This stone, the origin m yth claims, was tangled in the net o f a fisherm an and was later understood to be a m anifestation o f the deity E bisu. A shrine was built for it, and Ebisu is w orshipped there, w ith the stone as the mitamtzshiro (spirit vessel) or goshintai o f th e shrine.73 A nother example, recorded by the folklorist Sakurada Katsunori, paral­ lels this case: T he E bisu-gam i at a small village nam ed S eze-no-ura, o n Koshiki Island, Kagoshim a-ken, w hich I visited som e years ago, is a mere stone. This rock is roughly pear-shaped and was said to have floated ashore. T he rock is so enor­ m ous that n ot even tw o m en cou ld lift it and is definitely n o t a pum ice stone, which is light. A certain villager w h o saw this rock drifting toward the beach made a shrine for it on top o f a hill behind the village, d eem in g it an irreverent act to leave the stone in the water. Im m ediately, how ever, the sto n e began to roll back d ow n the hill and halted at the sp ot w here it is n ow enshrined. . . . It is further told that the stone had been very ligh t before it was installed there, but that it has since b ecom e very heavy.74

In Shimane Prefecture, stones that are found in fish o r fishing nets are called “Ebtsu-san” and are w orshipped o n kam idana in private hom es by fishermen. In Niigata Prefecture, on the island ofS ad o , the goshintai o f a

local shrine is a stone shaped like a piece o f B uddhist sculpture, w hich appeared o u t o f the ocean. I t is called “ E bisu .” We could cite m ore exam ­ ples o f worship sites along the n o rth e rn coast o f Japan in which the spirit vessel is a stone said to have appeared or surfaced.75 In th e O sum i Peninsula in Kyushu, the follow ing practice is recorded: O n the fifteenth day o f the new year, young people from the com m unity en ter the sea, and bring o u t stones they random ly pick up from the b o tto m . These stones are w orshipped as the goshintai for Ebisu in the shrine for the com ing year. A similar practice on the nearby island o f Koshiki-jim a alters th e practice slightly: if th e fishing th a t year is bad, a new stone m ust be found and the previous stone is th ro w n back into the sea. A fu rth er case from W akam atsu in n o rth e rn Kyushu tells o f a fisherm an finding a glistening stone in the w ater, w hich was th en w or­ shipped as E bisu.76 O r fu rth er still, consider the follow ing practice: A fisherm an comes across a corpse o f an unidentified person floating in the sea. T he corpse is pulled aboard th e ship and taken back to shore, w here it is w orshipped as a m anifestation o f the deity Ebisu before being prepared for a p ro p er fu­ neral.77 A ccording to N am ihira, this practice has a n u m b er o f restrictions, nam ely th a t a drow ned corpse from o n e ’s ow n group (i.e., a know n per­ son) is n o t considered to be Ebisu. R ather, Ebisu is the stranger, the o u t­ sider w ho com es ashore. T he practice o f picking up a corpse at sea seems to ru n co u n ter to a strongly enforcecj..taboo am ong fisherm en never to be polluted by death w hen at sea, because the sea is a sacred space. B ut a corpse o f unknow n origin floating in the sea com es from the “ o th er w orld” ( tokoyo) and can thus be deified. R o b ert Sm ith m entions this cus­ tom in his w ork on Japanese ancestors: W hen corpses are foun d floating at sea, they are b rou ght back to village for internm ent in a graveyard reserved for d row ning victim s. Called floating buddhas {n agare botoke), their spirits are th o u g h t to be the m essengers or ser­ vants o f som e major g o d . This represents a contem porary version o f the ancient con cep t o f ven gefu l god s [goryo-shin, ta ta r i-g a m i), that is, spirits o f disaster victim s that are th o u g h t to be dangerous if uncared for b ut b eneficent if deified and properly treated. In deed , these spirits are often appealed to in the expecta­ tion that they can increase the size o f the catch. W hen the corpse is hauled ou t o f th e water, th e fisherm en speak to it, asking for assistance in return for having taken it up for burial.78

F u rth e r examples from Ebisu shrines reveal th at the goshintai are fre­ quently stones w ith odd shapes, often shaped like hum an beings, said to have appeared o u t o f now here, floated o n to the beach, or been caught in a net. This m o tif o f the drifted deity appears to be central to E bisu’s m an­ ifestations to the hum an com m unity. W hile this aspect o f Ebisu worship

shares som ething w ith the stru ctu re o f m arebito w orship described in the previous chapter, there is one striking difference: unlike the m arebito, w ho will probably make its appearance at a set tim e o f the year (usually N ew Year), a floating deity may appear o u t o f now here at any tim e. T he hum an com m unity has a significant lack o f control over th e visit o f the deity, and its appearance (and disappearance) cannot be second-guessed. It therefore falls to ritual to contain these events. T he narrative device o f the drifted deity appears in the origin m yth o f the N ishinom iya shrine. It is m ore com plex than the above examples and shows a greater degree o f ritual developm ent surrounding the appearance o f the deity on the waves. This com plexity is probably due to Nishinom iya being the center o f Ebisu worship and having absorbed o th er Ebisu narra­ tives. Because this shrine has the status and prestige o f the center, its ow n narrative has been carefully sculpted to be as inclusive and definitive as possible. A nd o f course we cannot overlook th e possibility th a t perhaps evidence from o th er shrines, which we assume to be intact, may in fact be fragm ented, appearing less coherent than the N ishinom iya version be­ cause they have n o t been as well preserved.

D rifted D eity M o tifa n d the Nishinomiya Shrine As the center o f Ebisu w orship, N ishinom iya shrine’s engi has a norm ative function in th e Ebisu tradition; it can be seen as a paradigm o f Ebisu worship and practice. T he founding narrative o f this shrine is a fascinating example o f the drifted deity m otif. It is easy to see how this legend o f the center absorbs a n um ber o f related examples o f the drifted deity m otif. H ere is Yoshii’s synopsis: One day, a fisherman was out fishing about three kilometers to the east o f present-day Nishinomiya when he caught som ething in his net that appeared to be an effigy o f a deity \shinzd]. H e thought nothing o f it and threw it back into the ocean and continued fishing. T hen, a bit further along the sea line, in the area called Wadahama79 near Kobe, he caught som ething else in his net. He thought it weird, and looked at it, and again it was this same effigy hanging in his net. Thinking this must mean som ething, he took it home and worshipped i t . That night in a dream, he received an oracle from the deity. “I am the deity Ebisu. I have been traveling the country and have com e to this place, but I would like to have a worship hall built for me a little to the west o f here. I will show you the place.” The fisherman was overcome with fear and spoke o f his dream to other people in a village. Thereupon, they put the figure in a cart for transporting vessels containing sacred forces [o-mikoshi] and carried it to the

w est until they arrived at the beach in front o f w here the fisherm an had previ­ ously found th e figure. T here, they b u ilt a w orship hall. It faces w est [recall th at th e w o rd N ishinom iya m eans “w estern shrine” ] and to this day is the site o f the N ishinom iya Ebisu sh rin e.80

T he N ishinom iya narrative has a n u m b er o f m otifs th at attract o u r a t­ ten tion. First, the deity presents itself to a fisherm an, and he is slow to realize this object in his n et m eans som ething. Are we to assume th at fisherm en catch shrine sculptures in their nets so often it is com m onplace? This fisherm an’s catching the effigy twice enables the deity to make its intentions unm istakably clear: “I f you d o n ’t pick m e up in your n e t and keep m e, I will keep appearing until you do so .” N ext, the fisherm an d e ­ cides to take th e object hom e, and the oracle from the deity is delivered in a dream . T he oracle reveals w hat the deity wants: a place to be w or­ shipped. These m otifs— fisherm an, repeated appearance o f an object until it is recognized, oracle delivered in a dream , and co n struction o f a w orship hall— form the m ajor core o f th e Ebisu narratives. W hat does this case have in com m on w ith all th e o th er examples cited above? AU o f these appearances are in som e way connected w ith the sea. T he object w orshipped as Ebisu has appeared o u t o f now here and largely by chance. T he person w ho finds the object has no control over its appearance. I t either floats in, appears in the w ater, or continues to get caught in a net. In m ost o f the cases, there is som ething unusual ab o u t its appearance. W hen it is a stone, it is a stone th a t floats, a paradox if ever there were one. W hen it is a corpse, th a t to o is strange w hen found float­ ing in the sea. These tw o features o f the narratives, the strangeness o f the object (be it stone, effigy, or corpse) and the unexpected and uncontrolled natu re o f the arrival seem to be the essential aspects o f these divine m ani­ festations. Also p ro m in en t in m any o f these narratives is the role o f a dream or an oracle in which the identity o f the object and the intentions o f the deity are revealed. This device for in tro d u cin g a new awareness o f ano ther ord er o f reality is com m on in Japanese religions, b o th in B uddhist tales and in the larger realm o f legends and m yths n o t directly tied to the B uddhist tra d itio n .81 Perhaps the m ost o u tstanding feature o f Ebisu is its unexpected and spontaneous appearance. In h ere n t in the worship o f these unusual objects th at drift in from now here is the need to ritually control the unexpected and unknow n. E nshrinem ent is n o t m erely the first step in adoration, it is also the first step in containm ent. These unusual stones reveal the deity’s radically different n atu re, and their enshrinem ent makes sure this quality is carefully localized so it can be controlled th ro u g h w orship and ritual action. This is the message o f the floating deity (hydchakugami) phenom e-

non: other beings can appear from across the waves at any time, on their own accord, and it is up to the human community to see that they are properly handled. H ow can this m otif o f the drifting deity help us under­ stand other aspects o f Ebisu worship?

Linguistic an d Orthographic Evidence fo r the Development o f Ebisu A third, though highly problematic, source o f information for studying Japanese deities is the characters used to write their names and the de­ scriptive phrases that have been attached to them throughout Japanese history. Such an approach is problematic because characters were often attached to existing Japanese words with little concern for the original meaning o f the character in Chinese. Characters used in this way are called ateji. W hen looking for meaning in the name through reference to the characters used, therefore, one runs the risk o f being gravely misled. Char­ acters attached to the name Ebisu, however, do not appear to be part o f the ateji phenom enon. “Ebisu” was first a linguistically meaningful word with the meaning “foreigner” (or “barbarian” ), and characters with that meaning in the original Chinese were later added to the word Ebisu. There are a number o f characters used to write the name Ebisu, all of which continue to be in use today, and these suggest early aspects and perceptions o f this deity. The most com mon way to write the name is with a single character that means literally “barbarian,” “foreigner,” or “alien race.” “Ebisu” in ancient Japanese texts refers to the tribal peoples sub­ dued by the centralized government during the protohistorical period of state form ation.82 A second single character also often used for Ebisu also means “ barbarian” but carries the added connotation o f warrior. (In their Chinese readings, these two characters together, read jiii, mean simply “barbarian.” ) In the Heian dictionary Iroha Jiruishd, the word p ro ­ nounced “ebisu,” meaning “barbarian” or “foreigner” uses three other characters.83 A single character is still often used to write the name o f this deity, although the three-character com pound, all ateji characters used for sound only, is more common. Inherent in all o f these terms is the idea o f a stranger. Namihira goes so far as to suggest that the derivation o f the name Ebisu is from the word “Emishi,” a name given to foreigners in ancient Japan.84 Because o f this linguistic evidence, some scholars have suggested that the deity Ebisu is a rem nant o f a deity worshipped by a group o f people who preceded the ethnic Japanese population that later populated and dominated the Japanese archipelago. This theory may in fact have some

tru th to it, b u t it is utterly im possible to prove. A t any rate, discerning the historical origins o f such a com plex phen o m en o n as Ebisu worship and its transform ation does n o t go far tow ard illum inating its significance. O riguchi S hinobu, w ith m uch insight b u t little evidence suggests that the nam e Ebisu originally referred to a generic category o f “visiting dei­ ties” w ho w ould arrive in the hum an realm at certain tim es o f the year. These are m arebito, discussed in m ore detail in the previous chapter. An im p o rtan t aspect o f m arebito is the am biguous and undeterm ined nature o f the deity’s power. T he stranger deity represents unknow n pow er, which can be tu rn ed to benevolent or m alevolent ends depending on how the deity’s arrival and departure are ritually handled. T reated well, Ebisu bestows blessings. Slighted, he wreaks havoc in the form o f storm s, epi­ demics, d ro u g h ts, and o th er disasters. A lthough the double-edged nature o f sacredness is m ost pronounced in the example o f strangers, it is a com m on feature o f all Japanese deities, w ho have four aspects to their sacredness. Each deity (and to a lesser ex­ te n t each hum an being) has at once an a ra -m i-ta m a (rough and vengeful spirit), a n ig i-m i-ta m a (spirit o f union and harm ony), a kushi-m i-tam a (spirit o f m ysterious transform ation), and a saki-m i-tam a (spirit o f bless­ in g ).85 W hether or n o t the deity bestows its benevolent nature on people is largely dep en d en t on the pow er o f the ritual system to carefully m ediate and channel the potency o f the deity’s presence. Japanese ritual prayers indicate th a t people were well aware th a t one and the same deity could have different aspects, and so w hen a deity is addressed, it is invited to present a particular aspect suitable to the request. F u rth erm o re, classical narratives refer to various aspects o f deities, w hich are enshrined separately w ithin the same shrine precincts.86 This discussion o f aspects o f deities helps us situate fu rth er evidence concerning Ebisu. Takeuchi notes th at before the medieval period, the com m on nam e for this deity was “A ra-ebisu” {ara being the same w ord as in ara-m itumtt,, m eaning “ro u g h ” ) and was probably a deity o f curses, ta-tari-gami.87 This fear o f a negative action against the hum an co m m u ­ nity by a deity has been proposed to explain why Ebisu has been m ade into a god o f luck. C asting Ebisu in this light n o t only turns th at w hich is negative in to som ething positive, b u t also increases the ritual arena in w hich this deity’s pow er can be controlled and channeled. It is im p o rtan t to reiterate th a t Ebisu as a deity in Japanese religion is never m entioned by this nam e in the standard chronicles and m ythic ac­ counts o f the N ihongi and K ojiki. Perhaps Ebisu worship by this nam e was n o t a strong en o u g h tradition at the tim e these narratives were redacted to m erit inclusion in the standard chronicles. Perhaps Ebisu w orship was in ­ tentionally excluded. W hat is certain is th a t Ebisu is eventually tied into

the main accounts through identification with deities who do figure in them, deities that have something in com mon with the liminal nature o f Ebisu. These graftings no t only underscore certain aspects o f Ebisu in the deity’s development in Japanese religion but also serve to expand the im­ plications o f Ebisu worship. The graftings also serve to create mainstream traditions for Ebisu worship, and the use o f these Kojiki and Nihongi nar­ ratives suggests a certain type of mythological nostalgic discourse. Here, we turn our attention to the deities onto which the standard chronicles have grafted the worship o f Ebisu.

Ebisu as the A d u lt H iruko (Leech Child) The reader will recall that in chapter 2, the myth o f the failed creation attem pt o f Izanagi and Izanami was presented. The first o f the two deities with whom Ebisu is usually identified is the offspring o f Izanagi and Izanami, H iruko, the Leech Child—a legless, amorphous blob that emerged from a failed creation attempt. This is the association most commonly used in tying Ebisu into the larger Izanagi and Izanami myth cycle. It certainly dates from at least the fifteenth century. Scholars have proposed a num ber of interpretations o f this myth o f failed creation, presented in Kojiki and Nihongi. Tsuda Stikichi suggests that it reflects an early magi­ cal custom in which a body substitute o f a first-born child is placed in a boat and floated downstream to magically protect the infant.88 While such practices o f body substitutes protecting infants are n o t uncommon in Japan, this interpretation seems unlikely because it fails to take into account the m yth’s insistence that this child was a failure. It also ig­ nores the gender hierarchy that must be followed in male-female sexual relationships. Another interpretation, somewhat far-fetched, maintains th at although the word hiruko is written in the texts with the characters “leech” and “child,” these are ateji, characters that should n o t be taken literally. Rather, the w ord hiruko should be understood as the male version of hirume, “sun maiden,” and means “sun lad.”89 This interpretation has little to recommend it, save that it satisfies the desire to tie all myths in this redacted text into sun worship, and render unified that which is most likely more a case o f mythological patchwork through redaction. Both o f these lines of interpretation, one using the myth as a veiled ex­ pression o f history and the other assuming that the key to the myth’s mean­ ing lies in its phonetics, ignore the fundamental reality of how myths oper­ ate, survive, and are transformed in culture. Myths are created and survive precisely because they encode in a narrative structure existential meanings that cannot be expressed in other forms with the same results. Myths are

dependent on a core story line, which can remain constant despite variation and that allows a range o f possible interpretations o f meaning. The most credible interpretation o f this myth addresses it on its own plane o f reference as a story about creation. The birth o f the Leech Child refers to the initial failure o f creation.90 The basic m otif is that before the proper creation o f the world can be accomplished, some false starts and failures are sure to happen. These failures are then abandoned (in boats at sea, in the woods, along rivers, in the wilderness or open fields, under bridges—in short, in any undeterm ined and uncharted space) so that the work o f creation can go on. Lest we take the difficult work o f creation for granted, we need to be rem inded o f the failures o f first tries. Banishing the failure into the wilderness expresses the need to exclude things that did n ot turn out right from o f the final order o f a com pleted world. It also provides a narrative device for expounding on the structures and logic of the successful creation, the basic framework o f a cosmogonic system. Set adrift on the waves, H iruko seems to have been successfully ban­ ished from the realm o f creation. While nothing comes o f the legless child again in the main narrative o f the Kojiki and Nibongi, this deity is brought back into the mainstream o f Japanese religious practice by means o f an­ other myth and ritual tradition: H iruko grows up (or in some cases, merely comes ashore) and becomes Ebisu. The founding myths o f various Ebisu shrines around Japan adapt this m otif o f the abandoned and float­ ing child and claim it as the origin o f the shrine. There is a direct connec­ tion between the account o f H iruko floating on the waves and the m otif o f the floating deity discussed above. Indeed, the Leech Child is the float­ ing deity par excellence. This identification o f Ebisu as the adult Leech Child has become a stan­ dard aspect o f Ebisu worship, and it plays a major part in the worship of Ebisu at the Nishinomiya shrine. A text called Jingi ShojU, said to have been composed by Urabe Kanena from Choroku 4 (1460), mentions the Ebisu shrine in Nishinomiya as follows: “H irota Daimyojin. Chokushi o f the fifth rank. O ne worship at Hama-minami o f M uko-gori, province of Settsu, called Nishinomiya Ebisudono. Izanagi no M ikoto ’s third son, H iruko.”91 Four deities are worshipped at this site: Amaterasu O-mikoto, Okuni-nushi no kami (who as we shall see is the father o f the other deity onto whom Ebisu is grafted), Susanoo no O -m ikoto, and H iruko no M ikoto (the Leech Child, understood to be the child Ebisu). N o t one of these is called by the name Ebisu. Ebisu is worshipped as H iruko, the Leech Child. This pattern can be seen throughout Japan, as worship halls to H iruko are often found within Ebisu shrines. It seems that this deity can exist simultaneously as both child and adult. Worship can be ad­ dressed to both stages, suggesting perhaps that the child’s essence is al­ ways intrinsically present within the adult.

Nakayama Taro presents the following story from Nishinom iya to show this continuity o f ugliness. A ccording to Nakayama, every year on the n inth day o f the N ew Year period, the Leech Child o f the Settsu Nishinom iya shrine pays a visit to th e H iro d a shrine nearby, b u t the appearance o f this deity is so strange and ugly th at people hate to look at it. In contrast to oth er deity processions w here people line the streets to view the passing portable shrine (o-mikoshi) carrying the divine presence, on this day, p e o ­ ple close up their gates and doors and do n o t go o u t, for fear they may actually catch a glimpse o f this ugly deity lum bering alo n g .92 Ebisu, as we know, is obese, herm aphroditic, deform ed, and frequently intoxicated. As a child, to o , Ebisu is so ugly th at people refuse to look upon its passing form . W hat we have here is a worship practice which is in reality antiw or­ ship: people do the opposite o f hailing the o-m ikoshi— they avoid it. Ebisu, w hether child o r adult, is the ultim ate nondcity. O ne w ould w o n ­ der why there should be any b o th er w ith this ugly deity at all, since its presence evokes only aversion. But th e fact that the practice exists points to the need to make this religious point: chaos, disorder, and liminality cannot be denied. These attributes m ust be allowed to exist; they m ust be ritually handled and periodically invoked, even if we have to avert our gaze in the process. A ccording to N am ihira Em iko, there is yet another reason th a t people avert their gaze from the Leech Child. I t is said th at the Leech Child him self also hates to be seen and th a t anyone w ho looks upon this deity will be cursed ( taturureru ).93 T he m o tif o f a deity dem anding th a t hum an beings n o t look upon it is com m on in w orld religions, w ith different rea­ sons given for the prohibition in different traditions. This example o f the aversion caused by Ebisu suggests th a t the very pow er o f Ebisu’s sacredness is the ability to provide a focus for abhorrence and revulsion. H o w do you represent the unthinkable aspects o f c o n ­ sciousness? Ebisu, w hether child or adult, presides over and makes present the realm o f the horrid. Ebisu has been grafted o n to the m ain narratives o f the K ojiki and N ihongi th ro u g h the opportunistic segue o f an ugly and abandoned failed creation, a m iscarried and bloody blob best sym bolized by a Leech.

Ebisu us Kotoshiro-nushi no kam i Some traditions o f Ebisu worship attach the deity n o t to the Leech Child but to another fleeting character in the K ojiki narratives, K otoshironushi no kami. A ccording to Yoshii, the identification o f Ebisu w ith this deity appears to be relatively late, probably dating from the early Toku-

gawa p e rio d .94 This deity K otoshiro-nushi n o kami, w hose nam e means “ the m aster o f know ing th in g s” (referring to divine oracles) is the son o f O kuni nushi no m ikoto and his second wife Kamu ya tate him e no m ik o to .95 (T he reader will recall th a t O kuni nushi no kami is one o f the four deities w orshipped at the N ishinom iya shrine.) T he significance o f this deity K otoshiro-nushi is considerable, in spite o f the fact th at he is m entioned in only a few lines in the Kojiki. H ere is how the deity figures in these myths. O ne o f the central concerns in the first book o f the K ojiki is the p ro b ­ lem o f hegem ony. T he Sun G oddess, having been granted the rig h t to rule over the Plain o f H ig h H eaven, m ust also find a way to rule the land called Toyo Ashi H ara no Chi Aki no N aga Ih o Aki n o M izu ho no Kuni, “T he Land o f the Plentiful Reed Plains, o f the T housand A utum ns, and L ong Five H u n d re d A utum ns Fresh Rice Ears.” 96 She dispatches various people in her entourage to subdue the land and g rant perm ission to take it from O kuni nushi no kami. She sends Ame no ho hi no kami and then Ame no waka hiko, b u t neither is successful in their m ission.97 Finally, she sends tw o deities, Take m ika tsuchi and Ame n o tori-fune no kami. (N ote how the deity w ith w hom Ebisu is later identified conceals him self from the w orld.) We have been dispatched by th e com m and o f A m aterasu and Taka ki no kami to inquire: th e Central Land o f th e Reed Plains, over which you hold sway, is a land en tru sted to th e rule o f my offspring; w hat is your in ten tio n w ith regard to this?” T hen he [O kuni nushi no Kami] replied, “ I cannot say. My son Ya-pe K otoshiro nushi no kami98 will say. H ow ever, he has gone o u t to am use him self h u n tin g for birds and fishing at the Cape o fM ih o , and has n o t yet re tu rn e d .” H e re u p o n , Am e n o T ori-fune no Kami was dispatched to sum m on Ya-pe K otoshiro-nushi n o kam i, w ho, w hen inquiry was m ade o f him , spoke to his father th e great deity, saying, “W ith fearful reverence, let us present this land to the offspring o f th e heavenly deities.” H e th en stam ped his feet and o v erturn ed th e boat; and by clapping his hands w ith a heavenly reverse clapping, he transform ed it in to a green tw ig fence, and concealed him self.”99

T he tale continues as attendants o f the heavenly m aiden inquire o f the old m an if he has o th er children w ho need to be consulted before he can hand over his land. H e replies th at indeed he has one o ther son, Take-m i na kata no kami. This son, w hen asked, also displays magical powers as he engages in a battle w ith the forces o f the Sun G oddess, b u t in the end, they h u n t him dow n and are ab o u t to kill him . Pleading for his life, he agrees to give up control o f the land: “ Pray d o n o t kill m e. I will go to no o th er place. Also, I will n o t disobey the com m ands o f my father, O kuni

nushi no kami, and will not disobey the words o f Ya-pe Kotoshiro nushi no kami. I will yield the Central Land o f the Reed plains in accordance with the commands o f the offspring o f the heavenly deities.” 100 Following this, Okuni nushi no kami agrees to surrender his land, provided a worship hall is built for him. Further, he agrees to “conceal him self’ and wait upon the new rulers.101 It is a m atter o f dispute just where Okuni nushi and his children are said to have concealed themselves. M otoori suggests he retired to the land o f Yomi, the land o f the dead, where Izanagi w ent to fetch Izanami after her death.102 Philippi notes that other prem odern com mentators conclude that they withdrew to Izum o,103 and he suggests that “Probably Opo Kuni nusi merely retires to the unseen world o f the spirit.” 104 Supporting this, book 9 o f the Nihonjji (the account o f Jingu K ogo) refers to this deity as “the deity who rules in Heaven, who rules in the Void, the gem-casketentering prince, the awful K otoshiro-nushi.” 105 Given the evidence in the myths, if we read them as myths, Philippi’s summation seems most plausi­ ble. The myth describes how one order o f meaning supplants another, which withdraws into invisibility. Why has Ebisu been identified with the figure Kotoshiro-nushi no kami in this myth? O n an obvious level, we see in this brief passage a deity associated with hunting and fishing, and this identification is also m en­ tioned in passing in the Nihongi. Kotoshiro-nushi no kami is also clearly capable o f the mastery o f ritual language. But more im portant is his final fate in the text: he disappears from creation and becomes invisible. T hrough his foot stamping, ritual clapping, and utterances, he brings about a magical transformation and withdraws from creation, supposedly so as not to get in the way o f the new rule. It is this feature o f the myth that most attracts our attention: here again, much as with Leech Child, a deity has been removed from creation, in this case through his own submission to the new rule (though one is always suspicious about these matters). Kotoshiro nushi has gone from being a visible deity to an invisible one,106 and like his father, he has agreed to serve as “the rear and the vanguard” to the new ruling forces. The text states that if he “serves them respect­ fully, there will be no rebellious deities.” 107 The standard interpretation o f this myth has been again decidedly euhemeristic: A ccording to this reply, O p o Kuni nushi agreed to relinquish the political, ex ­ ternal rule o f the land to the em perors, w hile retaining a religious, cerem onial role. This is similar to the events o f the Taika reform o f 6 4 6 , w hen the local rulers ( kuni no miyatiiko), w ho had formerly possessed hereditary powers in political and religious m atters, were n ow deprived o f political pow er and replaced by governors dispatched from the central governm ent, although

allow ed to retain their religious functions. N o d ou b t the story o f the abdication o f O p o-k un i-nu sh i is a m ythological reflection o f som e such process o f h istori­ cal d evelop m en t.108

W hile the similarity betw een this m yth and the historical events leading up to the establishm ent o f the early Japanese state are striking and highly recom m end such an in terp retatio n , accounting for a m y th ’s origins in his­ to ry does n o t exhaust the m eanings it assumes as a religious idea. O ur atten tio n to this identification o f the deity Ebisu w ith K otoshiro-nushi can therefore n o t be satisfied w ith a euhem eristic interpretation. Rather, we m ust ask w hat is it ab o u t this figure from the m ythical chronicles that invites identification w ith the deity Ebisu, in light o f w hat we have already seen o f Ebisu worship th ro u g h o u t Japanese history. W hy does a deity o f am biguity and lim inality becom e the deity from the invisible realm , the deity “w ho knows things?” To answer this question, we have to read this m yth as a m yth and n o t as a veiled and sym bolized history, alth o u g h it may have w orked as bo th. T he m yth seems to be ab o u t the pow er o f the invisible (spiritual) w orld to insert itselfin to the visible (physical) realm o f the new order. T he m yth suggests the workings o f the unconscious o r hidden aspects o f hum an spirituality. O ne o f its concerns is w ith the fear o f reprisal o f sup­ pressed forces. W hat w ould happen if these forces, w hich have been com ­ m anded to relinquish their control over the affairs o f the land, should try to resurface? Fear o f reprisal may have been a real concern. K otoshiro nushi’s transform ation from son o f the ruler to “th e rear and the van­ g u a rd ” (to guarantee th a t there will n o t be any vengeful deities) suggests th is.109 A nother example from early Japanese religious literature suggests this interpretation. In 9 2 7 C.E., over tw o h u n d red years after the K ojiki m yths were re­ dacted and recorded, a n u m b er o f ritual prayers used officially by the gov­ ern m en t were recorded as p art o f the E n g ish ik ijm O ne prayer is entitled “T ataru Kami wo U tsushi-yaru” (To Drive Away a Vengeful D eity). T he concern in this m yth is w ith keeping vengeful deities from interfering w ith affairs o f state. B ut strikingly, th e handing o f land over to the forces o f A m aterasu by O kuni nushi form s the structu re o f the prayer. T he prayer im plores this deity and his forces w ith these words: M ay the Sovereign D eitie s111 d w ellin g w ithin the heavenly palace N o t rage and ravage, Because as deities they are w ell acquainted W ith the matters begun in the H eavenly H igh Plain, M ay they rectify their hearts in the m anner o f the rectifying deities . . . A nd may they g o from this place and m ove to another place o f lovely m ountains and rivers

W here they can took o u t over the four quarters, A nd may they reign over that as their place. . . . I place these n ob le offerings in abundance upon tables Like a lo n g m ountain range and present them Praying that the Sovereign D eities Will with a pure heart receive them tranquilly As offerings o f ease, As offerings o f abundance, And will n ot seek vengeance and ravage, But will m ove to a place o f w ide and lovely m ountains and rivers, A nd as deities dwell there pacified. W ith this prayer, I fulfill your praises. Thus I hum bly speak.112

As this prayer makes clear, th e m ythical account o f O kuni nushi su rren ­ dering the land, even tw o h u n d red and fifty years after the com pilation o f the K ojiki, was still the definitive narrative to express the fear o f vengeance and reprisal o f suppressed forces. From a religious perspective, these dei­ ties were m ore than previously defeated political forces. They represented the pow er o f the invisible w orld to bring hardship, vengeance, rage, and havoc on the hum an realm. O kuni nushi and his sons are the pow er o f the unseen w orld. W hy has Ebisu been identified w ith deities whose nam es appear in the m ainstream mythical narratives? This identification is a Japanese example o f syncretism , an appropriation o f one religious system by another, in this case, the Shinto tradition absorbing a w ider (and perhaps earlier) m yth tra d itio n .113 W hat these tw o cases also reveal, however, is th at Ebisu is connected to the repressed, the invisible, the deform ed, and the aban­ doned. As the ICojiki narratives gained in pow er as representing standard religiosity, o th er religious systems so ught ways to be identified w ith this pow erful tra d itio n .114 These tw o examples suggest th a t m ainstream religions have a strong need to find a place for the unseen and unknow n w orld. T he fact th a t this abandoned deity finds its way back into m ainstream Japanese w orship sug­ gests th at those parts o f creation th at are half-form ed, am biguous, ugly, and even potentially dangerous have an essential place in the ord er o f things. Banished, they reappear. Repressed, they force their way back into consciousness. R endered invisible, they reclaim the visible realm . A nd how they resurface may be even uglier than before, w hen we first pushed them aside. B uilding a w orship hall for Ebisu is, at a spiritual level, m uch safer and m ore com forting than the im age o f an unform ed and m isshapen child floating endlessly on the waves o f the prim ordial ocean, w aiting to come ashore from tim e to tim e to seek vengeance.

Absorbins an d H a n d lin g Epidemic D eities In Japanese religions, there is a category o f deities referred to as ekibydgami, or epidem ic deity— a term created by scholars o f Japanese re ­ ligions. I f properly w orshipped, these deities can help prevent an epi­ dem ic; if neglected, they can cause one. T h ro u g h o u t Japanese history, there have been a large n u m b er o f deities th a t have been connected w ith epidem ic disease in som e way, and the relationship betw een certain dis­ eases and deities becam e very specific.115 For the m ost part, ekibydgami in Japan are related to the m ost co m ­ m on— and m ost incom prehensible— epidem ic disease th a t Japan has faced in its history: smallpox. In fact, deities associated w ith smallpox com prise their ow n class o f epidem ic deities, called hosogami (smallpox deities).116 A ccording to the historian Nakajim a Yoichiro, smallpox reached epidem ic p ro p o rtio n s in Japan approxim ately one h u n d red times from the first m ajor record o f a w idespread epidem ic in 7 3 5 -7 3 7 until the end o f the n in eteen th century. T he physician and historian o f m edicine Fujikawa Yu is m ore conservative and lists only fifty-four occurrences o f sm allpox, w ith intervals betw een the outbreaks decreasing by the late tw elfth century, suggesting th a t th e disease had becom e endem ic to the Japanese p o p u latio n .117 W illiam M cN eill, in his classic study o f the im pact o f epidem ics on hum an dem ographics, Plagues a n d Peoples, has argued th a t because o f Japan’s relatively isolated geographical location vis-a-vis the m ajor p o p u ­ lation centers o f Asia (primarily C hina), epidem ic disease in Japan was quick to spread and devastate th e population b u t slow to acquire endem ic status and mimic the disease patterns o f the Asian c o n tin e n t.118 This m eant th a t every tim e an epidem ic spread th ro u g h o u t Japan, it w ould have disastrous results, sweeping from one end o f th e country to the other. T he virus o r bacteria w ould th en die o u t completely, only to be followed by a new infection o f the population half a generation or a gener­ ation later, w ith equally high m ortality rates. B oth Japan’s isolation and relatively low population density prevented smallpox from becom ing en ­ dem ic in Japan until ab o u t the th irte e n th century. A fter this period, we see the disease often referred to as a childhood disease. T h ro u g h o u t East Asia, varieties o f pestilence and disease— bubonic plague, smallpox, measles, influenza, cholera, typhus— were often u n d er­ sto o d to be the result o f neglected deities, m alevolent spirit forces, and epidem ic dem ons. C onsequently, m ajor rituals o f appeasem ent were u n ­ dertaken to control the sickness by pacifying and expeling these forces from society. Japan was no exception to this description, and som e o f Japan’s m ost elaborate ritual perform ances, including the G ion M atsuri in

K yoto, began as spirit appeasem ent rites for epidem ic deities. T he connec­ tion betw een Ebisu w orship, puppetry, and the appeasem ent o f epidem ic spirits is a difficult area o f research, largely because so m uch o f the practice o f epidem ic spirit appeasem ent was unw ritten. W hile we do have sources docum enting governm ent-sponsored appeasem ent rites during tim es o f pestilence in Japanese history, only a few sources, often m ade in passing, refer directly to the relationship betw een puppetry and epidem ic spirit appeasem ent. The record suggests th at the Usa H achim an H ojo-e began as th e result o f an epidem ic and expressed a dual use o f puppets, first in the military conquest o f the H ayato and later in the ritual control o f the epi­ demic said to have resulted from their defeated, m alevolent spirits. This case is representative, and even w ith the paucity o f sources is one o f our better recorded cases. T he o th er source o f inform ation for studying this connection comes from the fragm ents o f evidence describing a small w or­ ship hall at the N ishinom iya Ebisu shrine, dedicated to th e tutelary deity o f puppeteers, H yakudayu.

Hyakudayii, Women of Pleasure, Children, Puppets, and Smallpox The worship hall dedicated to Hyakudayii is m entioned in an eleventhcentury text by O e no Masafusa ab o u t the w om en w ho lived and w or­ shipped near the w hat is now know n as the N ishinom iya shrine. His Yujoki (C hronicles o f the W om en o f Pleasure) contains th e following lines about the H iroda shrine (and also gives us a date for situating kugutsu and perform ers at this site): “To the south is Sum iyoshi, and to the W est is H iroda. They [the w om en] m ade these the places w here they pray for agreeable om ens. In particular [they pray to ] Hyakudayu. This is one nam e for dosojin} 19 T he num ber o f these [character missing from text] they carve is in the hundreds and thousands. They can m elt people’s hearts. This also, is an old practice.” 120 O e no M asafosa’s reference to the worship o f H yakudayu is intriguing, because it suggests th a t perhaps H yakudayu was originally som e sort o f a carved object n o t unlike the effigies discussed in chapter I . T here is a n ­ o th er story, this one from the N ishinom iya shrine itself, explaining the origins o f this connection betw een H yakudayu, the Ebisu shrine, and puppeteers. A text o f the Kyoho era ( 1 7 1 6 -1 7 3 5 ) entitled Meisho Nishinomiya A n n a i Sha contains the following story, which we can read as a reflection o f the founding m yth o f the w orship site: North and a little bit separated from the main shrine precincts [of Nishinomiya] there is a small worship hall [hokora\. In the era of the deity Iwakusubuna, in this inlet, there was an old man named Dokun. When he heard that the deity at this shrine at the age of three still could not walk,121 he made a puppet to

entertain and raise the child deity. Even today, when a newborn child is one hundred days old, the parents bring the child to this worship hall, and here give the child a name. Objects worshipped at this site include five painted effigies. In this, we can see a practice which has been handed down from the age o f the gods. Nishinomiya puppeteers, and also the practice o f calling puppets “D okum bo [literally “the stick o f D okun” ] comes from this deity. Today, this place is where one comes to pray for the longevity o f a newborn child.122

T his text ties th e practice o f p u p p etry at the N ishinom iya shrine to the H yakudayu worship hall. It also supplies w hat am ounts to a founding m yth o f puppeteers at N ishinom iya: pu p p etry was invented for the p u r­ poses o f entertaining and appeasing the Leech Child. T he nature o f H yakudayu helps us understand m ore ab o u t ritual pu p p etry at the N ishi­ nom iya center. Tknother N ishinom iya shrine d ocum ent, N ishinom iya D a ijin H onki, contains the follow ing explanation— basically the same evi­ dence— u n d er the heading “H yakudayu W orship H a ll” : “O nce upon a tim e, in this inlet there was an old m an nam ed D okun, and to entertain and appease the deity H iru k o [Leech C hild], he fashioned a small p u p p et and m ade it dance. This is the said origins o f the N ishinom iya puppeteers. M oreover, people in this area carry their new born babies to this shrine on the baby’s h u n d red th day, and p aint their baby’s face w ith a pow der from here. This is said to be g o o d for the baby. T herefore, the nam e H yaku­ dayu com es from this practice.” 123 W hy com e to H yakudayu to ask for th e p ro tectio n o f a baby? A brief vignette from a text entitled Setsuyo Ochiboshu by H am am atsu U takuni (1 7 7 6 -1 8 2 7 ), dated Bunka 5 (1 8 0 8 ), adds a fu rth er dim ension to the relationship betw een H yakudayu, puppetry, Ebisu, and the p ro te c tio n o f children. U n d e r the heading “A bout the N ishinom iya H yakudayu” there is the following: In Nishinomiya, to the north o f the Ebisu shrine, there is a small shrine in which a statue o f a child about three years o f age is enshrined. The child is not a deity. Every year at N ew Year, there is the custom o f taking about three or four cups o f powder and painting the face o f the statue with it, and then leaving it on. People in the area w ho have had children born that year rub the face o f the statue and then rub the powder on their own children’s faces. This is for the prevention o f smallpox. It is also said here that this doll is the origin o f Japanese puppetry. This doll is from the Kasai family o f Nishinomiya, and the Naniwa puppet troupe came from that group. These people are called tayu, and this is why the statue is called Hyakudayu.

In Yamazaki Yoshinari’s Tanki M anroku, volum e 2, we find a section entitled “A Reduced-Size C opy o f the Smallpox P rotecting D eity in the G enuine H andw riting o f the Venerable K dtaku.” 124 This piece o f w riting is alm ost identical to the same a u th o r’s piece in his tw enty volum e w ork

K airoku, written between 1820 and 1838, under the heading “Hyakudayfl: Protective Deity o f Smallpox.” Here is what his heading in Tanki M anroku says: Hyakudayuden in the Genuine Script by Kotaku-Sensei: The origins of the Protective Talisman (O-mamori) during the Kyoho Era at the time of a smallpox epidemic A ccording to an oracle, before a person manifests smallpox, one should p u t up a protective talisman o f (this) deity in various places. T he m ediator o f the above-m entioned oracle n o ted th at at the Settsu N ishinom iya shrine there is a hall called the H yakudayuden, and this deity enshrined there is a protective deity for smallpox. T he oracle says th a t in the event o f smallpox, one should quickly call th e priest o f th e H yakudayil den to present the protective talis­ mans. T he priests say th a t th e talism an is n o t effective unless the greatest callig­ rapher u n d er heaven is sum m oned and, w ith the instructions o f the priest, asked to w rite the four characters for H yakudayuden upon it and hang it near oneself. I f o n e has a devoted heart, th ere is n o need to w orry a b o u t the smallpox epidem ic. T he calligrapher H osoi K otaku was called, and according to the aforem en­ tio n ed p riest’s instructions, he w rote tw o sheets w ith the characters “ Hyakudayuden” on them . T he priest presented one o f them to the H yakudayu hall and gave the o th e r one to K otaku himself. Kotaku stores the piece in his hom e. A fter this event, Kotaku d id n ’t contract sm allpox, and was very pleased, but his son B unzaburS g o t it anyway. N evertheless, shortly hereafter, w ithout m uch ado, the boy was healed. T h en KStaku said th at he should give the sheet w ith the characters o n it to one o f his disciples, M orooka N anrin, w ho has many children. H e th en did so. T he HyakudayQ origins are w ritten ab o u t in so m uch detail th at I w o n ’t b o th er to d o so h e re .125 A b o u t th e aforem entioned K otaku Sensei, I heard this story from his disciple N anrin-Sensei. This story was typeset on a board and w ritten dow n in A n’ei 7 [1 7 7 8 ], in th e fo u rth m onth.

This text refers to a practice that was com m on during the Tokugawa period o f Japan, namely, using amulets for protection against small­ pox, and refers to the efficacy o f the Nishinomiya Hyakudayu to ward o ff smallpox. Kugutsu attached to the Nishinomiya shrine considered Hyakudayu to be their tutelary deity (ujig am i), and continued to do so until quite late. A source dated Bunka 13 (1817) indicates that when a group o f Nishinomiya puppeteers presented a joruri performance (against the wishes o f the Nishinomiya shrine), they dedicated the performance at the Hyakudayu worship hall.126 Sometime around the beginning o f Meiji, this wor­ ship hall dedicated to Hyakudayu was moved into the shrine precincts.

T he HyakudayQ w orship hail still stands today, although the Sanjo district is no longer distinct, and its environs have been reabsorbed in to the larger Nishinom iya shrine. W hat can we make o f these shreds o f evidence— hints o f ritual puppetry, glimpses into the nature o f the enigm atic Hyakudayu, suggestions ab o u t Ebisu w orship— and w hat was their relationship to the Ebisu shrine and the people whose prim ary occupation was the ritual h an ­ dling o f Ebisu? T he interesting feature o f puppetry at this cerem onial center concerns the m yth in which a hum an officiant, D 5kun, uses a pup p et to entertain the deity child H iru k o , an event th at is u n d erstood to be the origin o f N ishinom iya puppetry. This use o f p uppetry has been generalized so that visits to this site are believed to prevent smallpox in children.127 In local versions o f these stories, there is a clearly im plicit relationship am ong the following motifs: ( I ) a child deity is unable to walk and is abandoned on the sea (and floats ashore as a drifting deity), (2) a hum an shrine officiant creates a puppet to entertain and appease the child deity, (3) parents can then bring their child to this shrine on its h u n d red th day, (4) pow der is ru b b ed on the child’s face or on the effigies in this w or­ ship hall, (5) smallpox can be prevented by this practice, and (6) even the name o f the Hyakudayu site in understood to be efficacious against the dreaded disease. Fitted together, these fragm ents o f evidence start to form a recogniz­ able picture: puppetry was understood to be a practice capable o f appeas­ ing this dreaded deity H iru k o , later connected (even equated) with Ebisu. T he effigies in the worship hall o f H yakudayu were understood to have been either body substitutes for children, o r replicas o f the puppets used in appeasem ent rites. T ouching these effigies safeguarded one against the dangerous forces o f the Leech Child. It is interesting th a t evidence points to a connection betw een an effigy used to ward o ff disease and the origins o f puppetry. T he practice o f using effigies to w ard o ff disease has a long history in Japan. It was probably borrow ed from China, although may also have developed independently. Felicia Bock writes this com m ent on a line in her translation o f the Enjjishiki (com piled during the Engi era, 9 0 1 -9 2 2 ): “A nother custom b o r­ row ed from China to avert epidem ic disease was th at o f fashioning in clay small images o f cows (or bulls) and children and placing them at main thoroughfare gates around the palace.” 128 Such effigies w ould serve as targets for an epidem ic spirit’s w rath, and real children (or animals) w ould be spared the illness. In all likelihood, ritual puppetry draws on this under­ standing o f the pow er o f the body substitute. AU the evidence relating to epidemics and the Nishinom iya center sug­ gests a highly localized association o f H yakudayu, Ebisu, and ekibyoJfam i— part o f the Ebisu tradition th at was absorbed into the Nishinom iya

shrine com plex and “cordoned o f f ’ into the single figure o f H yakudayu. AU these pieces o f evidence date from th e Tokugaw a period. Regardless o f the actual historical connection betw een these cases, the connection betw een N ishinom iya as a center for puppeteers and Nishinom iya H yakudayu as an ekibyogami was well know n. This com m on per­ ception may well have contributed to a fear o f puppeteers as those w ho traffic w ith epidem ic spirits.

The Gradual Breakup o f the N ishinom iya Kugutsu In spite o f all we know, it rem ains to be seen w hat these Ebisu-kaki actu­ ally did. In chapters 4 and 5,1 present a detailed description o f the origins o f the Awaji tradition and Awaji Ebisu rite, w hich grew o u t o f the earlier N ishinom iya ritual and still retains som e associations w ith this form er in ­ terpretation o f the deity. T he Awaji case is probably the best source o f evidence we have for understanding w hat these N ishinom iya Ebisu-kaki did w hen they presented ritual perform ances. References to the Ebisu-kaki from Nishinom iya are num erous b u t dis­ appointing. T heir num bers indicate th at the practice was w idespread. T he historical record is m arred by the fact that by the end o f the sixteen cen­ tu ry ritual p uppetry perform ances were so com m on en o u g h th a t everyone knew w hat they were and they did n o t require description. By the early seventeenth century, for example, Ebisu-kaki already was a com m on term for puppeteer. H ere is a text dated the n inth m o n th o f Keicho 19 (1616): “It was raining. Som e perform ers came and presented a piece called A m id a no M unew ari. They w ere Ebisu-kaki types.” 129 A lthough A m id a no M unew ari (Am ida’s Riven Breast) is a piece o f early joruri describing th e deeds o f Am ida B uddha to save a young and pious girl, anyone p re ­ senting a play w ith puppets was “ an Ebisu-kaki type.” 130 We know that these puppeteers presented talismans for the Nishinom iya shrine and re ­ ceived som e protection and a small salary for their services. As we shall see shortly, by the eighteenth century this center for puppetry began to shift, and perform ers living in the Sanjo district attached to the Nishinom iya shrine began to perform nonreligious ballads. T he accom panying draw ing o f an Ebisu-kaki is from J in r in K um m o z u i, a seven-scroll collection describing everyday life in the G enroku era, drawn in G enroku 3 (1690). T he seventh scroll, dedicated to artists and w om en o f pleasure, includes this picture, titled “Ebisu-m ai” and bearing the caption, “They im itate N o h and do lots o f th in g s.” 131 Ritual puppeteers attached to the Nishinom iya shrine clearly perform ed duties, ranging from traveling around the country presenting Ebisu rites to w orking w ithin the shrine precincts, and possibly even acting in some

A N ishinom iya kugutsu-mawashi, called here “ E bisu-m ai.”

capacity to help prevent the spread o f smallpox th ro u g h special rites in ­ volving their puppets at the H yakudayu shrine. But by the early m odern period, a new trend was afoot. This tren d , w hich as we shall see led to a gradual decline o f Nishinom iya as a center for puppetry rites, gave birth to a new period in the history o f puppetry. This was p art o f a larger develop­ m ent in Japanese perform ing arts, which later becam e the highly nuanced ningyo joruri trad itio n , a new context for p uppetry as dram atic e n te rta in ­ m ent. This new developm ent created a bit o f consternation on the p a rt o f N ishinom iya shrine authorities and led to an eventual split and the estab­ lishm ent o f the new Awaji tradition. This m ovem ent o f puppeteers o u t o f the shrine appears to have been well underw ay by the early eighteenth century. A record from the twentyseventh day o f the th ird m o n th o f Kyoho 5 (17 2 1 ) indicates that Nishinom iya puppeteers perform ed in a tem ple in Amagasaki; and on the nine­ tee n th day o f the te n th m o n th o f Kyoho 8 (1 7 2 4 ), a person from the N ishinom iya Sanjo, in opposition to a shrine priest there presented a p u p ­ p etry perform ance inside the shrine precincts. And again in Kyoho 9 (1725) and Kyoho 12 (1 7 2 6 ), there were listings o f perform ances by p u p ­ peteers from the N ishinom iya Sanjo district at tem ples in Amagasaki. So, twelve or thirteen years later, in the Kanpo era (1 7 4 1 -1 7 4 4 ), a line in shrine records dated the fifth day o f the eleventh m o n th o f Kanp 5 I (1 7 4 1 ) reads, “In recentyears, Sanjo m ura distractions have been distress-

in g .” This com m ent indicates th at a lo t o f these new “pu p p et perfo rm ­ ances” were seen as distractions from their em ploym ent at the shrine. But it appears th at n o t just one group was absenting itself and conducting its ow n affairs. In Bunka 13, Sanjo m ura’s Yoshi Fusaburo (second son) and six others gave a puppetry perform ance and dedicated it at the H yakudayu shrine.132 These references suggest th at while puppeteers w ere still co n tin ­ uing their practice o f situating their new perform ances in the context o f H yakudayu w orship, they were n o t m erely acting as representatives o f the shrine; they were developing their ow n dram atic tradition, which was b e­ com ing an increasingly popular one. It was in the context o f Nishinom iya puppeteers breaking away from their duties at the shrine th a t we see the rise o f th e Awaji puppetry trad i­ tion as a self-consciously constructed new center for spirit appeasem ent using puppets.

4 A D ead Priest, an A ngry Deity, a Fisherm an, and a Puppet: The Narrative Origins o f Awaji Ningyo w a r n i n g : T h is art is for th e ap p ea sem en t o f k a m i.

P eop le after th is sh o u ld n o t take this lightly. I f th ey d o , it w ill w e ig h h eavily u p o n th e m . H e r e ­ after, p eo p le sh o u ld b e so rely afraid. — from th e fo u n d in g narrative o f th e Awaji trad ition , D dku m bo D e n k i ( 1 6 3 8 )

I n t h e last chapter, w e saw that p u pp eteers called E bisu-kaki lived in the Sanjo district near th e N ish in o m iy a shrine and w ere em p lo y ed there as lo w ranking officials. T h ey p erform ed rituals to spread E bisu w orsh ip and w orsh ip p ed an ep id em ic d eity called H yakudayu. T h ese Ebisu-kaki w ere extrem ely popular by th e m id to late six teen th century, ju d g in g by the nu m ero u s referen ces to th em at th e tim e. By th e m iddle o f th e sixteenth century, th ey began to expan d their range o f perform ances to in clu d e dra­ m atic skits. T h e tradition o f appeasing th e deity E bisu th ro u g h ritual p u p ­ petry was said to have started w h en a priest nam ed D o k u n m ade a p u p p et to entertain and raise th e L eech C hild. In this chapter w e turn to a d iscussion o f th e d e v elo p m e n t o f A w aji’s Sanjo district in Ichi as the cen ter o f ritual p u ppetry in central Japan aroun d th e m id d le o f th e six teen th century, a d e v e lo p m en t c o in cid in g w ith a splintering o f th e N ish in o m iy a k u gu tsu tradition. A m ajor im petus for th e rise o f Awaji as a n ew cen ter w as the increasing a m o u n t o f traveling and p erfo rm in g that p up p eteers w ere d o in g ou tsid e th e N ish in om iya area. T h ey may have w anted to establish a n ew cen ter o f their o w n . A study o f the d ev e lo p m e n t o f ritual p u p petry o n Awaji provides hints as to h o w new perform ance trad itions arose in late m edieval Japan and h o w they so u g h t to leg itim a te th em selves as d istin ct traditions. A central m o tiv a tio n o f puppeteers 0 1 1 Awaji w as estab lish in g puppetry as a ritual practice free from th e jurisd iction o f th e cerem onial authority o f N ish in o m iy a w h ile appropriating the religiou s authority o f a priest from that shrine. T his w as accom p lish ed in large part th ro u g h th e p rod u ction o f a un ifying narrative. B asing th eir n ew tradition o n a K an’ei 15 ( 1 6 3 8 ) d o c u m e n t describing th eir origin s, D dkum bo D en k i1 Awaji puppeteers at-

tem pted to create a unique and legitim ate identity for them selves as free­ lance ritual specialists, traveling the country. T he text em phasized the loosening o f attachm ents to the N ishinom iya religious center. T h e rise o f itinerant ritual puppetry on Awaji participates in a larger dynamic in Japa­ nese religious history: w hen unordained and non-ecclesiastical religious specialists becam e highly popular, m ajor religious centers (b o th B uddhist and am algam ated B uddhist-Shinto centers) so ught ways to control their activities. T here are basically tw o lines o f interpretation th a t account for the exis­ tence o f the Awaji tradition. O ne argues th a t Awaji has been a center for puppetry from the late H eian period and th at this center developed paral­ lel to the N ishinom iya kugutsu groups. This interpretation provides the Awaji tradition w ith a longer history than historical docum ents can vali­ date. T he o th er line o f interpretation, which I shall develop here, argues th at Awaji puppetry developed in the m id-sixteenth century w hen Nishinom iya puppeteers left th a t center. From my reading o f the fragm entary evidence, the latter interpretation is the only verifiable account o f the ori­ gins o f the Awaji tradition. I shall begin, however, w ith a discussion o f the first interpretation, w hich represents a m odern version o f an older dy­ namic in the Awaji tradition— autonom y from the N ishinom iya center and the uniqueness o f Awaji as a place.

T h e C radle o f A w aji P u p p etry: A Sanjo D istr ic t o n th e M ihara P lain T he cerem onial center o f Awaji puppetry lies in the m iddle o f the M ihara plain in the lower half o f the island. T he tow n is called Ichi (w ritten w ith th e character m eaning “ m arket” ) and the district is Sanjo. T he center is a small shrine now called O m ido H achim an D aibosatsu. Just how and w hen this site became a center for p uppetry is a p o in t o f controversy, as historical records describing the early history o f Awaji ritual puppetry are vague. A cluster o f hypotheses, popular am ong folklorists, look for the roots o f Awaji puppetry in the im agined religious universe o f the early m aritim e people o f the Inland Sea.

The Amazoku Hypothesis A ccording to archaeological studies conducted on Awaji, the central plain o f M ihara-gun is am ong the oldest settled sites in the Inland Sea. F or this reason, there have been attem pts to tie the developm ent o f Awaji p u p ­ petry to th e people w ho populated this island in the protohistorical pe-

NARRATIVE ORIGINS OF AWAJI NINGTO

139

Sanjo • (6mido Hachiman Daibosatsu)

Map of Awaji showing the location ofSanjo and two of Awaji's larger towns .

riod, maritime people called by the general term amazoku or ama (sea people). How are we to understand the popularity of this kind ofhypothesis, and what light can it shed on the dynamics of tradition formation in Awaji puppetry? Wakamori Taro, the folklorist who in the 1950s supervised a major investigation of Awaji's history, mythology, archaeology, and folklore, has argued that by the protohistorical, or late Kofun period (roughly the fifth

to the seventh centuries C . E. ) , tribes o f sea people were already organized into a so rt o f confederation in this region, so th a t the people living on the small island o f N ushim a and also at the w estern end o f Shikoku could be b ro u g h t to n o rth e rn Awaji (across from present-day Akashi) to work as navigators o f the Inland Sea and divers for pearls and seaweed. W akamori suggests th at perhaps the southern p art o f Awaji, called Ama, and the plain o f M ihara may have been th e center o f pow er o f the am azoku o f the w estern Inland Sea.1A lthough Awaji is quite m ountainous, this plain c o n ­ sists o f a large stretch o f relatively flat and arable land. M iyam oto Tsuneichi writes th at Awaji, Awa (on Shikoku), and the Settsu region (near present-day N ishinom iya) constituted a region d o m i­ nated by am azoku. H e cites a do cu m en t dated the fifth m o n th o f 844 which says there were m ore than three thousand ama settled in groups along the beaches and bays o f Awaji.2 Perhaps the central plain o f M ihara may have been the center o f the “confederated tribes” o f ama w ho c o n ­ trolled th e w estern Inland Sea. Little is know n ab o u t th e am azoku o th er than w hat can be gleaned ab o u t them from readings o f K ojiki and Nihongi, interpretations o f ar­ chaeological digs along the Inland Sea, references to the m em bers o f their tribes w ho, w hen subjugated, becam e perform ers for the imperial family, and speculation on their contributions to o th er religious traditions in Japan, particularly H achim an w orship.3 F rom these sources, a hazy pic­ ture emerges o f these people: T heir leaders were female sham ans w ho used oracles to organize their governance. T heyw ere involved in the prac­ tice o f diving and fishing and w ere also adept at navigation from the sea using m ountain peaks as points o f reference, a skill th a t proved useful to the governm ent o f the em erging Yamato state as it so ught to conquer w estern Japan.4 M any o f the m yths found in K ojiki and N ihongi clearly have m aritim e them es and are considered to be traditional m yths o f these tribes, redacted to make a coherent narrative for the imperial line o f Yamato rulers. It is widely speculated th a t Ebisu may have originally been a deity o f the am azoku, a hypothesis th at w ould partially explain why Ebisu worship is found all over Japan from an early tim e and why the deity is called only by th e nam e m eaning “foreigner” or “barbarian” and is not m entioned in the redacted myths. O ne can postulate th at perhaps an older tradition o f divination, oracles, and appeasem ent o f m alevolent spirits is the com m on d enom inator be­ tw een Awaji, N ishinom iya, and Usa. T he coincidence o f Sanjo districts where puppetry is practiced in lower Awaji, N ishinom iya, and Usa; the apparent relationship betw een H achim an and Ebisu worship and ama,zoku·, and the fact th a t small worship sites dedicated to Hyakudayu are present in all three sites are striking. T h e reader will recall th at one inter­ pretation o f the origins o f the w ord kugutsu proposed by O riguchi sug-

g ested th a t am azo k u w o u ld actually carry p u p p e ts in th e ir baskets (kapo o r kug u) used fo r g a th e rin g seaw eed. T his hy p o th esis has n o evidence save th e m o st far-fetch ed p h ilological specu latio n s to back it u p , b u t it p o in ts to a s tro n g ten d e n c y in so m e tra d itio n s o f Japanese folklore stu d ies to tie th e m y sterio u s a m az o k u to th e d ev e lo p m e n t o f Japanese pu p p etry . T h e re is n o t e n o u g h evidence to s u p p o rt th e h y p o th esis, yet th e very h y p o th esis itse lf is so m e th in g o f a d a tu m in th e stu d y o f Awaji pu p p etry . Like th e d e b a te over w h e re th e a n c ie n t ru le r P im iko w as b u rie d , th e q u est fo r th e w o rld o f th e am azo k u c o n stitu te s s o m e th in g o f a fad in Japanese fo lk lo re stu d ies. S u b ju g a te d in th e early p e rio d o f th e Japanese sta te , th ese m aritim e peo p les re p re se n t a b u ried an d fo rg o tte n p ast a n d have b eco m e ico n s fo r a lo st p a rt o f th e self in c o n stru c te d id e n titie s in Japan today. T h e a tte m p t to find th e ro o ts o f Awaji n in g y o in th is an c ie n t p ast seem s to be n o stalg ia o f a u n iq u e k ind. T h e sea re p resen ts a fluid, m y sterio u s, d a n g e r­ ou s, a n d p o w erfu l o rd e r o f m ean in g . R eclaim ing it as p a r t o f o n e ’s p ast is a p o w e rfu l act. T h e lu re o f th is hy p o th esis was p ow erful e n o u g h to fuel th e d e v e lo p ­ m e n t o f a sh o rt-liv ed th e m e park o n A w aji, w h e re o n e co u ld “ experience life th e way th a t th e a n c ie n t am azo k u d id ” by living in h u ts, m ak in g salt o u t o f sea w ater, an d d ig g in g irrig a tio n d itch es for basic ag ric u ltu re. A r­ ch aeo lo g ists a n d h isto rian s o n Awaji have b een m o re circ u m sp e ct a b o u t this th e o ry , a n d th o u g h m o st o f th e m find th e hyp o th esis app ealin g , th e re is a tacit a g re e m e n t th a t it is u n p ro v a b le , to be left to en e rg etic y o u n g scholars a n d arch aeo lo g ists w ith a flare for th e sensational.

The Usa Hypothesis A parallel, an d n o t to tally u n re la te d h y pothesis a b o u t th e o rig in s o f Awaji p u p p e try co n cern s th e re la tio n sh ip b etw e en th e sp read o f H a c h im a n w o r­ ship fro m th e U sa c e n te r, th e use o f ritu al p e rfo rm an c e in deity ap pease­ m e n t, a n d Awaji p u p p etry . In th e late 1 9 5 0 s, W akam ori an d H a n d a Yasuo (a folklorist fro m O ita U niversity) p ro p o se d th e hy p o th esis th a t th e Awaji Sanjo site was actually a div in atio n d istrict c o n tro lle d by m iko (fem ale sh a­ m an s) w h o , u n d e r th e ju risd ictio n o f th e U sa H a c h im a n sh rin e, m ain ­ ta in e d a ce n te r o f oracles a n d d iv in atio n o n Awaji. W akam ori an d H a n d a a rg u e d th a t th ese sham ans m ay have u sed p u p p e ts to appease deities, an d th e y sp ecu lated th a t p u p p e te e rs fro m a d iffe re n t k u g u tsu g ro u p m ay have co m e later fro m N ish in o m iy a, jo in e d forces w ith these sham ans, a n d d e ­ v elo p ed p u p p e try fro m a strictly ritu al p erfo rm an c e in to a m o re dram atic fo rm o f e n te rta in m e n t.5 S uch a h y pothesis w o u ld su g g est th a t Awaji p u p p e try aro se as a parallel tra d itio n to th e N ish in o m iy a ce n te r, a n d al­ th o u g h Awaji p u p p e te e rs w ere stro n g ly c o n n e c te d to N ish in o m iy a, they

may have had roots predating the breakup o f the N ishinom iya kugutsu groups. T he intention o f this hypothesis was to form ulate a history for Awaji puppetry th at gave it a unique history n o t derived from the Nishinom iya kugutsu g ro u p s.6 W akamori and H anda never developed their hypothesis beyond a n u m ­ ber o f basic questions, bu t these slowly assum ed the form o f an argum ent in the folklore studies m ovem ent. Like the am azoku hypothesis, and in som e subsequent presentations on the subject, the Usa hypothesis was so often cited as fact th at it assum ed the status o f d ata.7 B oth the am azoku and Usa hypotheses ren d er the history o f Awaji p u p ­ p etry exotic. They imply that this perform ance tradition holds the key to th e hidden past o f Japanese religious life. W hen seen as cultural p ro d u c­ tions, b o th o f these hypotheses are fascinating; as historical reconstruc­ tions, they are less com pelling. B ut if we can co n ten t ourselves with a m ore recent (and perhaps less scintillating) account o f origins, the early developm ent o f Awaji puppetry reveals a great deal about the production o f a tradition. W hat can we actually know ab o u t the developm ent o f ritual p u p p etry on Awaji?

Awaji Puppetry’s Birthplace and Center: O m ido H achim an D aibosatsu T he center o f Awaji ningyd is a small shrine on a back street o f the dis­ trict called Sanjo in the tow nship o f Ichi. T he shrine, Sanjo O m ido H achim an D aibosatsu, is known simply as O m ido by people living nearby. O ne local legend I heard regarding this nam e m aintained th at the origin o f the nam e O m ido came from o-umi-dd (birthing hall) and th at this shrine had been a place where w om en w ould give b irth .8 This theory may arise from a need to make reference to the fact th at the center is located in Sanjo. M ost likely, the term is simply a double honorific referring to a place o f worship. T h ro u g h o u t the Tokugaw a period, the Sanjo district su rrounding this shrine was w ritten w ith the characters m eaning “birth place,” b u t it is possible that earlier the characters “ place o f divination” were used.9 To arrive at the shrine, one leaves the m ain road th at runs from Fukura to S um oto and heads straight west. A narrow side street passes directly th ro u g h the Sanjo district w here a h undred years ago num erous puppet theaters such as U em ura G ennojo and Ichim ura R okunojo stood side by side, and w here ritual puppeteers m ade their hom es. N ow only the crum ­ bling walls o f these old theaters remain; b u t the shrine w here they w or­ shipped their tutelary kami still stands. O ne pulls o ff to the left o f this small street in to a tiny parking lot. T here, outside the concrete torii (gate)

dem arcating the entrance to the shrine, there stands a large stone erected in the 1960s. Engraved on it are the characters “Awaji N ingyo Joruri no H asshochi” (The Cradle o f Awaji P uppet Dram a). O ne walks th ro u g h the torii into an open area flanked on one side by a large w ooden structure th at once served as a stage and on the o th er by huge C ryptom eria trees, which keep th e spot shady and cool on even the h o ttest o f sum m er days. At the end o f the open precincts stand the two structures th at com prise the shrine: a haiden (worship hall) in the Irim oya style w ith a small stage for rituals, in which the deity H achim an is en ­ shrined and to the left, another smaller worship structure (hokora) dedi­ cated to four figures, all o f which are represented in clay effigies inside: Hyakudayu, D okum bo, Ebisu, and Akiba (the latter said by local people to be th e wife o f H yakudayu, th o u g h few are sure o f th is).10 T he reader will recall th at in the previous chapter I n o ted th at H yakudayu was an epidemic spirit w orshipped at the N ishinom iya Ebisu shrine, and the tu te ­ lary deity o f puppeteers w ho lived and w orked there. O n Awaji, Hyakudayu is th e m ajor figure in a very different kind o f drama. T he date these effigies were m ade and p u t into the shrine is n o t know n, b u t the sculpture said to be H yakudayu has a m arking on the back o f it indicating th at it was repaired in Bunka 2 (1805). Several scholars o f art history on Awaji suggested th at these effigies w ere probably m ade during the G enroku era (1 6 8 8 -1 7 0 3 ). This H achim an shrine is in the Usa lineage.11 In front o f the H achim an worship hall there are tw o lanterns and two guardian dogs (koma inu). Inscriptions on these offerings indicate th at they were dedicated to the shrine in 1805 by puppeteers from the puppet theater know n as U em ura G ennojo, considered the earliest and m ost powerful theater tro u p e on the island. N iim i Kanji refers to O m ido as a subsidiary shrine ( keigaisha) b u t does n o t provide the nam e o f its parent shrine.12 T h ro u g h o u t Japan, shrines are affiliated with m ajor centers th ro u g h a com plex system o f parent and child shrines. F or example, the Nishinom iya shrine is the center (parent) o f Ebisu worship in Japan, but has branch (child) shrines scattered th ro u g h o u t Japan. N ear each o f these branch shrines one will find still smaller local shrines, w hich m aintain some so rt o f affiliation to the re­ gional shrine as a means o f participating in the pow er and authority o f the large center. We see the same pattern w ith H achim an worship as well. As studies o f H achim an w akam iya (child shrines) have show n, the original affiliations frequently becom e blurred or even forgotten in the case o f very small local shrines.13 This is certainly the case o f Ebisu shrines th ro u g h ­ o u t Japan. T here are no w ritten records describing the original affiliations o f O m ido, and it is n o t m entioned in the records o f any larger shrine on Awaji.14

The worship hall housing the sculptures of the mythical founders of Awaji puppetry.

From left to right: Akiba, Ebisu, Dokumbo, Hyakudayu.

T he center for pu p p etry appears at present to be affiliated w ith the K6da H achim an shrine ab o u t ten kilom eters away, evidenced by the fact th at m o st people in Sanjo consider them selves to be ujiko (literally, clan children) o f K5da H achim an. P uppeteers from Sanjo were also in stru ­ m ental in m aintaining K oda H achim an in the early n in eteen th century, for offerings at this site bear inscriptions indicating p uppetry troupes were co n trib u tin g to its u p k eep .15 This affiliation betw een O m ido and K oda has been largely implicit. Ev­ eryone understands th at these tw o sites are connected, b u t since the affili­ ation is n o t m entioned in any records, it is im possible to prove. T h e p a t­ tern o f affiliation o f m ajor shrine (honsha.) to subsidiary shrines in Japan usually has them in closer proxim ity than the K oda H achim an shrine and Sanjo O m ido. A lthough evidence suggests the Sanjo O m id5 center was affiliated w ith Koda H achim an, it is possible th at this affiliation came ab o u t relatively recently, perhaps during the G enroku era. Perhaps the Sanjo shrine originally had o th e r affiliations. W hat w ould a shift in shrine affiliations reveal ab o u t the form ation o f the Awaji tradition? O m ido is actually quite near a larger Ebisu shrine. This shrine, the larg­ est Ebisu shrine on the island, lies just at the edge o f Sanjo and is the Awaji representative o f the N ishinom iya center. T he reader will recall th at at the N ishinom iya center, just outside the m ain shrine precincts in a Sanjo dis­ trict, there was a small w orship hall dedicated to the deity H yakudayu. I propose th a t the Sanjo O m ido site on Awaji was originally affiliated w ith the big Ebisu shrine, itself a child o f the N ishinom iya center. T he location o f the H yakudayu shrine in relationship to th e Ebisu center was m odeled on the N ishinom iya center. T he Awaji center for p uppetry was therefore originally som ething o f a m icrocosm ic version o f the N ishinom iya center. As puppeteers entered into conflict w ith the N ishinom iya center, I sug­ gest, they w anted to break o u t o f this affiliation. So they so u g h t an o th er affiliation for their shrine, shifting their focus to the Koda H achim an shrine even th o u g h it was m uch fu rth er away. O m ido has been the cerem onial center o f Awaji p uppetry for as long as there have been records. In fact, m any o f the w ritten records discussed below situate the origins o f p uppetry in this place. As p uppetry troupes becam e num erous on Awaji in the m id to late seventeenth century, they set up operations around this shrine in the Sanjo district, and O m ido b e ­ came a local place o f w orship. H e re , every N ew Year, any itinerant per­ form er setting o u t from Awaji to perform appeasem ent rituals w ould p resent him self before the worship hall o f H yakudayu and pray for p ro te c ­ tion and success. Later, local joruri troupes on the island w ould com pete for th e rig h t to present Sanbaso before this w orship hall on N ew Year’s Day. (T he h o n o r was usually alternated from year to year betw een the two largest theaters on the island.) Today, the O m ido shrine is quiet b u t is still

used as a local worship site. Its m ain attraction is as a gateball space, and every m orning d u ring the sum m er and fall m onths, elderly local people w ho w ork in th e fields gather to play gateball at around ten o ’clock in the m orning. W hat kind o f tradition was centered here? H o w did it becom e a tradition in its ow n right?

O n o g o r o jim a ’s Sanjo D 5 k u m b o : E arly R eco rd s o f A w aji P u p p eteers It is n o t clear w hen puppeteers becam e active on the island o f Awaji. U sing scattered evidence from the late sixteenth century th ro u g h the early p art o f the tw entieth century, it is possible to piece to g eth e r a probable picture o f the form ative period o f the tradition. O u r earliest record describing the activities o f an Awaji puppeteer is dated 1570. It is a docum ent a puppeteer carried w ith him as he traveled the country perform ing ritual puppetry. It reads: ‘O n o g o ro jim a ’s Sanjo D rikum bo transm its H ikida Awaji’s title. Now , he carries the seasonal cerem onies o f the three shrines to the Im perial co u rt, H e was given the title o f fourth rank. I t is his M ajesty’s Im perial Will w hich decrees this case. R ecorded in the second m o n th o f the first year o f G enki.” 16 We learn from this piece o f evidence th a t by 1570 there was a person nam ed H ikida w ho had already been granted a nam e and a rank and given the rig h t to perform and travel about the cou ntry presenting ritual per­ form ances, particularly those relating to seasonal rituals o f the three shrines, Ise Jingu, Iw ashim izu H achim an, and either Kam o or Kasuga, all directly connected to the Im perial family. T he type o f perform ance to w hich this text refers is probably the Sanbasri ritual, to be discussed in detail in the following chapter.17 T he right o f this puppeteer to perform was generalized to a category o f perform er, referred to in the text as “Sanjo D okum bri.18 In this docum ent, Awaji is referred to by a m ythical nam e, O nogorojim a. T he N ihongi and K ojiki creation accounts tell o f an island called O nogorojim a, the “self-curdling island” th at congealed at the end o f the heavenly jew eled spear low ered into the undifferentiated brine by Izanagi and Izanagi. T he substance th a t dripped o ff the spear tip form ed in to land. T he two deities th en descended o n to this spot, and from this small bit o f land they proceeded to create the rest o f the cosmos. W hat is the significance o f this reference in this d o cum ent and w hat does it tell us ab o u t the early Awaji tradition o f puppetry? In certain readings o f the K ojiki and N ihongi m yths, Awaji is regarded as the placenta o f creation, and the m yths say th at Izanam i and Izanagi “took no pleasure in it.” 19 In symbolic term s, a placenta can be under-

sto o d as a useless by-product o f b irth — necessary for the creation o f life b u t o f no value after life has been started. T h ro u g h o u t prem odern and medieval Japanese history, Awaji was a liminal place, th e island in betw een H o n sh u and Shikoku, a land w here exiled em perors lived, w here imperial princes w ent to kill gam e, w here th e defeated H eike w arriors were said to have fled. This vision o f Awaji presents it as a by-product o f creation, a failed creation attem pt. B ut there is an other, com peting vision o f Awaji to be presented, one th a t makes it n o t only th e first island created b u t also the very navel o f the universe: O nogorojim a. A ccording to Awaji legend, the actual O nogorojim a exists w ithin th e island o f Awaji. In an area near the b order betw een Seidanchb and N andancho in so u th ern Awaji, th ere is a small w orship hall dem arcated by torii. I t is su rro u n d ed com pletely by w ater, creating a tiny island. T h e circum ference o f the m oat is approxim ately twenty-five m eters. A m iniature footbridge connects it to the road and one can walk over to it and stand on the tiny bit o f land. This site is considered to be the first piece o f congealed land created by Izanam i and Izanagi, and it is indicated by a sign explaining th e m ythological event. W hile the small shrine is n o t set in very triu m p h an t surroundings, m ost people I spoke w ith w ere well aware o f it and the story connected w ith it: It was O n ogorojim a.20 W akam ori n o ted this tendency to claim th e primacy (and centrality) o f Awaji in the local un d erstan d in g o f the w orld: “W hile there are different ways o f narrating it, alm ost everyone claims th a t Awaji was one o f the first islands c reated .”21 T he Awaji ningyo tradition appropriated and in som e ways co n trib u ted to this vision o f Awaji’s identity. T he d o cu m en t from 1570 q u o ted above puts this vision o f Awaji in the foreground: Awaji was the center o f the cosm os, the location o f O nogorojim a, and using this nam e for a person from Awaji was a way o f assigning prim acy to b o th the island and its tradition o f puppetry. T he text rem inds the person w ho reads it th a t this perform er is from the very center o f the universe, travel­ ing th ro u g h o u t the land w ith the seal o f approval from no n e o th er than th e em peror himself. As will becom e apparent later (chapter 6), a m ajor dynamic in the Awaji tradition has been insisting on the uniqueness and centrality o f this perform ance tradition vis-a-vis others in Japan.

Ecclesiastical and P opular R eligiou s A u th ority P art o f the tension th a t led to the developm ent o f the Awaji tradition as distinct from th e N ishinom iya center is a com m on dynamic in Japanese religious history. T h ro u g h o u t the M urom achi period and the develop­ m en t o f Tokugaw a Japan, perform ers o f all types were attached to various

tem ples and shrines and were expected to perform at certain tim es o f the year. These Sanjo hoshi, as they were often called, were allowed to engage in w ork outside the shrines at o th er tim es o f year, and this usually m eant “ freelance” ritual and th eater work. As we have seen, puppeteers living in the Sanjo district o f the N ishinom iya shrine w ere clearly conscripted per­ form ers w ith duties related to th e shrine, b u t they also gave occasional perform ances elsewhere. Various references in Japanese history and theater narratives describe the problem s th at arose w hen perform ers w ith responsibilities at shrines and tem ples found th a t w orking on their ow n was m ore lucrative (and possibly m ore fun) than being attached to ecclesiastical centers. F or exam ­ ple, Zeam i w rote o f this problem and adm onished his perform ers to stick to their obligations: “U n d er the pretext o f m aking tours, som e actors neglect the Religious Services; some arrive late, others stay away from the religious service at Kasuga shrine. T herefore, they will com e to a bad end. Even if they are prosperous for a while, at the end they will be punished. T he religious services are ou r basic duty, while tours are just to secure our livelihood in the spare m om ents in betw een.”22 We know th a t while puppeteers were affiliated w ith and lived in the precincts o f the N ishinom iya shrine, their status w ithin the shrine was decidedly low. For the m ost part, they were subcontractors w ith the shrine, operating u n d er the jurisdiction o f the shrine as they spread the w orship o f Ebisu around the countryside by perform ing rites dedicated to th e deity and distributing Ebisu am ulets for a fee. They were required to rem ain attached to the N ishinom iya center in som e way and needed perm ission from th a t center to perform Ebisu rites regardless o f w here they w ent. In this way, the shrine m aintained control over n o t only the pupp eteers’ lives but local access to Ebisu rituals, understood to be helpful in securing a good fishing catch for the year.23 T h e real problem came up w hen they becam e popular in their ow n rig h t and discovered th a t w orking on their ow n was highly lucrative. (It is also possible to discern a new concern w ith artistic freedom . In perform ances outside the shrine, there was greater possibility for artistic and dram atic developm ent.) As we saw at the end o f the last chapter, m any puppeteers started to operate w ith ­ o u t securing proper perm ission from shrine authorities, and their status at the shrine became m ore tenuous. From this m ovem ent arose Awaji puppetry. A key problem in inventing a new tradition concerns finding the structures and histories to legitim ate it, which often entails borrow ing and appropriating existing devices from com peting (and frequently par­ ent) traditions. AU o f the texts we consider below trace the origins o f the Awaji tradi­ tion to th e teachings o f a non-ecclesiastical sacred specialist. An examina-

tio n o f these texts reveals th a t while puppeteers w ere concerned w ith liberating them selves from the Nishinam iya cerem onial center, their chief concern on Awaji seems to have been to establish a parallel sphere o f reli­ gious au thority and practice w ith Awaji as its center, elevating the p u p p e ­ teers to the level o f sacred specialists even th o u g h they w ere n o longer part o f a recognized center. “Awaji p u p p e try ” m eant m ore than just another place for puppeteers. I t co n stitu ted a new trad itio n , constructed against the authority o f th e center in N ishinom iya. R ather than viewing th e m ­ selves as m erely the “little” version o f the “ g rea ter” traditions o f those centers, these itinerant artists pro d u ced their ow n notions o f religious a u thority and legitim ation, founded on their ability to enact magical transform ations o f m atter and spirit and bring a b o u t spirit pacification th ro u g h the m edium o f puppetry. O n the stren g th o f w hat was presented as a sacred and secret d o cu m en t describing th e origins o f Awaji puppetry, they m aintained th eir au to n o m o u s and even imperially sanctioned right to present Ebisu appeasem ent and purification rites th ro u g h o u t the c o u n ­ try. T he case o f itin eran t perform ance gives us som e evidence to suggest th a t for m any Japanese people, the activities o f m ajor religious centers were som etim es largely irrelevant to their lives, and m eaningful and sig­ nificant religious action was m ade available by u nordained, “non-ecclesiastical” sacred specialists. R ecords th ro u g h o u t the Tokugaw a period indicate th a t ritual p uppetry (using the term D o k u m b b m awashi) was a com m on occupation on Awaji. Census records from the m iddle o f the Bunka era ( 1 8 0 4 -1 8 1 8 ), currently stored in the M ihara m unicipal office, indicate th a t in Sanjo alone, 92 o u t o f 144 households listed D ok u m b o mawashi as their occupation. Similar records for (then) neighboring Ichi indicated th a t 22 o u t o f 173 h o u se­ holds were those o f D ok u m b o m awashi, and o th er villages in M iharagun had approxim ately 2 0 p ercen t o f their families listed as puppeteers.24 T hey traveled widely in Japan and yet m aintained a com m on identity as Awaji puppeteers. Just how far from Awaji they traveled is answ ered in p art by a recent discovery. In 1987, a w om an nam ed Ai Suzue living in M orioka in n o rth ­ eastern Japan was cleaning o u t her family storehouse w hen she found p u p ­ pets and docum ents relating to Awaji p up petry (including a copy o f Ddkum bd D enki, to be discussed below ). Research subsequently done by scholars from th e Iw ate Prefecture historical m useum revealed th at these had been b ro u g h t to M orioka in 1641 by Suzue Shirobe, one o f M rs. S u zu e’s ancestors and an Awaji puppeteer. This discovery indicated th at by the m id-seventeenth century, Awaji puppeteers were already traveling th ro u g h o u t the country, presenting them selves as m em bers o f a distinct pu p p etry tradition and p erform ing before lords and even the em peror.25

W hat was the relationship o f these puppeteers to the N ishinom iya cen­ ter? H o w did they understand the nature o f their work? To answer these questions, we have tw o types o f w ritten sources, each follow ing a different line o f interpretation: O ne em phasizes the origins o f Awaji puppetry in a conflict betw een a Nishinom iya shrine priest and a puppeteer em ployed there. T he o th er downplays the N ishinom iya origins, and carefully appro­ priates this earlier tradition while legitim ating Awaji puppetry as a tradi­ tio n in its ow n right. T h ro u g h o u t the years, these two types o f narratives have been interw oven, and oral versions on Awaji reflect a synthesis o f the tw o. AU versions, however, reflect the tensions betw een a Shinto priest attached to a shrine (the Nishinom iya shrine) and a puppeteer w ith o u t a formal shrine affiliation. O u r earliest record, from the m iddle o f the seven­ teenth century, is the formal origin m yth o f the Awaji tradition, consid­ ered to be a sacred docum ent. This d o cum ent provides a fluid central narrative th a t unites puppeteers on Awaji into one tradition and places the practice o f spirit appeasem ent at the center o f their art. This carefully fashioned narrative absorbs the mythical tradition o f the N ishinom iya center, blunts issues o f difference, and all the while insists on the Awaji trad itio n ’s legitimacy. H ere, I present a translation and discussion o f this docum ent following a presentation o f o ur second kind o f evidence, n ineteenth-century histori­ cal references to the early history o f Awaji puppetry. I have chosen to present these sources o u t o f chronological o rder because the apparent intentions o f the earlier docum ent are best appreciated w hen seen in c o n ­ text o f the o th er narratives dating nearly a h u n d red and fifty years later.

A Remembered Past: Α νβαίηβ Priests, L ib a tio n , and a New Tradition T he reader will recall th at puppeteers on Awaji were involved in b o th rit­ ual activity and the presentation o f dram atic en tertainm ent in the joruri tradition. W hile ritual p uppetry tended to dom inate the early develop­ m ent o f the tradition, the theatrical tradition o f ningyo joruri was the main activity o f Awaji pu p p et theaters by the early nineteenth century. Records from Bunka 7 (1811) indicate th at there were over tw enty full­ tim e perform ing troupes on Awaji, h alf o f them located in Sanjo. Several o f them had m ultiple repertoire com panies, which traveled independently from one another all over Japan. U em ura G ennojo, for example, was di­ vided into five separate groups. O ne or tw o always stayed hom e and presented theater on Awaji, while the others kept a heavily booked travel­ ing schedule.

M any ritual puppeteers found seasonal em ploym ent in the large th e a ­ ters, the recorded histories o f which contain passing references to the early history o f ritual puppeteery. D uring th e B unka period, as the popularity o f ningyo jriruri peaked in Japan, there was a flurry o f activity am ong p u p ­ p etry troupes on Awaji, each seeking to d o cu m en t its unique puppetry styles and prove th a t it had been the first and th e best. M any o f these docum ents, which describe the relations am ong different jo ru ri troupes on the island, m ake passing references to the origins o f Awaji p uppetry as a w hole. A n exam ination o f these docum ents provides som e o f the infor­ m ation we have for an early history o f ritual puppetry, even th o u g h it is fragm ented and som etim es unreliable. Because a D 5k u m b 5 mawashi was o f low er social status than a pu p p eteer w ho m anipulated puppets in joruri perform ances, these histories often ten d to dow nplay o r even om it refer­ ences to ritual puppetry. O ne o f th e m ost elaborate descriptions o f origins com es from a text published in 1825 called A w aji-gusa, a u th o red by a father and son, Fujii Yoshin and Fujii Shom in. It is clearly based on an interview w ith som eone in Sanjo describing the early history o f Awaji puppetry. First, said a village elder, at the tim e that Hyakudayu cam e to Awaji, he stayed for several days in the h ouse o f a p uppet-head carver nam ed Kikudayu.26 After that, he [ H yakudayu] m ade love w ith Kikudayu’s daughter, and she becam e p re g n a n t.27 A fte r m o re than a h u n d red days, H yakudayu su d d en ly d ie d from an illness. Because o f this, the imperial order o f H yakudayu28 has rem ained here, and in later years, there were argum ents and litigation b etw een the N ishinom iya puppeteers and the Awaji D o k u m b o . B ut Kikudayu was in p ossession o f the im perial order o f H yakudayu, and therefore it was determ ined that he was the very best in all the land.

A very similar version o f events is recorded twenty-six years later in 1851 in a text by G yosho Sei called A w a ji no ICuni Meisho Zue (A picture book o f Awaji fam ous places).29 This book also contains a draw ing o f an itinerant Awaji puppeteer. Meisho zu e (picture books o f fam ous places or nam e places) constituted a genre o f popular travel guide books during the E do period. A w a ji no K u n i Meisho Z ue is th e Awaji example o f this g enre.30 This d o cum ent seems to have relied on tw o sources, D okum bo D enki (discussed below) and Awaji-j$usa, so it gives us little new inform ation. T he picture it c o n ­ tains may well have been based o n verbal descriptions o f puppeteers, for it closely resem bles the picture o f an Ebisu-kaki presented in the previous chapter. T he text states th at “N ishinom iya’s H yakudayu came to Awaji and b ro u g h t along a p u p p et, and for a long tim e he stayed in this village [i.e., Sanjo] at O m ido while he was sick. A t som e p o in t in tim e, a puppet-

Drawing of an Awaji puppeteer from an 1851 travel text describing Awaji. head carver o f this village by the name o f Kikudayu accom panied Hyakudayu, and w hen he b ro u g h t H yakudayu hom e with him, Hyakudayu made love with Kikudayu’s daughter.” These two versions (which I will consider tog eth er, since the latter seems to have relied on the form er) have a num ber o f im portant points. First, they suggest that H yakudayu was the founder o f Awaji p uppetry and that he originally came from Nishinom iya. Second, they note th at he stayed at the hom e o f a puppet-head carver, which implies that there was already some so rt o f puppetry activity on Awaji prior to the arrival o f the Nishinom iya puppeteers. T hird, they m ention the m arriage o f an outsider to a local girl, a narrative device that allows for the localization o f an im-

p orted practice.31 F ourth, they b oth refer to the litigation b etw een the N ishinom iya center and Awaji puppeteers. This last p o in t, th e controversy over th e rights o f Awaji puppeteers to perform ind ependently from the perm ission o f th e N ishinom iya center, is the focus o f this sou rce.32 Apparently, there was a problem w ith Hyakudayu (or w hoever this person was) bringing ritual puppetry to Awaji and p erform ing. H is imperial order served as his license to perform , similar to the 1 5 7 0 d ocu m en t q u oted earlier. T he n ext version does n o t even m ention H yakudayu but rather begins im m ediately w ith the issue o f litigation. T he puppeteer loses and has to leave N ishinom iya. In an Ansei 4 (1 8 5 7 ) text by a father and son nam ed K onishi T om on ao and Konishi K inko, Ongyoku M ichi Shirube, w e find the follow ing: In the Settsu Nishinomiya Ebisu Shrine, a priest [kannushi] by the name o f Mori Tango had an argum ent over the authority o f the shrine with a priest [of lower rank]33 employed in the same shrine named Mori Kanedayu. In the deci­ sion from the administrative office, Kanedayu was defeated, and so he entrusted his young son with Mori Tango and went to a temple called Shonenji in Amagasaki, where he stayed.34As a means o f making a living, he made small puppets, and using an old sutra box, he walked around the streets o f the town, doing puppet performances based on stories he made up himself, which he sang the way one sings the Tales o f the Heike.35 The spectators really praised him. H e then went up to Kyoto and lived the life o f a wandering puppeteer [ningyo mawashi]. At about the time that there was a fire in the imperial court, and from the land left by the fire, through a crack in the hedge, a young prince saw the perform ance.36 So these performances were seen by people high and low in society.37 T hen, from this prince he received many gifts. The prince also pro­ claimed a title: “O f the various and assorted arts, the greatest puppeteer in Japan; the name Uemura (or Yamamoto) Kanetayu is bestowed upon him .” Later in the fifth year o f Tensho [1578], it was delivered orally by the prince. Later, Uemura Kanedayu went to Awaji M ihara-gun Sanjo mura, and there he taught the art o f puppetry to the poor farmers. H e then received permission from the Lord o f the Castle [probably a reference to the Hachisuga family in Awa, discussed below] and began to give puppetry performances. This was the beginning o f joruri shibai. This d o cu m en t su ggests that it was a form er N ishinom iya puppeteer, kicked o u t after a dispute w ith another person from that shrine, w h o was given the nam e U em ura and w h o cam e to Awaji to start puppetry, al­ th o u g h it makes no m en tion o f ritual puppetry and uses the neutral term ningyo m awashi. D u rin g the Tokugawa period, family nam es were granted to peasants or artists to give special recogn ition for m erit, and possibly the puppeteer was given this nam e to h on or his perform ance.

W hat attracts ou r atten tio n is the detail th at this puppeteer entered into a conflict w ith the Nishinom iya center, becam e a w andering puppeteer w ith no affiliations, stayed for a while in a nunnery in Amagasaki, and eventu­ ally ended up on Awaji. T he next version appears in an anthology edited by the scholar o f Edo history, M itam ura Engyo, entitled A w a ji A w a Ningyo Z a K yiiki Zassan (Assorted old docum ents for the Awaji and Awa p uppetry tro u p e s).38 O ne docum ent from the anthology is Ningyo Ganso H onke Uem ura Gennojo Jireki (T he original puppeteer family U em ura G en nojo’s history). T he text does n o t have a date b u t is th o u g h t to have been w ritten in the Bunka period (1804—1817). It states: D oku m b b was from the Sanjo district o f K yoto, and so he was called Sanjo D o k u m b b . M anipulating puppets, he served the L eech Child o f the N ishinom iya shrine. After his death, a priest \kan n u sh i\ nam ed M ori Kanedayu succeeded D ok u m b o. A nd then, after that, Hyakudayu o f N ishinom iya, the founder o f puppetry, starting with the Leech Child o f N ishinom iya, traveled all over the country serving in village shrines, both along the coast and inland. Eventually, he came to Awaji, and there he stayed a lo n g tim e at the h om e o f H ikida. T here, he married H ikida’s eldest daughter, and she gave birth to a baby son. This child was called Shigedayu, and after H yakudayti’s death, he took the name “G en n ojo” (T he O riginal). This Hikida Awaji nam e becam e the originator o f m anipulated puppets, and he taught the art and techniques o f Hyakudayu to the farmers and established the Awaji troupe [z a \. At the tim e o f Em peror G oyozei, he perform ed before the em peror. A nd then in 1 5 7 0 , in the second m onth, an imperial name was b estow ed on him. After this, H ikida took the stage name o f U em ura.

This docum ent has three notew orthy features: First, it gives us a source for D 5kum b5, asserting th at he was from a Sanjo district, a twist we do n o t see in any other source, even the N ishinom iya sources for D bkum bo. Second, it traces the origins o f Awaji puppetry from its mythical beginning w ith a priest nam ed D 5kum b5 to a priest nam ed M ori Kanetayu, w ho som ehow hands the rights over to a person nam ed H yakudayu; the latter comes to Awaji and m arries in to the H ikida family, which finally changes its name to U em ura. T hird, it indicates th at this person was given his nam e by an em peror, establishing the oldest puppetry troupe on Awaji. In another nineteenth century text, Zdho Jdruri Taikeizu (The new and revised joruri lineage chart), we learn the following: A man named H ikida, w ho hailed from Awaji, had from birth a great deal o f talent in all sorts o f perform ing arts o f th c f u r y ii style, and he was able to make different faces for you n g and old , m en and w o m en , and he devised clever ways to express a w ide range o f em o tio n s,39 and was able to m anipulate puppets

so that they were just like hum an b ein gs. This was seen by m any lords, and he was praised as “hallmark and founder o f Japan’s m any perform ing arts.” O n o n e o f those occasion s H id eyosh i heard ab out h im , saw his perform ance, and was deeply m oved . H id eyosh i granted him perform ance space in the Shijo Kawaram achi district in K yoto, and there his perform ances were seen by m any p eople. A fter this, Shigedayu was su m m on ed to th e im perial court, and the em peror saw the perform ance and gave him the title H ikida Awaji T ayu.40

A lthough th e text m akes no m en tio n o f the N ishinom iya origins, it indi­ cates th at this im p o rta n t Awaji pu p p eteer was granted a title by the em ­ p e ro r before w hom he perform ed. I t is n o t possible to m aintain th a t all o f these versions w ere based o n the same source. O u r in te n tio n was to determ ine how they represent tensions concerning the right to perform . T w o ideas stand o u t in these sources w hen they are read to g ether. First, it is clear th a t Awaji puppeteers originally came from N ishinom iya and strug gled to find an identity for them selves in relation to this center. Sec­ o n d , these early puppeteers w anted to have their ow n tradition. These origin accounts suggest th a t having a h istory and being a tradition was a form o f legitim ation in its ow n rig h t.41 N ow we tu rn back in tim e from these sources to a discussion o f the earliest version o f a “ h istory” o f the Awaji trad itio n , a text th a t, I w ould argue, was carefully pro d u ced to create a unified tradition.

A Dead Priest and an Angry Deity: Dokumbo Denki, Kan’ei 15 (1 6 3 8 ) Puppeteers w ho were centered on Awaji in th e Sanjo district were n o t m erely a scattered group o f freelance ritual artists: they were a tradition, created n o t th ro u g h consensus ab o u t perform ance style b u t th ro u g h the p ro d u ctio n o f a unifying narrative. To be an Awaji puppeteer m eant one shared the same story w ith o th er puppeteers ab o u t w ho you w ere, w here you came from , and why you did w hat you did. O th e r aspects o f tradition were secondary to this central story. Part o f the w ork o f tradition is to ensure its continuity and survival, to provide a vision o f consensus and center, and to be, in the m ost basic o f term s, hegem onic ab o u t discourse.42 T he founding m yth o f the Awaji trad itio n , D okum bd D en ki, is a fasci­ n ating do cu m en t in Awaji p u p p e try ’s history. Because it is clearly a m ythic narrative, it has been left aside by scholars studying th e origins o f the Awaji tradition. F o r exam ple, in his essay “ G enroku ki Awaji Ayatsurishibai no C hiho K oki,” Sakaguchi B unno discusses the history o fU e m u ra G ennojb tro u p e b u t dismisses this text w ith one sentence: “ C oncerning

the origins o f Awaji puppetry, beginning w ith the anonym ous Dokumbd D enki, sources narrate som ething about a connection betw een p u p p e­ teers attached to Nishinom iya and Awaji Sanjo-m ura.”43 But the structure o f this m ythical docum ent, particularly w hen it is read nonchronologically, reveals its purpose. Clearly both mythical and historical, the text narrates the founding event o f the tradition and creates and legitim ates Awaji puppetry as a tra ­ dition in its ow n right. Every Awaji puppeteer carried a copy o f this do cu ­ m en t as he traveled th ro u g h o u t Japan. H aving a copy o f the scroll m eant you were part o f the Awaji tradition. My hypothesis is th at this text represents a creative fusion o f a large num ber o f elem ents already present on Awaji at the tim e it was w ritten; it serves as a confederative docum ent for various factions; and it reveals how a perform ance-based cultural practice creates a m ythical tradition. We are able to see the considerations th at create and give form to a tradition. The basic narrative m utes the historical issues while absorbing the m yth tradi­ tion o f Nishinom iya and even the larger issues o f Ebisu w orship. I t also blunts the conflicts betw een different claims to origins on Awaji (U em ura, Kanedayu, etc.) and creates a new identity for the enigm atic Hyakudayu. Hyakudayu, I argue, is the nam e given to the first puppeteer in this narra­ tive because he is the paradigm atic puppeteer o f the tradition, capable o f being at once every one and no one (n o t unlike the puppets he is said to m anipulate). Even th o u g h it is o u r earliest w ritten narrative, I argue that this text is a late creation in the phenom enon o f puppetry on Awaji. R ecorded sixtyeight years after our docum ent identifying Awaji puppeteers (the 1570 “license” ), it was designed to m eld a n um ber o f com peting claims to ori­ gins on the island, thereby enabling Awaji puppetry to possess a m yth o f origins legitim ating it as a distinct tradition against the pressures from Nishinom iya shrine authorities seeking to reassert their control. The Awaji tradition developed from com peting factions each claiming to be the first. Dokumbd D enki reflects w hat becam e the canonical version o f events, free to be nonhistorical and even fanciful. As I have discussed elsewhere in my studies o f Japanese ritual and its relationship to m yths, m yth as a narrative structure makes it possible to include a num ber o f diverse elem ents and synthesize them in a unique way.44 It is n o t merely th a t m yth allows one to be loose w ith the facts. Rather, m yth makes facts secondary to religious and ideological senti­ m ents and concerns, which becom e pow erful and at tim es quite danger­ ous. M ythical tim e is n o t just about “realities lived and stories to ld .” I t can be about realities rew ritten and stories retold o r left o u t as well. The text o f Dokumbo D enki is solely in Chinese characters.45 T he narra­ to r claims that the contents are based upon a previous w ritten docum ent

lost at the tim e o f its com pilation, an o th er rhetorical device th at allows the text to transcend history while at the same tim e claim ing to narrate it definitively.46 T h e to n e o f the tex t gives th e im pression th at it is narrating an old oral trad itio n , and it uses a standard genre for origin texts in Japa­ nese religions, know n as engi, in w hich the beginnings o f the text fre­ quently reiterate lines from the Japanese m ythical narratives o f the N ihongi and K ojiki (tw o heavily constructed narratives in their own rig h t).47 T he text is divided in to tw o rath er distinct sections. T he total n u m b er o f characters in th e text is 1079, and an additional sixty-eight characters appear at th e end o f the text in the p ostscript ( okugaki), indicat­ ing the date and conditions u n d e r w hich it was recorded and the nam e and rank o f th e scribe and reciter. Scrolls o f Dokum bd D enki w ere regarded as sacred objects by p u p p e ­ teers and w ere treated w ith special care and devotion clearly because this d o cu m en t provided a center for th e tradition. T he scroll was kept on a kam idana in theaters on Awaji and was carried in the box containing ritual im plem ents for th e p u p p e t Sanbaso w hen a p u p p eteer traveled. AU the p u p p e t troupes originating from Awajishim a d u rin g the Tokugaw a period to o k a copy o f this d o cu m en t w ith them . T he d o cu m en t has been found in the locations to w hich Awaji p u p p etry spread: Yamanashi, N agano, and M orioka. In each discovery, th e text had been carefully stored, frequently in th e same box w ith the sacred p uppets o f Sanbaso, O kina, and Ebisu. These puppets, unlike those depicting jo ru ri characters, are afforded a special status, since they are used in th e perform ance o f sacred rites. W arnings in the th eater lore often state th a t in the event o f fire, the San­ baso p u ppets and the Dokumbo D enki m ust be saved, o r else terrible things will ensue. To date, the existence o f eight copies o f this text have been ascertained: one in th e K uroda N ingyb Za in N agano Prefecture, a th eater started by an Awaji p u ppeteer in th e eig h teen th century; three in the Awaji N ingyo Joruri Kan (Awaji P u p p et T heater) in Fukura; three in th e hom e o f the T oyota Hisae family in Sanjo m ura o n Awaji, descendants o f the U em ura G ennojo tro u p e ;48 and one in M orioka in 1987 in the hom e o f Suzue Ai, the descendant o f an Awaji p u ppeteer w ho perform ed in M orioka in 1648. A fu rth e r copy, discovered in Yamanashi Prefecture and m entioned by N agata K5kichi in his book Ikite Ir u Ningyo Shibai, has been subse­ quently lost 49 T he calligraphy styles o f the eight scrolls differ considerably. T he date o f all eig h t is K an’ei 18, a lth o u g h som e are dated in the sum m er and som e in th e tw elfth m o n th o f th a t year. A ccording to the okugaki o f all the texts, the person w ho narrated th e c o n te n t o f the text is a certain Sakagami N yudo, whose nam e indicates th a t he was a m em b er o f the B uddhist clergy b u t a b o u t w hom n o th in g is know n.

That each subsequent branch o f the Awaji tradition had as its founding narrative the Ddkum bd D enki accounts for the tradition’s perceived co h e­ sion over several centuries, despite the itinerant status o f many pup p e­ teers. T he text provided a vision o f the center for the tradition. G eograph­ ically, the center is Awaji. In narrative terms, the center is the story o f Hyakudayu. Religiously, the tradition centers on appeasem ent rituals. T he narrative can be divided into tw o parts.50 T he reader will recall that in the previous chapter, the deity Ebisu was grafted o n to the mythical failed creation attem pt o f Izanagi and Izanam i. This small child deity, the Leech Child, was abandoned on the waves by its parents. T he first part o f the narrative (n ot translated here) is an abridged but verbatim account o f the N ihongi creation account. T he text presents that myth up to the point where the Leech Child is set afloat. This is usually w here the Leech Child is left in standard m ythologies in Japan. This is w here the Awaji tradition o f puppetry begins. Dokum bo Denki [second part] T he Leech Child drifted on the waves for m any years and m onths. Before that, he arrived at W akokuzaki and had the shape o f a wheel. T here he became a kami o f light. A t th a t tim e, there was a fisherman by the nam e o f M urogim i. (From this tim e, the leader o f the fisherm en began to be called M urogim i. Later his surnam e becam e Fujiwara, and his first nam e was H yakudayu Masakiyo.) A t a certain tim e, he was riding in a fishing bo at w hen suddenly the sky became dark, clouds g athered and darkened the sun, and lightning flashed all around. N o tin g the strangeness o f this occurrence, Hyakudayu drew near to a small child he saw floating in the water. T he child had the shape o f a kami and was only ab o u t twelve or so. T he child tu rn ed to face him and delivered the following oracle: “ I am the Leech Child o f long ago. U ntil now, I have had no w orship hall. Build me a tem porary w orship hall on the seashore.” T he hall built as a result o f this oracle was the Nishinom iya Daim yojin (Ebisu Saburb D en) in Nishinom iya. In this shrine, a person by the nam e o f D o kum bo was capable o f m ediating for this kami and receiving his messages. But after the death o f D bkum bo, there was no one to appease the deity, and so the Leech C hild caused heavy rains and winds, fishing disasters, and mishaps o n land. H yakudayu reported this to the head o f the Fujiwara family [in K onoeden], w ho was in the capital, and an imperial o rd er came dow n. It said th a t he should make a puppet w ith the same face and posture as D bkum bb. Follow ing this order, he m anipulated the puppet before the w orship hall o f the Leech C hild, and the deity’s spirit calm ed dow n. After this, puppets o f this type in th e likeness o f D bkum bo were unusually effective in appeasing the Leech Child. A fter this, H yakudayu w ent around to

m any provinces, w orsh ip pin g m any god s. After this, H yak ud ayu-D oku m b d 51 stopp ed at Awaji and transm itted this art. It was at this place that the god s created the great cou n try o f Japan, and here is the island O fT oyoakizushim a.32 After that, he was titled ayatsurim on o [a p up peteer], H yakudayu lived at Awaji M ihara, Sanjo, and there he transm itted this art. It is said that here the people o f old w orshipped the myriad (eigh t m illion) kami. After his death, he was w orshipped som ew h ere in N ish inom iya. H e was given the imperial ed ict from the capital to appease divine spirits, and as an appeaser o f kami H yakudayu was later given th e fo llo w in g proclam ation: “T he great cou n try o f Japan is a divine country. T herefore, this person w h o appeases divine will is the ultim ate person o f m any talen ts.” T his art is for th e appeasem ent o f kam i. H ereafter, p eo p le should n o t take this lightly. I f they d o, it will w eigh heavily upon them . P eop le after this sh ould be sorely afraid.

w a rn in g :

p o s t s c r i p t : T here was a record o f this story before, but it has been lost, and the previous was based u p on a secret letter from the h ou se o fY o sh id a , but since the old d o cu m en t has n o t been w ell ch ecked, I have believed only the oral version.

Sakagami N yu d o M idsum m er, Kan’ei 15

It is apparent from looking at this tex t, even in translation, th a t it as­ sum es a certain a m o u n t o f know ledge o f the characters and events in ­ volved. T he transcriber claims he was attem p tin g to recreate a w ritten d o c u m e n t (th e secret letter) by listening to som eone narrate its contents. This “ secret le tte r” was itself probably based u p o n an oral tradition. I t is im possible to determ ine w h eth er th e original d o cu m en t ever in fact ex­ isted. Regardless, this device in the text allows for the creation o f a new version o f events, while n o d d in g to the authority o f an other, older record. H e re , I w ould like to focus on five strategies o f trad itio n form ation in this docum ent: the appropriation o f m yths and m otifs from a larger and stable religious center; the form ulation o f a religious problem and a ritual solution; the expression o f legitim ation from outside the tradition; the establishm ent o f a new geographical center; and th e expression o f the new tra d itio n ’s pow er and im portance. These five strategies indicate th at this d o c u m e n t was concerned n o t m erely w ith recording w hat had happened before, b u t also w ith launching a new tradition o f ritual puppetry, distinct from th e tradition o u t o f w hich it developed. This text absorbed tw o m yths from th e N ishinom iya tradition and fram ed them to highlight th e Awaji pu p p etry tradition. First, the text re ­ fers to the arrival o f the Leech C hild at the beach at W akokuzaki. T he reader will recall th at the engi text o f the N ishinom iya shrine narrates

th e strange occurrence o f a floating effigy being repeatedly caught in a fisherm an’s net until finally in a dream it is revealed th at this object is none other than the deity Ebisu. T he Nishinom iya version specifically m entions a place-nam e—W adaham a. T he Awaji narrative has absorbed this part o f the N ishinom iya engi and has altered the nam e slightly to read W akokuzaki.53 T he second m yth th a t this Awaji docum en t takes from Nishinom iya concerns the puppeteer D ok u n , here called D okum bri. A ccording to the N ishinom iya m yth o f puppetry, recounted in chapter 3, “ In the era o f the deity Iw akusubuna, in this inlet, there was an old m an nam ed D okun. W hen he heard th at the deity at this shrine at th e age o f three still could n o t walk, he m ade a pup p et to entertain and raise the child deity.” In Dokumbd D enki, this mythical tradition is continued b u t altered slightly. T he N ishinom iya text specifically says th at a person nam ed D dkun used a doll to appease the Leech Child. T he Awaji tradition states th at this per­ son nam ed D okum bo (a nam e w hich already m eant “ pu p p eteer” ) merely entertained the Leech Child. But the Awaji version then adds a new d i­ m ension to the account, and this in tu rn creates a new ritual developm ent. A ccording to the Awaji tradition, a crisis arises w hen the priest D dkum bo from the N ishinom iya shrine dies. This Leech C hild, a m al­ form ed and failed creation abandoned by its parents w ho com es floating in o ff the waves, is suddenly w ithout a hum an officiant. W ho will see that his potentially m alevolent nature is appeased, now th at D okum bo is dead? W ith no one to take D o k u m b o ’s place, disasters o f all sorts occur. The point could n o t be m ore clear: Awaji puppetry was started by the person w ho literally stepped in and to o k th e place o f the original priest, m aking this new tradition the rightful heir to D ok u m b o ’s ritual authority. T he text then uses this narrative m om ent to launch a new beginning (that is to say, a retelling o f “history” ) for ritual puppetry. W hereas in the Nishinom iya myths it was D okun w ho first used a doll (or p u ppet), in the Awaji version o f events the first pup p et is used by a person nam ed H yakudayu. Dokumbd D enki narrates th a t H yakudayu received his orders from the imperial family indicating w hat should be done: make a puppet in the likeness o f D okum bo to appease this angry deity. T he text relates that n o t only was this particular pup p et effective, b u t thereafter all puppets in this likeness w ere effective in appeasing the Leech Child. This fact, n o ted in the text, allows for the tradition to expand to include m ore than one puppeteer by m aking the potency o f the pu p p et generalizable to m any puppets. It is n o t en o u g h for Hyakudayu to have such authority. His au­ thority m ust be extendable to subsequent perform ers. T he pu p p et has becom e the priest, a device w hich allows the Awaji tradition to absorb the ritual authority o f the Nishinom iya center while deflecting the challenge this entailed.

A t this p o in t, it is im p o rta n t to n o te w hat seems to be a veritable co nfu­ sion o f nam es and places in D okum bo D enki. W hy does the place-nam e W adaham a g et changed in the Awaji text? W hy does a priest called D 5kun here get called D 5kum b5? W hy does a deity nam ed H yakudayti in Nishinom iya becom e a person in the Awaji account? O n the one hand, the text could be read as a garbled account o f the N ishinom iya records, som ehow confused in the process o f transm ission. Such an in terp retatio n , however, is based on the assum ption th at oral traditions have no safeguards for accuracy at the level o f detail and can m aintain only the m ost general o f p lot lines. T his is clearly n o t th e case. I w ould argue th a t th e confusion o f nam es and places in this tex t is intentional. W hat w ould such a shifting o f nam es accom plish for the process o f trad itio n creation? D okum bd D enki, as a d o cu m en t unifying Awaji puppeteers in to one self-conscious trad itio n , so u g h t to p resent a narrative th a t could be any p u p p e te e r’s narrative b u t th at rem ained distinct from th e N ishinom iya center. By relying o n the figure o f H yakudayu as its central character, the tradition allow ed for a m ythical hero. Just as H yakudayu’s authority gets generalized in the text, so to o does his prestige in the tradition. H yakudayu could be any Awaji puppeteer. A nother feature o f this text stands out: its insistence th a t the decision to make a p u p p et “in the likeness” o f D o k u m b o was th e result o f an imperial order. W hen com bined w ith the title th a t H yakudayu is said to have re ­ ceived, the result is a legitim acy for the new trad itio n th a t the N ishinom iya authorities could n o t challenge. T he latter p a rt o f th e text focuses on the transm ission o f the a rt o f pu p p etry to the cen ter on Awaji, referring to it by yet a n o th e r m ythical nam e underscoring its primacy, Toyoakezushim a, “ th e w o rld-opening island.” This m id-seventeenth cen tu ry d o cu m en t served to unify the Awaji tra­ d itio n early in its developm ent. L ater texts revealed different textures w ithin this trad itio n , b u t Dokumbo D enki gave it a narrative center.

C on trol o f Awaji P u ppetry by th e H achisuga Fam ily A lthough puppeteers w ere able to break away from N ishinom iya and es­ tablish their ow n tradition o n Awaji, it w ould be inaccurate to assum e th at they enjoyed a w orld free from the constraints o f status codes. Awaji p u p ­ peteers, as we n o ted in chapter 2 , were regarded as outcasts in Tokugaw a Japan, and althou gh there were o th e r outcast groups w ith low er status, ritual puppeteers suffered from restricted m obility and discrim ination. Ritual puppeteers in particular had a low status b o th because they were itinerant perform ers and because they perform ed sacred rites in which they were p olluted by co ntact w ith beings from an o th er realm. W hile

m ost outcasts in Japan were under the control o f the leader o f the o u t­ casts, D anzaem on, Awaji puppeteers were not; they were controlled by the same family th at controlled the rest o f the Awaji-Awa region, the H achisuga family, whose main castle was on Shikoku in Awa, near T o ­ kushim a. It is widely th o u g h t th at this exception to the rule o f D anzaem on’s leadership explains why Awaji troupes (which num bered over forty during the nineteenth century) becam e such a dom inant puppetry tradition and enjoyed so m uch freedom to travel and perform . H ow Awaji puppeteers came to be u n d er the control o f the H achisuga family is the subject o f an interesting legend, the veracity o f which cannot be ascertained. A ccording to the legend, because this family supported the Tokugawa family, w hen the head o f the H achisuga family w anted to send letters to E do to discuss the war, he had a difficult tim e enabling his retainers to travel there from Shikoku. T he fastest and safest route was overland across Awaji, b u t this area was controlled by the opposing army. Samurai w orking for the H achisuga family from Awa, so the legend m aintains, w ould disguise themselves as puppeteers from the U em ura puppetry troupe. They th en could travel back and forth across Awaji and even over o n to H o n sh u , and in this way were able to carry o u t their duties unnoticed. T here is an old docum ent discovered in N ishinom iya in the hom e o f the Taida family that narrates this story. A ccording to this docum ent, a m an nam ed Taidai G enzaem on received an order from his lord, the head o f the H achisuga family, saying th a t he was to go to the battle camp o f Tokugawa Ieyasu, and so he set o ff from the dom ain o f H achisuga in Tokushima. G enzaem on w ent overland th ro u g h Awaji, and in ord er to break th ro u g h the netw ork o f the opposing forces, he m ade an arrange­ m ent w ith the head o f the U em ura puppet troupe and disguised him self as a puppeteer. W hile they were doing p uppetry perform ances, he slipped away and m anaged to get to Ieyasu. W hen he arrived at Ieyasu’s cam p, he was wearing the damask coat o f the U em ura puppeteers over his armor. G enzaem on sang the praises o f U em ura to Ieyasu, and from his ow n nam e he to ok one character and gave it to U em ura, and after th at U em ura was called U em ura G ennojo. T herefore, the legend m aintains, the H achisuga family protected the puppeteers because they had allowed their samurai to pass as puppeteers in the decisive final days o f the war. U nder the protection o f the H achisuga family, Awaji puppetry was able to flourish. In addition to the num erous ritual puppeteers w ho lived on Awaji in Sanjo, m any p u p p et theaters developed and were able to m aintain rigorous touring schedules all over Japan. A standard phrase on Awaji referred to the “Awaji forty,” although this large num ber probably counted the repertory com panies w ithin large troupes as individual thea­ ters. A ccording to records dated 1811, there were tw enty troupes on

Awaji, half in Sanjo and the o th er half in Aibara. Aw aji-gusa (1825) lists eighteen troupes. T h e U em ura G ennojd tro upe divided into five reper­ to ry com panies. T he num ber o f these troupes rem ained constant until the beginning o f the Meiji period (1 8 6 8 ), w hen they slowly declined. By the m iddle o f the Showa period (1 9 2 5 -1 9 8 9 ), only four theaters rem ained, and shortly after the war, they existed in nam e only. T he history o f these p uppetry tro u p es is beyond the scope o f this study o f ritual puppetry, b u t it should be n o ted th a t m any o f the rituals we examine in the next chapter were m aintained by puppeteers w ho w orked m ost o f the year for these larger troupes. We turn next to a discussion o f w hat ritual puppeteers on Awaji did w hen they “m ade D okum bo dance.”

5 Puppets o f the Road, Puppets o f the Field: Shiki Sanbaso, Ebisu-m ai, and P uppetry Festivals on Awaji

Kctdozuke N in g y δ: Puppets o f th e R oad W hen these D 5kum b5-m aw ashi from Awaji visited hom es and perform ed on fishing docks and near shrines, w hat did their rites and puppets look like? H ow were these puppeteers organized am ong themselves? H ow were their perform ances received? W hat happened to their puppets w hen they were broken and could no longer be used? H ow do people living on Awaji today rem em ber these performances? These are some o f the ques­ tions addressed in this chapter. As I have pointed o u t, evidence for a study o f ritual puppetry is frag­ m ented, due to the low status o f puppeteers and the popular, oral, and highly localized nature o f their traditions. By the tim e we get to the devel­ opm ent o f the Awaji tradition, there is m ore evidence. P uppet kashira and im plem ents, descriptions o f perform ances by people w ho rem em ber see­ ing them as children in the prewar period, transcriptions o f the w ords recited in the perform ances, old photographs, and even films o f som e o f the last perform ances o f the early postw ar period provide enough infor­ m ation to piece to g eth er a fairly accurate picture o f ritual puppetry. T h ro u g h studying and com paring the puppets, records and rituals that developed o u t o f the Awaji tradition, including revived perform ances on Awaji, it is possible to get a clearer idea o f w hat Awaji D okum bo-m aw ashi did in p uppetry’s heyday during the Tokugaw a period. In this chapter, the discussion focuses on the sacred rituals o f the itiner­ ant puppeteers w ho hailed from the center o f ritual puppetry, Sanjo, on the island o f Awaji. These puppeteers, the tradition m aintains, are the professional descendants o f the fisherm an-turned-deity-appeaser Hyakudayu. They perform ed appeasem ent and purification rites from d oor to door at set seasons o f the year, before audiences ranging from the poorest peasants to the imperial family. W hile m any o f them also w orked at the m ajor theaters on the island th ro u g h o u t the year as puppeteers for ningyo joruri, they drew sharp distinctions betw een these perform ances and their ritual duties, referred to as shinji or kamigoto (sacred m atters). M any also farm ed small plots o f land.

T heir ritual perform ances were o f tw o types. O n the one hand, there w ere Sanbasd perform ances, solem n and graceful rites o f magical purifica­ tion and revitalization en treatin g three deities to visit th e hum an realm to bestow felicity and blessing th ro u g h the apotheosis o f a p u p p e t into a visiting deity. O n the o th er hand, there were h um o ro u s deity plays d ep ict­ ing and h o n o rin g the deity Ebisu, his love o f sake, and his control o f the fishing catch.

P u p p e tr y as a P ro fe ssio n : N o m e n c la tu r e , S ta tu s, an d T erritories T h e reader will recall th a t puppeteers on Awaji to o k their nam e from the m ythical priest at the N ishinom iya shrine, D okun (D rikum bri according to the Awaji texts), w ho entertained and appeased the Leech Child. U p o n D o k u n ’s death, th e puppeteer H yakudayu is said to have m ade a pu p p et to appease this angry (and perhaps grieving) Leech Child. This mythical narrative, and its religious ideas o f deity appeasem ent and co ntainm ent th ro u g h ritual perform ance, form ed the core o f the Awaji p uppetry trad i­ tion w ith the D rikum bo-m aw ashi as its sacred specialist. To be a Drikum bri-m aw ashi m eant th a t one presented one or b o th o f these rites, Sanbasri and E bisu.1 Shiki-Sanbasri and Ebisu-m ai rituals should be considered a pair in the Awaji tradition. These tw o perform ance form s have distinct lines o f devel­ o p m en t in the history o f Japanese p erform ing arts, b u t by the tim e th e Awaji tradition em erges, they are interrelated, mainly th ro u g h the story presented in Dokumbo D en ki.2 T he tw o rituals are opposed to one an o th er in their intentions. Ebisu is regarded as a polluted deity, w hereas the deities invoked in th e Sanbasri ritual are those o f purity and vitalization. W hen seen as p art o f the same tradition, the tw o rites suggest the awareness th at b o th p o llution and purity are pow erful and necessary. By the early n in eteen th century, Drikum bri-m aw ashi began to special­ ize in either Sanbasri or Ebisu perform ance. I t is tem p tin g to argue for this developm ent on religious or even aesthetic grounds, b u t it should be n o ted th a t this specialization also coincides w ith the developm ent o f larger puppets. W hereas th ro u g h o u t the sixteenth and early seventeenth centuries the puppets were small eno u g h to be presented from a box hung around the neck, by the end o f the seventeenth century, as the practice o f ningyri jriruri becam e popular and Awaji kashira becam e larger, the size and w eight o f the puppets m ade it difficult to carry so m any at one tim e. A specialist in Ebisu m ight carry only one pu p p et, whereas a Sanbasrimawashi m ig h t carry one, tw o, or even three puppets along w ith ritual im plem ents.

A nother reason this specialization occurred is likely due to the higher status afforded to Sanbaso perform ers, as we shall see shortly. T he Ebisu rite, because it has a simpler series o f perform ance dem ands, was easier to m aster, and those w ho were skilled in the Sanbasd perform ance, which included m usic, percussion and elaborate puppet m anipulation, were likely regarded as m ore skillful artists.

Perform ance T erritories a n d Patronage Clearly it was n o t possible to su p p o rt a family th ro u g h ritual puppetry alone. Consequently, m ost D okum bo-m aw ashi had o th er form s o f em ­ ploym ent to augm ent their incom es. Even th o u g h ritual puppetry may have occupied som e o f these perform ers for only several weeks a year, it is notew orthy th at those census records kept during the Tokugawa period th at m entioned puppeteers at all listed people on Awaji w ho did puppetry as Dokum bri-m aw ashi, regardless o f how m uch o f their incom e this repre­ sented. Being an itinerant puppeteer was a quality o f being, som ething th at gave one a fixed (albeit low) status in society. Itinerant puppeteers were organized in tw o ways. A large num ber o f them w orked for the m ajor puppet theaters on th e island. D uring the periods o f the year w hen they w ere n o t to u ring with these theaters, they were free to perform as D okum bo-m aw ashi. They were essentially sub­ contractors in these theaters. O th e r itinerant puppeteers had no affiliations w ith p roper theaters but depended on alternate form s o f livelihood. M any o f these freelance p u p ­ peteers had small plots o f land, given to them in the early seventeenth century w hen land was given to outcasts by the governm ent to improve the yield o f villages and open up new land.3 These people were regarded as kawata hyakusho (unclean farm ers), and th eir presence was often re ­ sented by people living near them . B ut this m eant th at as itinerant pu p p e­ teers, they were free for a greater p art o f the year than those em ployed at the m ajor theaters. These itinerant perform ers, regardless o f their patronage and affilia­ tions, presented their rites th ro u g h o u t the island o f Awaji. Niimi notes th at while m any o f them stayed close to the center o f Awaji, som e traveled to its n o rth ern tip and across to the H anshin area (Kobe and Osaka). Som e to u red Shikoku and w ent to Kyushu. T here are also reports o f Awaji D okum bo-m aw ashi perform ing in Kyoto, and o f course the case o f Suzue Shirobe in M orioka, discussed in the previous chapter, shows th a t some ventured far from Awaji and never came back. As noted earlier, puppeteers, agreed upon territories in order n o t to overlap w ith one another and cause unnecessary hardship. Niimi relates an

interview w ith an elderly pu p p eteer in 1970, in which they discuss these territorial restrictions, called nawabari. T he puppeteer, Mr. Saitri, was in his eighties at the tim e o f the interview, which m eant he was perform ing as a Dokum bri-m aw ashi in the 1920s and 1930s. We learn from this inter­ view th at individual puppeteers agreed upon territories th ro u g h the main perform ance theaters w ith which they were connected. So, for example, if one were affiliated in som e way with the H isadayu tro u p e, a p u p p et theater on the island, one was free to perform in the Ama region o f southern Awaji, b u t n o t in Seidancho. Som e areas were very strict about their terri­ torial agreem ents, but the area surrounding Fukura (where the Awaji Ningyri Jriruri Kan is today) was an open territory, and anyone was allowed to perform there. But if a puppeteer from H isadayu was asked to perform in Seidanchri, for example, they w ould have to refuse the offer and tu rn the request over to the D okum bo-m aw ashi who controlled th at territory.4 Mr. Saito n o tes, however, th a t som e people had their ow n patrons outside the established theaters, and these puppeteers w ould ignore the territories and perform w here they wished. H e was probably referring to puppeteers w ho did n o t rely on established theaters for their seasonal em ­ ploym ent. These puppeteers w ere referred to as f u r ifu r i Sanbasd (casual, “ by chance” Sanbasri). I asked a n u m b er o f people ab o u t this category, b u t few recognized it. O n e m an told m e th a t puppeteers w ith no theater affiliations w ere seen as being o f low er social status th an those from the theaters. T hey were considered less skilled and artistic. These territories served a n u m b er o f purposes. They ensured th a t a single given area w ould n o t be flooded by ritual perform ers and thus m ade it m ore likely th at puppeteers could make som e so rt o f a profit. They also ensured th a t their w hereabouts were know n to th e authorities, and th at they w ould re p o rt their earnings to the H achisuga family in Awa th ro u g h their base theaters. P uppeteers kept a record o f w hat they were paid w hen they perform ed in the kadozuke context. I recall being show n such a d o cu m en t in 1984 w hen I was invited to the hom e o f an elderly m an w hose father had been a D okum bo-m aw ashi. T h e d o cu m en t re ­ corded earnings o f m oney, rice, lodgings, and sake received du ring several N ew Year periods.

Recollections of Prew ar Performances D uring my studies on Awaji, I interview ed m any people w ho rem em bered the seasonal visits o f D okum bo-m aw ashi from th e period before th e war.5 M any o f th em w ere children at the tim e, and their recollections were often hazy. B ut there was a clear p a tte rn in their responses, for children tended to notice th e same things a b o u t the perform ances. Frequently, answers to

questions w ould develop over several m eetings, and people w ould even phone back to say they had rem em bered som ething else they had failed to tell me the first tim e. Interview s, from the m ost form al to the passing conversation, consisted o f a series o f questions. I began each inquiry by asking people w hat, if anything, they rem em bered ab o u t the visits o f p u p ­ peteers to their hom es. Was the perform er som eone the family knew in some way? W here did they recall the perform ance taking place? W hat did family m em bers do before, during, and after the perform ance, and how did they relate to the performer? W hat particular features o f the experi­ ence rem ained in the person’s m ind after all these years? A m arked p attern em erged in response to my questions, namely, a sharp distinction betw een the perceived status o f Ebisu and Sanbaso per­ form ers. F or the m ost part, Sanbaso-mawashi were m ore frequent in com ­ m unities n o t directly bordering on the sea, and they were o f course m ore popular at the N ew Year, w eddings, and boat launchings. M ost people connected Sanbaso w ith agriculture, which m any people n o ted was a “clean” occupation, as opposed to fishing. But Sanbaso-m awashi were also regarded w ith greater respect than those puppeteers w ho only p re ­ sented Ebisu rites. I was repeatedly told by people involved in farm ing that an Ebisu perform er w ould never be allowed into the house b u t w ould be kept at the gate (or maybe the kitchen), whereas a Sanbasb puppeteer could be allowed up into the zashiki. T he explanation given for this differ­ ence was th at an Ebisu-mawashi was o f lower social status than a Sanbasomawashi.6 Two people com m ented th at they considered the Sanbasbmawashi to be “cleaner” and better dressed, an attitude th at probably reflects the ambivalence about Ebisu at some level.7 O thers told me families engaged in business had m ore respect for an Ebisu-m awashi. M ost people were n o t sure why they still rem em bered this distinction, b u t it came up often enough to constitute a p attern in the interviews. M any rem em bered Ebisu perform ances taking place in the kitchen. In m ost old farm houses, the kitchen was a part o f the house where one did n o t have to rem ove o n e ’s shoes, since it had a dirt floor. Perform ing there indicated th at the perform er was n o t being allowed fur­ th er into the house. T he reader will recall the draw ing from a scroll in ­ cluded in chapter 3, showing a puppeteer standing on the d irt inside a doorway, beneath the householders sitting inside. M ost people rem em ­ bered that Ebisu always caught a fish and th at the perform ance had to do with good luck for the com ing fishing season. M any people rem em bered the m oving eyes o f the Sanbaso p u ppet, w hich has symbolic significance in the rite b u t also appeals to children and has great entertainm ent value. W hen people w ould hear the Sanbasb-mawashi was com ing near their neig hborhood, they w ould return hom e if they were o u t, prepare for the

perform ance, p u t on nice clothes, and wait for the perform er to arrive. Responses to w h ether or n o t the perform ers were know n depended on w here the person answ ering the questions lived. M ost people said they were n o t aware th a t their family knew the perform er. I f they were were children at the tim e, they rem em bered being called in by the parents and grandparents to w atch the perform ances, and m any o f them rem em ­ bered seeing an exchange o f food, m oney, o r rice for the perform ances. Q uite a few people told m e th at n eig h b o rh o o d children w ould som etim es follow the puppeteers from house to house while they m ade their rounds. M any people said th at the puppeteers often gave the children sweets and trinkets. This appeal to children is a feature th a t com es across clearly in the early draw ings o f the itinerant puppeteers presented in chapter 3. W hile these perform ances were n o t in ten d ed as en tertain m en t for chil­ dren, a p u ppeteer knew th a t if his perform ance was appealing to children, his popularity at a hom e was likely to increase. B oth o f the ritual p e rfo rm ­ ances we shall consider below include elem ents th a t m ake them appealing to people o f all ages. I t was also clear th a t a few people on Awaji had n o t liked the itinerant perform ers. O n n um erous occasions people had vivid recollections o f the prew ar pu p p etry festivals w here jo ru ri had d om inated the day and were pleased to consider these festivals as p art o f th eir Awaji identity. But w hen asked a b o u t their recollections o f Sanbaso-m aw ashi and Ebisu-m awashi, they gave vague and noncom m ittal answers. Pushed further, a n um ber o f people suggested th a t these perform ers w ere beggars and should n o t be considered p art o f the Awaji tradition. I was occasionally directed away from this line o f research w ith the adm onition th a t a study o f the religious aspects o f Awaji pu p p etry was either o ff th e m ark, u n im p o rtan t, or w o rth leaving o u t o f the record altogether. O ne m an I interview ed to ld me it was a real sham e I was only interested in such “ em barrassing” parts o f the tradition. I t will be rem em bered th a t th ro u g h o u t the Tokugaw a period, laws were enacted th a t m ade it im possible for one to escape from outcast status and th at m axim ized the distinction betw een peasants and outcasts, so th at the m ilitary governm ent and local lords could exert m axim um control over society and suppress peasant unrest. This carefully crafted social ide · ology, w hich m ade certain groups o f people defiled outsiders, m eant th at social discrim ination against outcasts, ritual p uppeteers included, was n o t only ram pant b u t also legally enforced. D iscrim ination was seen n o t as an evil to be overcom e b u t as a system the violation o f which w ould upset the o rd er o f society. M any m ajor studies o f Japanese pu p p etry have neglected a study o fitin e ra n t puppeteers altogether, and th a t this attitu d e clouds the recollections o f th e trad itio n even by people on Awaji.

Playing w ith the Color o f the M orning Sun: Sanbaso Ritual Puppeteers w ho perform er Sanbaso ritual were called Sanbaso-m awashi, and their rite had a num ber o f different nam es, depending on its context: Η δ η ό Sanbaso (C onsecrated Sanbaso), Shiki-SanbasS (C erem onial Sanbasri), Okina-m ai (D ance o f O kina), Okina-w atari (O kina Crossing O ver), K uro-O kina (Black O kina). Like the Ebisu rite, this one has also undergone a gradual transition to being perform ed on the stage, w ith a very established set o f rules for its presentation. B ut it is also possible to preserve the basic structure while abbreviating it considerably, as a com ­ parative exam ination o f Sanbaso rites clearly indicates. T he Sanbas5 rite is com m on to all Japanese perform ing arts. A lthough its origins are debatable,8 a num ber o f things are clear: It was originally intended for divine audiences, and is understood to be a ritual w ith the magical pow er to bring ab o u t a spiritual revitalization. T he text is o f u n ­ know n authorship, b u t it probably reflects a ritual chant used in the past. It closely resembles the descriptions o f Senzu m anzai, adding felicitous words and incorporating Saibara into the chant. Clearly the tw o are re ­ lated. T he basic perform ance we see here appears to have been adopted into N oh, w here the ritual chant was w ritten dow n and the choreography and iconography o f the rite were largely determ ined. A Jthough the N oh versions o f Sanbasb and those found in Kabuki, Bunraku, and ritual p u p ­ petry all differ, they share significant aspects o f the perform ance such as chants, costum es, and music. W hat is interesting about the puppet case, however, is how the itinerant context o f this perform ance kept the magical intentionalities o f the ritual alive. These ritual perform ances have the over­ riding them e o f the creation o f the cosmos and the establishm ent o f the cosmic order. AU the characters in the perform ance— Okina (an old m an), Senzai (a youth o f one thousand years), and Sanbaso (a th ird old m an)— are visiting deities from across the sea ( marcbito) w ho bring special bless­ ings (and, potentially, curses) in their seasonal visits.9 In N o h , the Sanbaso dance is presented as the dance o f O kina.10 T he puppetry rite uses the same text as th at used in N o h , w ith some variation. Because large parts o f the SanbasS rite were forgotten in the last fifty years, in the process o f reconstructing it, gaps have been filled in by using the N oh piece and standard yokyoku (N oh text), m aking it appear alm ost e n ­ tirely derivative o f the N oh Okina dance.11 M ost likely, the puppetry ver­ sion was originally distinct, b u t it also was m ore vulnerable to the vicissi­ tudes o f tim e because it was presented at a popular level. Surviving Sanbasb perform ances from the Awaji lineage in N agano Prefecture show a perform ance structure which is quite long, has little or no chanting, and

focuses on dance alm ost exclusively, w ith m any repetitive gestures. It is quite clear th a t this perform ance stru ctu re, akin to Kagura, was clearly n o t m eant to entertain a hum an audience, since it can be quite boring (lasting for hours). I t was presented, like K agura, at festivals in shrines in fro n t o f the w orship hall to entertain th e deities while hum an beings w ent ab o u t their business o f eating, gossiping, and playing gam es, occasionally glanc­ ing over at the perform ance. W atching a fou r-h o u r Sanbaso perform ance w ith rapt a tten tio n is behavior th a t could be expected only from a super­ natural being.

Occasions fo r the Sanbaso R ite O n Awaji, the Sanbaso rite was used for a n u m b er o f dram atically different purposes. I review them here according to the typology m ost people on Awaji intuitively use w hen they classify perform ance occasions.12 W ithin this typology, there w ere basically tw o styles o f perform ance. O ne, Shiki Sanbaso, consisted o f m usic, dance and p o etry and was b o th interesting to watch and relatively sh o rt (lasting from five to fifteen m inutes, m aking it suitable for a kadozuke perform ance), and appealing to hum an audiences. T he o th er, H o n o Sanbaso, was long and consisted alm ost exclusively o f dance and m ovem ent; it was used in shrines as a form o f Kagura w ith puppets. It does n o t survive on Awaji b u t does in N agano Prefecture at Waseda Jinja near Iida City and resem bles the Sanbaso dance found in Kagura. T he perform ances discussed below are Shiki Sanbaso rites, unless otherw ise indicated. Kotobuki Sanbaso. M ost com m only, Shiki Sanbasd was perform ed at the beginning o f som ething: w hen the m ain pillar o f a house w ent up (:mune-age no sai), at a w edding party, w hen a boat was first p u t in to the w ater ( um i-ire), or on the m o rning o f the th ird day o f the N ew Year. AU o f these events have tw o things in com m on: they are ritually pure m o ­ m ents, and they are characterized by kotobuki (felicity), symbols o f w hich are painted, sewn, and carved in to the Sanbaso pu p p et o r sung in the chant. Because everything in this rite points tow ard to newness, vitality, and longevity, the ritual invocation o f these powers in the cosmos served as a felicitous blessing o f the event. Hono Sanbaso. Situations w here th e audience was clearly and exclu­ sively divine required th e H o n o Sanbasd (hono m eaning consecration or worship). F or exam ple, rites were perform ed at w ater sources in the m ountains, in rice seedling nurseries, and before shrines at N ew Year. U nlike th e Shiki-Sanbaso rite ’s lim ited use o f the flute and drum followed by chanting o f p o etry and dancing, in this rite the flute and drum are played th ro u g h o u t the perform ance to attract and soothe th e kami, and

U m azum e M asaru holds a small Sanbaso puppet, o f the variety used by itinerant perform ers in the eighteenth and nineteenth centuries.

the repeated dance m ovem ents o f the puppet purify the four directions using the sleeves. A lthough a hum an audience was som etim es present at these rites, they were n o t the fundam ental recipients o f the perform ance. K am i-okuri Sanbaso. T he Sanbasd rite was also used in the process o f purification o f a hom e or village from evil, pollution, and noxious forces. This appears to be p a rt o f a w ider context o f ritual puppetry called by the general name kam i-okuri ningyo (sending the kami puppet) o r tnmayori ningyo (spirit possession o f a puppet). In this rite, worshippers w ould first gather and watch a puppet perform ance o f Sanbasd and then proceed th ro u g h the streets o f the com m unity to a fixed point outside the b o u n d ­ aries o f the village carrying the pup p et on a palanquin. T he pu p p et was understood to be a spirit vessel, and epidem ic and o th er spirits on the

PUPPETS OF THE ROAD AND FIELD

173

This photograph, taken before the war, shows a Sanbaso performance at a small worship site in the middle of a rice field. Courtesy ofFud6 Satoshi.

road, attracted by the performance, would possess the puppet as it passed through the community, and so be carried away. 13 Particularly important was the attraction, capture, and expulsion ofthe spirits of insect pests, which could ruin an entire harvest. This rite has not been performed on Awaji since before the war. I was curious about the efficacy of this rite and how people had understood it. Was this all people did to prevent the infestation of insects? "Of course not!" was always the reply. "People weren't stupid!" I was told that farmers hand picked insects off crops, used rotational planting, and smoked fields to rid them of insects. The purpose of the puppetry rite was to address the spiritual aspect of insects. When expelling them from a village, it was necessary to consider their feelings, lest they come back the next year and

174

CHAPTER 5

This puppet from the Waseda shrine in Nagano was used in annual rituals to purify the village. The puppet was understood to attract noxious forces and the spirits of harmful insects and carry them beyond the village boundaries.

create even more problems. This required the ritual use of puppets. Puppets were beings for dealing with problems in the spiritual, not the material, realm . We do not know what this kind of rite looked like on Awaji, but we can study extant examples of the ritual use of Sanbaso at other centers to which Awaji puppetry spread. In the Shinshu region (Nagano Prefecture), where an Awaji puppeteer established the practice of ritual puppetry in the early nineteenth century, the practice of the Sanbaso puppet being used as a tamayori ningyi5 has recently been revived, based on descriptions of the practice from the early 1950s, when it died out. The ritual begins at the local shrine, called Waseda Jinja, where the puppet is consecrated before the worship hall. This performance uses the Hono (consecratory) Sanbaso, as opposed to the Shiki (ceremonial) Sanbaso. The sweeping move-

m ent o f sleeves in purification is perform ed repeatedly; so too is the chant “ Oosai ya Oosai ya, yorokobi ariya” as the puppet purifies the four direc­ tions. After the dance, the puppet is carried out o f the shrine, through the streets o f the small tow n, and o u t to a designated spot at the edge o f town, where another rite is conducted to divest the puppet o f the pollution and spiritual forces it has absorbed. The puppet is said to attract spirits o f insects and also epidemics.14 Sanbaso R ite as R itu a l Purification o f the Home. As discussed in previ­ ous chapters, the Sanbaso rite was used to ritually purify a home or com ­ munity. This process was accomplished by two interrelated principles. One was the understanding that through the power o f the performance, any dangerous spirits or noxious forces in the homes would be attracted to the puppet and possess it, following the same logic as the tamayori ningyo discussed above. But there was also the understanding that through the ritual o f revitalization expressed in the Shiki-Sanbaso chant— and actualized as the body o f the puppet literally “comes alive” when it is picked up and moved— the dom inant form o f ritual pollution, namely, the degeneration o f the year, could be driven away. Vitality replaced exhaus­ tion, and purity replaced pollution. The rite affirmed the reality o f the new, the vital, the pure, and the felicitous.

A m agoi Sanbaso (O kina w a ta ri) Sanbaso puppets were also used in rites requesting rain, called amagoi (petitioning for rain). Lack o f rain was and still is a serious concern in rural Japan. Rice depends on an abundant water supply, and major famines have resulted from droughts like that o f 1732, when one third o f the peasant population o f Japan is said to have died.15 The rite performed in the amagoi context is the Shiki-Sanbaso, but it is frequently referred to as O kina-w atari. Lines taken from the chant— “The sound o f the waterfall, the sound o f the waterfall, even if the sun is shin­ ing, will not cease”— refer to the blessing o f abundant w ater even if there are many sunny days. The rite is th o u g h t to be efficacious in petitioning for rain. Puppetry appeases angry deities who are withholding their benef­ icence from the hum an com munity out o f revenge and chastisement. Amagoi rites were n o t uncom m on on Awaji. Frequently, the Sanbasb performance would be followed by a longer festival in which other joruri pieces were presented, a pattern o f ritual activity we consider the end o f this chapter. In Kyoho 9 (1724), dry w eather threatened to destroy the planted crops. The Uem ura Gen no Jo troupe perform ed a rain rite using Sanbaso puppets in the Hachim an shrine in Sum oto, followed by the fiveact joruri piece “Sugawara D enju Tenarai Kagami.” As a result, it rained

176

CHAPTER 5

nonstop for twelve days. This is m entioned in shrine records at the Sum oto H achim an shrine. Seventy years later, during another m ajor d ro u g h t in Kansei 6 (1794), puppeteers from this same troupe perform ed a rain-requesting rite, b u t it did n o t rain as a result. They then decided to follow the exact pattern they had followed at the H achim an shrine in S um oto seventy years before. They again perform ed the am agoi rite and th en the five-act joruri piece, the same one presented before. Rain soon fell, and it continued for ten days. I t is said th at the O kina mask used in b o th the Kyoho and Kansei perform ances was magical. A black O kina mask, it reflected the color o f clouds full o f rain. D uring a d ro u g h t in early Showa (the late 1920s), U em ura troupe p u p ­ peteers again perform ed an amagoi rite in the Hiyoshi Shrine in M atsubara in Seidancho. Follow ing the presentation o f Okina-w atari, a dragon appeared on stage, the symbol o f a storm y sky, and a long prayer for rain was recited. T he next p art o f th e perform ance was a presentation o f joruri pieces, and as the tayu was finishing chanting, it suddenly started to pour. T he puppets, their costum es, and all the perform ers and audience got entirely soaked. I t was rep o rted th a t the kami from this shrine really hated puppets, and so he soaked them before the puppeteers had a chance to p u t them away.16

Sbiki Sanbasd T he basic ritual used in these perform ances consisted o f a set use o f m usi­ cal instrum ents, chant, and puppets. W hat follows is an analysis and an n o ­ tated translation o f the Shiki Sanbaso rite. I have relied on a series o f sources: old films o f the last perform ances o f this rite in the kadozuke context, photographs, firsthand recollections from people on Awaji w ho both saw and perform ed the rites, the perform ances at the theater today, and interviews w ith U m azum e M asaru, w ho in the late 1940s was tau g h t the perform ance by elderly puppeteers in Sanjo.

K itu al Symbolism T hree puppets are used in the rite: O kina , Senzai, and Sanbasd. At the beginning o f the puppetry rite, before the puppets are picked up, a flute and d rum are played. This attracts the spirit— in m ost cases u n d erstood to be the deity o f the rice fields— and invites it to enter the puppet. Each puppet, preparing to be possessed, in tu rn covers its face w ith th e sleeves o f its elaborate costum e, indicating its subm ission to the greater spirit which has been sum m oned.

T he initial m ovem ents o f the first p u p p e t are in ten d ed to purify the su rro u n d in g space p rio r to the possession, w hich will take place w hen the deity enters the p u p p e t’s body. First, the arms o f the p u p p et m ove in the fo u r directions, creating a dram atic effect as th e long sleeves o f the cos­ tu m e snap w ith each gesture. T h e p u p p et shakes a rattle, to furth er purify th e space and bless the audience. This gesture is also understood as a ta m a fu ri (spirit shaking), w hich invigorates the perform ance (although m any farm ers w ith w hom I spoke said it was an im itation o f sow ing seed). O ne p u p p e t, Sanbaso, slowly begins to dance and eventually goes in to a frenzied and ecstatic tran ce,17 m ade apparent by a change in the shape o f his eyes: they go from norm al hum an eyes to ro u n d eyes th a t look in to the o th e r w orld— eyes o f fear, pain, w onderm ent, intense rage, frenzy. As the p reparation for the possession goes o n , the m ovem ents o f the p u p p e t be­ com e erratic: he stam ps his feet, rolls his eyes, and waves his arms. This foot stam ping (a-shibyoshi) is also u n d ersto o d as an act o f driving o u t evil and su m m oning up pow ers from the earth. AU this is to b u t a preparatory stage to th e pow erful transform ation a b o u t to take place, the actual p o s­ session. In this way, the p u p p e t becom es a body substitute on yet a nother level: he takes th e place o f th e sham an, w ho also may un d erg o an ecstatic transform ation in preparation for possession. I t is u n d ersto o d in this rite th a t the deities sum m oned are to o pow erful to be contained w ithin the body o f a hum an m ediator, so the p u p p et stands as a bridge betw een the hum an and divine com m unities. T he actual possession in full form is m ade apparent by p u ttin g masks on th e puppets (or som etim es just on the p u p p et Sanbaso) in the m iddle o f the rite. I f we consider th a t puppets are once rem oved from the realm o f h u m an perform ers, we have now a second level o f rem oval from th e o rd i­ nary, profane w orld o f m eaning w hen a p u p p et has a mask on its face. Even th o u g h puppets are substitutes for actual hum an beings in th e rites, they to o m ust be m ust be symbolically insulated from the sacred. This is accom plished by w earing a mask, after w hich, the p u p p e t’s m ovem ents becom e m ore controlled, although m ore forceful.18 A nother level o f pow er has settled into th e p u ppet, and th e deity is now in full control. T he purpose o f the p u p p e t’s possession— w hich originates in a tra d i­ tio n w herein a dead person returns in the form o f a p u p p et, D 5 kum bo, and w hich continually harks back to this o rigin— is to bring fecundity and new life to th e people in the audience, their hom es, th eir rice seedlings, and th e com m unity in general. Just as the pu p p et m ade o f lifeless m atter is b ro u g h t to life during th e first stage, w hen it is m ade to move and th en is, so to speak, ontologically upgraded to a being possessed by the divine, so to o is the com m unity’s vital energy restored and stren g th en ed . These tw o aspects o f th e p u p p e t’s m ovem ent suggest vitality: th e initial m ove­ m en t o u t o f inertia and the secondary possession by a divine force.

This prew ar p h o to g rap h o f a Sanbaso perform ance shows the p u ppet w earing a mask. C ourtesy o f F udo S atoshi.

The Pupfet Costume F u rth er evidence th at the pup p et em bodies a hope for vitalization— or revitalization, as the case may be— can be seen m ost clearly by exam ining the p u p p e t’s costum e. The costum es consist o f a series o f superim posed images referring to longevity and eternity. The m ost identifying aspect o f Sanbaso, so m uch so th at it has become his tradem ark, is the hat. O n it, we see thirteen stripes o f dark and bright, indicating the waxing and waning m oons, images o f both the passage o f tim e and a cyclical renew al.19 O n one side o f the hat is a full m oon, and on the other, a crescent m oon, further re­ inforcing the tem poral, cyclical, lunar aspect o f the hat. At the base o f the

179

PUPPETS OF THE ROAD AND FIELD

Sanbaso kashira

hat, the stitching connecting the hat to the kashira is a series of six-sided shapes, considered to represent the shell of a turtle (symbol of ten thousand years). (This six-sided pattern also appears on the costume of Senzai.) Next, there is the clothing worn by the puppet. The turtle appears again on the back of the small kimono, as does the crane, which signifies a thousand-year longevity. Sanbaso's two companions, Okina and Senzai, both embody the dimension of age , since Okina is old and Senzai (a mere thousand years of age) is youth. These images are often beautifully drawn on the costumes or sometimes crudely embroidered. As the puppets perform and move their sleeves about, these visions of temporality flash in and out of view. One sees a turtle, then a crane, and now and again a waxing or waning moon.

180

CHAPTER 5

This rare puppet, discovered in Morioka in 1987, shows the symbolism on the costumes of puppets in the Sanbaso performance.

The Chant Perhaps the most problematic aspect of this rite is the ritual chant (tonaegoto or jiuta) used in it. It seems to defy any logical translation, and at times appears to be a medley of symbols and a salad of words. The problem with the text, however, lies perhaps in our expectation that it should behave like a more conventional text-have a narrative structure, be translatable, etc. On one level, the text can be seen as the "speaking parts" for the human officiant (the puppeteer as shaman) and the deities who enter the puppets during the rite . The use of poetic imagery, literary references, and ritual utterances of untranslatable meaning suggest, however, that another level of language is at work here. Let us examine the

special n a tu re o f the language in this rite before tu rn in g to a translation o f the ch an t itself. As M aurice B loch has correctly n o te d co n cern in g th e unique n a tu re o f language in ritual co n tex ts, “ Ritual is . . . a place w here, because th e o rd i­ n ary form s o f linguistic co m m u n icatio n are changed, we c an n o t assum e the sem antic processes o f m ore o rd in ary c o m m u n ic a tio n .” 20 Ritual lan ­ guage, and ritual p o e try in particular, is u n d e rsto o d to have th e m agical capacity to create th e realities it describes. T his u n d e rsta n d in g is n o t an u n c o m m o n feature o f ritual language in th e h istory o f Japanese religions. T w o celebrated cases from the h isto ry o f Japanese religions elo q uently underscore this point: An early reference to th e pow er o f language to create realities can be found in th e ICojiki. In the m yth, th e god Susano, b ro th e r o f th e sun goddess A m aterasu, declares he has w on a m atch for pow er w ith his sister, and he “ rages w ith v icto ry ” by destroying the o rd er th a t his sister had created in th e form o f dikes betw een rice paddies, an d defecates in “the hall w here first fruits are ta s te d ,” th e n spreading his excrem ent a b o u t the hall. H is sister’s response is incom prehensible, unless we consider the pow er o f language. She says, “T h a t w hich appears to be feces m ust be w hat my b ro th e r has v om ited and strew n a b o u t w hile d ru n k .21 A lso, his breaking dow n th e ridges o f th e paddies and covering up th e ditches— my b ro th e r m ust have d o n e this because he th o u g h t it was w asteful to use the land th u s.” T h e text continues: iiEven though she spoke thus w ith good intentions, his misdeeds d id not cease, b u t becam e even m o re flag ran t.” 22 As Philippi no tes co n cern in g this passage, her a tte m p t to “ speak g o o d w ords correctively” {nori-naoshi) m o st p robably reflects th e belief th a t “ one could tu rn evil in to g o o d by speaking well o f it.” 23 Second, the Japanese preface by Ki no Tsurayuki to th e early-tenthcen tu ry te x t K okinshu (or K okinw ctkushu) directly refers to this capacity o f th e Japanese language, properly used: The seeds o f Japanese poetry lie in the human heart and grow into leaves o f ten thousand words. Many things happen to people o f this world, and all that they think and feel is given expression in descriptions o f things they see and hear. W hen we hear the warbling o f the m ountain thrush in the blossom s or the voice o f the frog in the water, we know every living thing has its song. It is poetry which, w ithout effort, m oves heaven and earth, stirs the feelings o f the invisible gods and spirits, sm oothes relations with men and w om en, and calms the hearts o f the fierce warriors. Such songs came into being when heaven and earth first appeared.24

A fu rth e r reflection o f an ancient Japanese conviction o f th e m agical efficacy o f ritual language, absorbed in to m any later Japanese ritual c o n ­ texts, can be seen in w hat is referred to as kotodam a shinko, o r th e belief

(shinko) in the spirit (tama) o f words (koto). According to Konishi Jin’ichi, this archaic and ancient view o f the power o f ritualized speech—phrases uttered in the proper context and with the proper tone and pronuncia­ tion—survived in Japan well into the Middle Ages, at the time the rite under discussion here was being formed. Konishi specifically notes a fea­ ture we will see shortly in the Sanbaso chant. He writes that “priests com ­ posed [poems] in the belief that speaking a great many longevity-related phrases would propel the kotodama in their utterances toward their sover­ eign and grant him long life. . . . The priests were aware that their [poem], a virtual list o f auspicious things, was recited in anticipation that the koto­ dama would function.”25 Finally, we can not overlook how the power o f language can express religious and political power relationships. As Gary Ebersole observes in his book R itu a l Poetry and the Politics o f Death in Early Japan “Song was frequently used and experienced as a form o f the exercise of power. It was a linguistic means o f manipulating religio-political power in the human sphere as well as o f manipulating the spiritual powers, in­ cluding the kam i and the spirits o f the dead. Song was sung and poetry recited not only for aesthetic pleasure but as a means o f ordering and controlling potentially dangerous aspects o f the world. This sense of the efficacy o f poetic language survived until much later in Japanese history and was prom inent in the Heian and medieval periods. Indeed, it is still found in the present in attenuated form in certain rural areas and ritual practices.”26 The text for the Sanbaso rite, which I translate below, has often been neglected by translators and commentators precisely because it relies so heavily on notions o f language extremely foreign to most people living in Japan and the West today.27 Part o f the difficulty in understanding this text, it seems, comes from the expectations modern readers have brought to it as text alone, divorced from its ritual context. When seen as an ex­ pression o f ritual speech, the incomprehensible opening lines, lack o f nar­ rative, and layering o f symbol upon symbol and literary reference upon literary reference, as we shall note below, all make sense. The realities of vitality, longevity, and harmony are invoked through the use o f potent language in a ritual setting. The following performance description is based upon that o f the Awaji puppet theater today. I have carefully compared it to a film made in the late 1940s o f two elderly Sanbaso-mawashi from Awaji, and the text and movements o f the puppets are nearly identical. The slight variations in ritual action between the present-day Awaji puppeteers and the elderly men from the 1940s need not concern us. After translating the per­ formance text, I offer a discussion o f the motifs and ritual speech ele­ ments in it.28

SANBASO

The performers begin by creating a sacred space fo r the performance. They cut white strips o f paper (gohei) to indicate the site is amarked off”from the profane. Next, they offer white uncooked rice and sacred sake to the deities invited. While one o f the performers plays a small flu te to summon sacred forces, a second performer beats a small drum and recites. P a r t O n e: I n v o c a tio n o f D e ity

C

horus

O

k in a

C

horus

O

k in a

C

horus

O

k in a

Ch

:

: :

: :

:

oru s

O

k in a

C

horus

tarari tarari ra Tarari agari rarari to Chiriya tarari tarari ra Tarari agari rarari to} 9 Live a long, long time We will also live a thousand autum ns— The ages o f the crane and the turtle30 Let u s enjoy good fortune in our hearts To to tarari tarari ra Chiriya tarari tarari ra Tarari agari rarari to T he sound o f the waterfall The sound o f the waterfall31 Even if the sun is shining Will not cease To ta ri a ri u do do do22

T o td

O k in a :

:

:

:

S e n z a i:

I t w ill n o t c e a s e

To tari ari It will continue as usual [ Senzai performs a slow dance, waving his sleevesfro m side to side to purify the space] S e n z a i:

C

horus

M a y y o u w ill liv e a t h o u s a n d y e a r s

:

A heavenly m aiden’s robe o f feathers Even if the sun is shining The sound o f the waterfall will not cease It will n o t cease Totari ari uto to to

[ Senzai is p u t down, and Okina is picked up] O

k in a

C

horus

: :

Agem akiya ton do ya Hiro bakari ya ton do ya

[ These two previous lines are fro m a Saibara and refer to a sexual moment between two lovers. D uring the following chanting, once again shared by the two performers, the drum m ing continues, and the Okina puppet bows,

spreads its arms, a nd then makes points sharply into the sky an d flips both sleeves behind its head in dance movement.] O k in a :

A lthough w e’re seated L e t’s begin! O k in a : We have been celebrating since a long tim e ago Since th e age o f the gods C h o r u s : Soyo yari chiya ton do ya O k in a : T he thousand-year crane Sings the M anzairaku song o f longevity And the tu rtle w ho has lived in th e p o n d for ten thousand generations Carries on his shell the three songs33 T he sand rustles and spreads o u t on the seashore A nd plays w ith th e color o f the m orning sun T he w ater cascades cool and pure Clearly floats the evening m oon Peace u nder heaven! Tranquillity' th ro u g h o u t the land This is today’s prayer

C

horus:

[ The invocation ends. O kina pauses briefly, as i f to listen to these following lines.]

h orus:

W ho are those old m en?34 W ho are they? W here are they from? Since this is the dance o f a thousand autum ns and ten thousand ages, let’s do one dance o f longevity, the M anzairaku.35 Manzairaku!

O

k in a :

M a n z a ir a k u !

C

horus:

Manzairaku!

C

horus:

O

k in a :

C

Part Two: T h e Trance D ance o f Sanbaso The puppeteer puts O kina down to his left a n d picks up Sanbasd. This p u p ­ pet has feet, as opposed to the other two, w ith only trousers. Sanbaso is hold­ ing a rattle made o f sm all bells with streamers attached. I t is inserted into a small hole in the puppet’s right hand. Shaking the rattle serves to purify the space, a n d the rattle itself serves as a to rim o n o, an object into which a sacred force will descend. When the puppet’s arm moves, the rattle sounds. He does a b rief dance to d ru m accompaniment, which includes the stamping o f feet— to drive out evil— a nd rapid movements o f the eyes— to in d i­ cate he is undergoing a transform ation into sacred status. Part Three: T h e A poth eosis, Sanbasd as Sacred Presence When Sanbasd’s dance isfinished, the puppeteer p u ts the doll down an d puts a black mask over his face. The drum m er puts aside his drum , puts a white

mask on O kina, a n d both puppets are m a n ip u la te d w ith o u t accom pani­ m e n t.36 N ote th a t in the fo llo w in g discussion, O kina serves as the “straight m a n ”fo r Sanbaso. In the Noh versions, this p a r t is played by Senzai or a m in o r actor (o m o te j. The conversation is carried o u t w ith both puppets being held w ith their faces covered by their sleeves. S a n b a s Ce O

k in a

:

Sanbaso: O

k in a

:

Sanbaso:

O

k in a

:

Sa n b a s o : O

k in a

:

Sanbaso: O

k in a

:

Sa n b a so : O

k in a

:

Sa n b a so :

Oh! Such joy, such joy, such joy! I w o n ’t let it slip o u t o f here! H o w felicitous! I will talk to th e lead dancer w ho dances in the later part. Aha! T h a t’s me! You are to dance th e felicitous th o u san d au tu m n s an d ten th o u san d ages for to d a y ’s blessing. You, th e black O kina. F o r to d a y ’s blessing, this black O kina is to felicitously dance the S enzu m anzai. T h ere is n o th in g I w ould like m ore to do! B ut first, sir, y o u m u st re tu rn to y o u r seat. I will n o t re tu rn to my seat. First you dance, sir, an d th e n I will re tu rn to m y seat. First, y ou re tu rn to y o u r seat. First, I am g o in g to enjoy th e dan ce, and th e n I will retu rn to my seat. First, you re tu rn to y o u r seat! No! First y ou dance! No! R etu rn to y o u r seat! H o w felicitous. B egin w ith th e bell! W h at a g reat deal o f tro u b le this is!

The conversation is thus concluded, a n d O kina is p u t down. The perform er who m a n ip u la te d O kina now picks up his d r u m fo r the masked dance o f Sanbaso, called in the Noh versions, the “bell scene. ” Both the mask a n d bell indicate th a t Sanbaso is now fu lly possessed by sacred forces. P a r t F o u r: T h e W o rd s o f B le ssin g Sanbaso dances frenetically to the follow ing chant. H is rattle cuts through space a n d shakes wildly fr o m side to side. H e leaps about in the air. The energy intensifies as the perform ers chant the text, beat the d ru m , a n d voice cadence calls. C

h o r u s

:

O h how grateful we are for this m anifestation o f th e deity, H o w blissful to give thanks to th e deity w ith this k a m i asobifi In d e ed , there are so m any d an cin g girls A t S um inoe we will h ear th e ir clear, lovely voices. T h ere th e blue waves are said to reflect th e shadow o f the waves. You and th e ro ad o f th e kami sh o u ld lead straig h t to th e lively capital in springtim e.

This is the dance o f Genjorakujii And for Manzairaku, you use Omi robes. W ith a striking arm, we purify you of all evil, and in our hands we gather longevity and good luck. The dance o f a thousand autumns caresses the people and lengthens their lives. Together, the pine trees and the wind enjoy making their voices whisper.39 We depend on the wind in the pines. As the chant concludes, the puppet ceases his dance, shakes his rattle from side to side as i f to bless the space, and then is placed in the sleeping position in the carrying box. The performers clap their hands together and bow and the end of the rite. T his use o f hig hly p o etic and felicitous language p re se n te d in a ritual c o n te x t unleashes th e p o w e r o f th e w o rd s them selves to create th e realities th ey so beautifully describe. H e re , le t us d irec t o u r a tte n tio n to th re e uses o f lan g u ag e in this ritual tex t: th e m agical use o f lan g u ag e as so u n d , the use o f lan g u ag e to invoke nostalgia an d tra d itio n , a n d th e p h e n o m e n o ­ logical use o f lan g uage to create im aginary a n d visual experience. T h e first stan za o f th e tex t opens and closes w ith th e u n tran slata b le syllables to to ta r a r i ta r a r i an d closes w ith a sim ilar chiriya ta r a r i ta ra r i ra. W hile these syllables have n o readily available m ean in g , th ey serve a n u m b e r o f p u rp o ses in th e text: F irst, they sensitize th e listener to th e o n o m a to p o e tic qualities th a t o ccu r later in th e te x t, an d hence to th e m agical use o f lan guage. S eco n d , th e y set th e stage for som e audial p uns in th e te x t, w h en th e sam e verb e n d in g rctri an d th e syllables do do do com e to have m ean in g . T h is allows th e listener to m ove b etw een tw o levels o f lan g u ag e, th e level o f m ean in g an d th a t o f so u n d . We see a su b tle shift from th e use o f lan g u ag e as so u n d to th e in c o rp o ra tio n o f these syllables in to m eaningful phrases. B u t th e n w e see th em used o n ce again o n o m a to poetically. By prefacing th e en tire piece w ith this s o rt o f ritu al u tte ra n c e , th e re fo re , th e p artic ip a n t o r audience is im m ediately fo rced to o p e n to various levels o f c o m p re h e n d in g ritu al so u n d . T h is ritu al use o f language I call a “ m agical use o f language as s o u n d .”40 T h e tex t th e n plays w ith this o p en n ess o f th e p articip an ts in a t least tw o fu rth e r ways. T h e n e x t elem e n t in th e tex t to w hich I w o u ld d raw o u r a tte n tio n is w h at I have chosen to call a “ m edley effec t.” A m edley, a s o n g o r piece w hich b o rro w s an d places in seem ingly ra n d o m o rd e r readily reco g n izab le bits and pieces o f p o p u la r songs o f th e day, is capable o f quickly creatin g a m o o d o r am biance. I t becom es unnecessary to re in v e n t th e w heel for each n ew ritual event. O n e has only to th in k o f th e in te n tio n s b e h in d and m o o d c reated by a b rie f m edley o f “T h e S tar S pangled B a n n e r,” “ A m erica

th e B eautiful,” “Y ou’re a G rand O ld F lag ,” and “ M y C o u n try ’Tis o f T h e e ” (w ith m aybe B ruce S prin g steen ’s “ B orn in th e U .S .A .” at the end to draw in a y o u n g e r crow d), o r o f th e freq u en t use o f m edleys in M e m o ­ rial D ay parades and F o u rth o f July celebrations. In d e e d , w hen a m edley is effectively used, it is often preferable to new pieces o f m usic w ith the sam e th em e, as new pieces lack the p o w er o f nostalgia and the au thority o f tradition. A n o th e r im p o rta n t aspect o f th e m edley effect is its e c o n ­ omy. T h e briefer th e reference to a n o th e r w ork, the better. A g o o d m edley uses only e n o u g h o f a piece to cap tu re th e m o o d before m oving on to another. T his Sanbaso ch an t em ploys such an apparatus to condense a n u m b er o f m eanings into one ritual m o m en t. A series o f fam ous one-liners from classical Japanese poetry, nam es o f o th e r perform ances (from K agura and G agaku, tw o o th e r p erfo rm in g acts w ith stro n g ritual origins), references to m yths, legends, and felicitous sym bols are all th ro w n in to th e text. F o r exam ple, “ th e so u n d o f th e w aterfall” is a line from th e R yojin hishd, and A gem aki and H iro bakari refer to K agura perform ances w ith sim ilar m otifs o f longevity, peace, a n d felicity. “T h e th re e so n g s” o n th e tu rtle ’s back are felicitous songs from th e N o h theater. T h e sim ple line “A heavenly m aid e n ’s ro b e o f feathers” calls fo rth an entire m yth and a p ro fo u n d n o ­ tio n o f eternity, the Vedic idea o f a kalpa as a m easurem ent o f tim e. A m aiden descends to earth once every one h u n d re d years and brushes a rock (forty Ii in b read th ) w ith h er ro b e o f feathers. As lo n g as it takes to w ear d o w n this rock is one kalpa.41 T h e result o f this “ m edley effect” is a high degree o f ritual co n d en satio n and efficiency. T h e ritual is literally packed w ith each o f these references. A w ell-educated Listener in m edieval Japan m ig h t have been familiar w ith a n u m b e r o f these references. Today m o st people a tte n d in g this rite will be largely unaw are o f them . T hey m u st rely on the pow er o f th e im ­ ages and poetic utterances w ith o u t recourse to th eir contextual origins. It is still possible, how ever, for a perso n w itnessing it to have a p ro fo u n d experience o f it at th e level o f language, because th e ritual ch an t includes a th ird use o f language, an appeal to a p h en o m en o lo g y o f the senses. R ef­ erences to w aterfalls, th e sun, tu rtle s, and cranes, and poetic lines like “T he sand rustles and spreads o u t o n th e seashore, and plays w ith the color o f th e m o rn in g su n ,” have an evocative pow er. O n e is able to visual­ ize th e realities described, and th ro u g h visualization in th e ritual, these peaceful and vibrant im ages becom e reality at th e level o f the spirit. W hen reinforced by th e o th e r sim ultaneous events in th e rite— th e m ovem ent and tran sfo rm atio n o f th e p u p p e t, the flute and d ru m m usic, th e designs o n th e costum es, and th e incessant ringing o f tiny bells— th e full m eanings o f these linguistic im ages can be felt. I have o ften interview ed people w ho have com e to w atch the Sanbaso perform ance, and m any often rem ark

th at the lovely visual images, understandable even today, create a soothing and pleasant m ood. If, therefore, we look for a narrative in this text, we will find a thin one: a few old m en show up from the land o f the gods and perform a dance after a bit o f hum orous negotiation. If, on the o ther hand, we direct our atten tio n to the ritual use o f language, we will see th at this text is n o t about narrative b u t about the creation o f m ood, im age, and effect. L an­ guage is in the service o f a greater analogical realization taking place in the rite: death is being tu rn ed into life. Aitd for the m ood to be fully felt, the text cannot be separated from the o th er ritual elem ents—m yths, symbols, tim ing, perceptions o f the perform ers.

An Endless Era o f Felicity: Ebisu-mawashi In chapter 3 we discussed how Ebisu is a liminal deity— am biguous, p o ­ tentially dangerous, and in continual need o f ritual appeasem ent and co n ­ tainm ent. While we do n o t know precisely w hat the N ishinom iya Ebisukaki did, we have a good idea o f w hat Ebisu perform ance on Awaji looked like. Ebisu-m awashi perform ed at private hom es, along fishing docks, and at the gates to small Ebisu w orship sites. Each perform ance space entailed a different idea o f host-perform er relations and gave the p er­ form er a different am ount o f control over the space. Let us look first at w hat is know n about the kadozuke context at private hom es (also som e­ tim es referred to as m ontsukeon Awaji, an alternate reading for the charac­ ter m eaning “gate” ).

Performance Context: Kadozuke Ebisu-mai Ebisu-m awashi traveled on foot to hom es w here they w ould present their short perform ance in exchange for som e sort o f rem uneration. It was co m m o n to see Ebisu perform ers during the N ew Year period, to inaugu­ rate the fishing season, or to bring in good business. O ld photographs o f these perform ers show them holding puppets th a t were about half as tall as they were. Frequently they had carrying bags hung on their backs. T heir heads were uncovered, and often they were wearing dark-colored pants bound up around the calves and straw sandals. Ebisu as well was dressed in the clothes o f an itinerant person, w ith straw sandals on the feet and pants tied up at the calf. T he puppet w ore an eboshi, a type o f hat th a t flopped dow n to one side, and often the puppet had a bell in one hand (tied th ro u g h a hole in the w ooden hand) and in the o th er fre­ quently held a fishing pole from which an artificial fish was hanging, as if

caught on a hook (this was to be a featured part o f the performance). Invariably, the puppet had a simple kashira w ith no internal moving parts, and a painted, smiling face and a small beard.42 The puppet was m anipu­ lated by one person. The puppeteer held the rod on to which the kashira was attached by inserting his left hand th ro u g h a hole in the back o f the puppet’s costum e. The puppeteer m anipulated the right hand o f the p u p ­ p et by inserting his right hand into the sleeve o f the p u p p et’s costum e. H e could then move the arm and sometimes the jointed wrist o f the puppet w ith some grace. Almost all itinerant Ebisu-kaki worked alone. O ne person carried the puppet and recited a simple story, which appears to have varied am ong perform ers. A ccording to Gunji Masakiyo, a standard version included three scenes. In the first, th e puppeteer w ould sing a lam ent describing the plight o f fisherman unable to catch fish. This song may have served as the perform er’s call announcing his arrival. In the next scene, known as “call­ ing the deity” (kam i yob&i), the puppeteer w ould then beseech Ebisu to appear and catch the first fish o f the season, so the year’s catch could be released. In the third scene, the puppet, understood to be the shintai for Ebisu, w ould appear and catch a live sea bream , the king o f fish (whose name, m i, is felicitous, since this syllable occurs in the w ord medetai, “fe­ licity” ). T hereupon th e puppet deity Ebisu w ould perform songs and dances for the spirit o f the fish, coaxing it to release the year’s catch. It was n o t possible to use a live fish if the perform er was presenting his piece from door to door, so this part o f the dramatic effect was reserved for Ebisu perform ances presented at fishing festivals. The piece w ould end w ith the recitation o f felicitous w ords.431 have n o t seen this version o f the perfor­ mance, but a few fishermen on Awaji rem em bered it. It is possible that this three-part version is an older example o f Ebisu ritual and may resemble the performances done by Ebisu-kaki from Nishinomiya. In some versions, Ebisu would ask the householder or people assem­ bled to share a drink as a toast to a good fishing season. In all perform ­ ances, Ebisu w ould “row o u t to sea” and catch a fish, the symbolic first catch o f the year. The actual fish in m ost o f the itinerant pieces was made o f papier-mAche or carved o u t o f light w ood and hidden in the puppeteer’s sleeve until it was tim e for it to appear. This perform ance was usually re­ ferred to as “Ebisu no tai-tsuri” (Ebisu catches a sea bream ). Puppeteers w ould recite the chant describing E bisu’s miraculous birth and then have the deity perform a dance. These dances were understood to be for the spirit o f the fish in the sea, made present th ro u g h the caught fish (real or artificial). O f note here is th at th e piece involves the invocation o f two O thers: Ebisu and the king o f the fish (who was never referred to by nam e). Ebisu was in essence the “keeper o f the catch,” the deity who em bodied the “luck o f the sea” (u m i no sachi).

Ebisu o-fuda distributed by Awaji itinerant puppeteers.

There were m any places in the chant w here the perform ance could be elaborated to include current events, make specific blessings, and in g en ­ eral assure its relevance. C om m on to all these perform ances was the u tte r­ ance o f the felicitous w ords at the end: “ Oki wa tairyb, oka wa m anzaku, shobai hanjo, tsukisenu godai koso m edetakere!” (In the offing is a big catch, in the hills, an abundant harvest! Business booms! This is an endless era o f felicity!) After the perform ance, the puppeteer w ould leave the householder with an o-fuda on which was drawn a picture o f Ebisu, usually depicting the deity smiling w ith a fish u n d er its arm. M any o f these o-fuda bore the nam e o f the Nishinom iya shrine. This could be placed in o n e’s ow n kam idana in the hom e or at a place o f business. O ften, it was carried on a boat. It served as p ro o f th at the person had th e rite perform ed and also em bodied som e o f the sacredness o f the perform ance. A t the end o f each

year, the o-fuda was throw n away in a special receptacle in a shrine,44 and a new one was received, either from an Ebisu-mawashi or directly from an Ebisu shrine. M ost o f the chants and plot lines used by itinerant Ebisu-kaki are lost, and we can form only a vague idea o f w hat their performances were like. As puppetry festivals lasting several days became m ore popular during the late eighteenth century, people w anted to see these rituals incorporated into the festivals as stage perform ances, and this entailed a transform ation o f the itinerant version o f Ebisu perform ance. These performances are still presented and have been carefully preserved. I have read the several ver­ sions o f Ebisu perform ance texts th at have been w ritten dow n and there are slight variations am ong them. The perform ance piece translated below, now the standard Ebisu-mai piece in the Awaji theater, seems to be representative.45 The lines recited by the puppeteer in performance (serifu) that I translate here reflect a late period in this process o f transfor­ m ation from ritual to stage.46

Perform ance Context: E b isu -m ai a t Shrines a n d on Docks Ebisu perform ance was also presented at festivals held in Ebisu shrines th ro u g h o u t the country, usually observed on the tw entieth o f the m onth, the festival day ( ennichi) for the deity.47 The m ost com m on setting for this rite in the Inland Sea area was on fishing docks or small Ebisu worship sites near docks, where fishermen w ould pray each m orning on their way o u t to sea. This m eant th at unlike the d oor-to-door kadozuke context, which required a great deal o f spontaneity and adaptability on the part o f the perform er, some attention could be given to the preparation o f the perform ance space. Rituals in these locations would have sponsors, per­ haps the head o f a fishing collective o r the local attendant at a small w or­ ship site. The sponsors w ould clean the Ebisu shrine worship hall, where the goshintai, o r spirit vessels, o f the deity were kept, and offer sake and newly cut branches o f the sakaki tree (considered to be sacred) to the deity. Fresh o-fuda could be prepared in advance to distribute to everyone w ho came. The Ebisu-mai perform ance was frequently part o f a longer series o f festivities, which included decorating boats and cleaning them o u t so that Ebisu could be invited aboard for the year. In 1960, as part o f a largerw ave o f nostalgia in Japan, people on Awaji sponsored a num ber o f fish festivals (jjyosai) based on those o f the prewar era. Fortunately, one o f these festivals was docum ented by the ph o to g ra­ pher M une Torasuke.48 H aving cross-checked his photographic record

w ith the recollections o f num erous people on Awaji, I am convinced th at this record is a reliable description o f an Ebisu ritual in a fish festival. T h e 1960 ritual was sponsored by the Fukura Fishing Collective.49 Sev­ eral days before the festival on April 3 ,1 9 6 0 , signs were posted around the docks which read: This coming April 3, we have arranged to have a fish festival, and therefore ask that you observe this day as a no-fishing day. Sanban [(Sanbasb)] and Ebisu-mai will be performed. The Fishing Collective50 Refraining from fishing on this festival day was im p o rtan t because it e n ­ sured an audience. M ore im portant, fishing on th at day w ould be a sign o f disrespect, since the purpose o f the festival was to ask Ebisu to ensure a good catch. People involved w ith the fishing collective began to build a stage several days before the event. Rolls o f m atting to be used as a floor for the audi­ ence and as a ro o f are visible in the background o f the photograph. The stage consisted o f a platform ab o u t three m eters square, around which was w rapped a skim enawa (sacred rope) and sakaki branches, dem arcating the space as sacred.51 The m orning o f the festival began w hen a puppeteer m anipulated Sanbaso before the local H achim an shrine. T he puppets and puppeteers faced the worship hall, and an audience looked o n from behind. A shin to priest sat nearby, presiding over the ritual.52 N ext, the proceedings m oved to the building which housed the offices o f the fishing collective. T he puppeteer m anipulated the Ebisu puppet before the kam idana set up in a tokonomn53 on th e second floor. T he tokonom a contained bottles o f sake and rice, and in it hung a scroll on which was depicted Ebisu holding a fish under his arm. These tw o rituals, clearly intended for a divine audience w ith hum an beings as “uninvited guests,” invoked the deities in whose h o n o r the festi­ val was being presented. T he m orning o f the festival, each m em ber o f the sponsoring group presented him seif before the w orship hall at the local Ebisu shrine, poured some sake over the sacred image, o r set a cup on the offering stand. Prayers for safety at sea and possibly abundant harvests were then offered. T he sake was poured over the deity to “get him ready” and “ loosen him u p ” before the perform ance. Sponsors then joined the spectators for the perform ance.54 In m idm orning the festival began in front o f the fishing collective, and people began to congregate. Because the collective was adjacent to the docks, the very attentive audience was able to sit on their boats, which had been decorated w ith banners and sakaki branches for the occasion. F arm ­ ers also came in their flatbed trucks and sat in them to w atch the perfor­ m ance o f Ebisu-m ai and Sanbaso, this tim e for hum an audiences.55

D uring the Ebisu rite, the puppet deity drank a round o f sake for each kind o f fish caught by m em bers o f the collective, allowing everyone to get m erry for the rest o f the festival.56 Crowds o f people gathered round the small platform laughing as the perform ance to o k place, all clearly having a good time. After the perform ance, o-fuda depicting the Ebisu were distributed, and fishermen then p u t these on the small kamidana in their fishing boats so they could “take Ebisu to sea w ith th e m .” 57 For the rest o f the day, there was a party and a feast, w ith various local groups providing enter­ tainm ent o f all kinds. The party was understood to be for b o th hum an beings and the divine guest. W hat was the co n ten t o f the actual Ebisu rite used in this fish festival? I describe here the Ebisu rite as enacted by a group o f ten perform ers (the usual procedure today on Awaji). It can also be presented by a single p er­ former, w ho enlists the aid o f a sponsor in offering sake to the deity in the place o f the landow ner indicated in the script. I present the m ore elabo­ rate version o f the rite here, referring to all stage directions from that version as “j o r u r i ” and the variations for a single-perform er version in the kadozuke context as “kadozuke.” The perform ance com bines details about the deity Ebisu—his birth, likes and dislikes, and powers—with d e ­ scriptions o f how to worship him properly, as well as hum orous expres­ sions o f his liminal and drunken nature. E b isu -m a i58 J6ruri: On a fla t space prepared in advance, a small chanting stand is

erected to the right o f center. Here sit a chanter who will recite the perfor­ mance and a musician who will provide flu te and drum m ing. A small screen is erected between the performers and the sponsors to demarcate the stage an d hide the activities o f the puppeteers fro m view. Altogether, five or six puppeteers are necessary. Each puppet— there are three in this elaborate version— is m anipulated by puppeteers wearing k u roko (black hoods). Puppets: Ebisu (m anipulated by three puppeteers), a landowner (also three puppeteers), a fisherman (one or two only) Chanting: one chanter and one shamisen accompanist K adozuke: When two performers present this rite, one serves as the m usi­ cian a n d chanter and one m anipulates Ebisu. Because these puppeteers were itinerant, they were not allowed to wear head covering. When there is only one performer, as was usually the case, the performance has no music; the puppeteer performs and recites the story while m anipulating Ebisu. Puppets: Ebisu Chanting: by puppeteer

From the north w est59 enters Ebisu. A fishing pole slung over his shoulder, he enters noisely. E bi SU: La

F

is h e r m a n

La

C

ndow ner

:

ndow ner

h o ru s

:

:

:

T he best god o f luck in the entire cosm os60 has arrived! Well, well! W elcome and do com e in! R ight this way! [ Speaking to his servant, a fisherman] Clean up the house and get o u t som e sacred sake.61 Yes, yes, yes. I understand! [ To fisherm an ] H u rry up a b o u t it! Ifth is Ebisu is offered three cups62 o f sake, then he will perform a dance.

Joruri: Ebisu is offered a cup o f sake a n d drinks it w ith g re a t relish, much to the am usem ent o f the sponsors. H e then slowly begins a dance interlude. Ify o u are asked ab o u t the date o f birth o f this felicitous Ebisu, tell them he was born in the first year o f F ukutoku, on the third day o f the new year, ju st before sunrise, in Shinshu Shinano, in a shrine called Takeigamiya! So easily he was born, he was born, he was born, he was b o rn !63 [Ebisu dances throughout this description o f his felicitous birth.] Take h o t w ater from the storeroom and lukewarm w ater from the warm spring, and mix the h o t w ater and the lukewarm w ater for the baby’s first bath w ater.64 This beneficent old m an, Ebisu Saburozaem on from N ishinom iya, will bestow luck on people w ho have faith in him. W ith the o th e r protective gods, he will bestow luck. Carefully wrap a sacred rope around the storehouse, gather to g eth er th e people o f the village and play a flute and drum . T h e child o f the turtle will mix w ith the voices and bells o f the m aidens, and Ebisu will get carried away, car­ ried away, carried away! [Ebisu dances, becoming more a n d more intoxicated] W earing a folded black h at65 and a h u nting robe, sharply creased, and four-eyed straw sandals, he comes flying into the storehouse w ith a shuffle, shuffle, shuffle! H e looks sharply in all four directions! E b is u :

T he luck o f the sea! T he luck o f the m ountain!

I heap them b o th up and pull b oth lucks to me. Shall we drink another round? Joruri: Sake is poured into a cup a n d is offered to the puppet. H e drinks it with g re a t relish. Each o f the sponsors fo r the rite also take a drink. Kadozuke: A t this point, a kadozuke perform er may have shared a fe lic i­ tous drink with his host. The piece m ay also have been chanted w ith no accompanying actions.66

L et’s drink a round so we can catch lots o f fish! L et’s drink a round so we can catch lots o f plaice. L et’s drink a round so we can catch lots o f whatever! Joruri: A t this point, each fisherman calls out the nam e o f his special fish, a n d a round is imbibed fo r each fish usually caught. This p a rt o f the rite, which can take up to h a lf an hour becomes more and more humorous as people become totally drunk and try to think o f every nam e o f every fish they possibly can to prolong the sharing o f sake.67 C

horus

:

And then, this Ebisu, feeling so drunk, reels to the right— Stagger, stagger, stagger— And then he reels to the left— Stagger, stagger, stagger, stagger!

Joruri: The puppet reels to the right and left, a clever im itation o f a d ru n k­ ard. Since the sponsors a t this point are also quite drunk, this p a rt o f the festivity can go on fo r quite a bit o f time. To this place, he transfers his luck, he transfers his luck! In a flash, he flies out to the ridge. H e rows ou t to the offing. Joruri: The fisherman who was the servant o f the landowner assists Ebisu in rowing. A s he pushes the boat o ff with a long pole, however, he gets left behind and is stranded holding the pole. He precariously balances but then falls into the water, and Ebisu has to rescue him fro m the waves. Once back in the boat, they continue on. H e is already in the offing! Suma and Susaki68 are visible beyond the cresting and falling waves. The voices o f the sand plovers on the beach can be heard calling to their friends, “chiriya\ chiririV’ Just as they scatter and fly away, Ebisu successfully catches and lands a fish. Jbruri: A t this point, a real sea bream, understood to be the most felicitous fish an d king o f all fish, which has been kept in a tub o f water below the performance area, is pulled fro m the tub and caught by the deity Ebisu. Kadozuke: A n artificial fish is used. T hen he dances and dances! [Holding onto the fishing pole, Ebisu performs a dance o f luck]69 In the offing is a big catch, In the hills, an abundant harvest! Business booms! This is an endless era o f felicity!”70 Joruri and kadozuke: Thisfinal line is recited in a slow and sonorous voice, fo r it is the magical chant o f felicity and good luck. Before the performance concludes, the fish is presented to the sponsor who has done the most to make the ritu a l possible. In some cases, it is prepared as sashimi and distributed to all present as p a rt o f a ritual meal following the performance.

196

CHAPTER 5

Ebisu catches a fish as a finale to the performance.

Nogake Butai: Puppets of the Field and Shrine Thus far in this book we have examined the practice of ritual puppetry by looking primarily at the itinerant context for these performances. Ritual puppetry within the kadozuke context continued in the Awaji tradition until the end of World War II, even though the basis of the tradition shifted from ritual to dramatic theater. Kadozuke ningyo performances were not the only expression of puppetry in a ritual context on Awaji, however. While Awaji puppet theaters were popular throughout Japan and maintained busy touring schedules all year, their greatest popularity was at home on Awaji, and here they combined ritual puppetry, joruri contests, and displays of costumes, stage settings, and implements to create the magical world of Awaji ningyo. Puppetry festivals (combining per-

formances from both the shinji and joruri dimensions o f the tradition) were a major cultural performance for Awaji people and an im portant source o f their identity. People on Awaji, I was repeatedly told, were dif­ ferent, more nonbiri (laid back) than other Japanese, and this was evident during these elaborate events. I turn now to this other aspect o f the Awaji tradition, the several-day puppet festivals ( nin£yo matsuri) presented on temporary stages built on harvested rice or onion fields (nogake butai, “stages set up on farmland” ) in the spring or fall.

Shibai wa asa kara, bento β α yoi kara The first time I interviewed someone on Awaji about their recollections o f puppetry, the discussion was dom inated by a detailed description o f the contents o f a picnic box (bentd bako). The woman with whom I spoke was in her late seventies. She described how she made one container o f chirashi zushi, another o f assorted dried fish, another o f various fresh veg­ etables including eda m a m e j1 a container o f various pickles ( o-tsukemono), and an entire box o f rice balls (ο-ηίβίΗ) for the children. She also said that she made a big jug o f cold tea and always remembered to tuck her husband’s pipe and smoking tobacco into the sleeve o f his ki­ m ono as they headed out. She also brought along a flask o f sake for him to share with his friends. As I copiously copied all this information into my field notebook, I thought perhaps I had been misunderstood. Why all this discussion o f mixed sushi, boiled soybeans, and cold tea when I had asked about puppets? After doing research on Awaji for several months, it became apparent that I had not been misunderstood at all. Food was a major part o f what people remembered when they thought about puppets. The Awaji expres­ sion “shibai wa asa kara, bento ga yoi kara,”72 (the plays are from the m orn­ ing because the picnic is good) reveals how people sustained themselves through marathon three- or four-day festivals o f puppetry performances. These performances, usually hosted by one or several o f the puppetry troupes on the island, were the main social event o f the year on Awaji. Many people told me that either they, their parents, or someone they knew had first fallen in love or met their spouse at a puppetry festival. As one man told me, “a ningyd matsuri was a great place to meet the girls.” Young couples could meet freely at these events and go for strolls. A group o f elderly men I spoke with played with the pun o f going to a ningyo matsuri to “look over the dolls.” Clearly a lot more went on at these events than sacred puppet rituals. Business got conducted, people caught up on gossip, and children had a chance to play freely with their friends.

T he tim ing o f these festivals was significant and reveals their underlying ritual structure. They were usually held im m ediately after the harvest or, if in shrines, perhaps right after planting in the spring. These tw o tim es, called in Japanese nokcmki and m eaning “free tim e from agriculture,” becam e occasions to offer rituals o f thanks to the deities w ho m ade plant­ ing and harvesting possible.73 T he presentation o f dram atic e n tertain ­ m en t from the joruri tradition was a secondary en tertainm ent following the dignified rites o f Sanbaso. T he nogake event was a form o f harvest ritual.74 T he atm osphere was festive and casual, and everyone w ould bring tatami m ats to sit on, elaborate box lunches, and gourds o f cool sake to share. In the m iddle o f the day, there was ab o u t a tw o -h o u r break, and everyone w ould break open the ben to and socialize. T he perform ers w ould som etim es join the audience for the feast, creating a social event w here the distinction betw een “perform ers” and “ audience” became n o n ­ existent and all were participants. F urtherm ore, the kami invited to the festival were also present. This picnic to end all picnics is a reflection o f the Japanese religious practice o f the kyoen, or naorai, a b anquet prepared in h o n o r o f the visiting kami. C oncerning this aspect o f a perform ance event, Raz writes: “T he kyden indicates one basic facet o f the Japanese theatre and its audience— the party spirit, pleasure-seeking banquet, in which dis­ tinctions betw een perform ers and spectators, th o u g h existing, are blurred by the nature o f the event.” 75 The nogake event began with the construction o f a tem porary stage, either in fallow rice fields right after harvest or in the open areas in front o f the haiden o f m ajor shrines (usually H achim an shrines). The construc­ tion o f the stage to ok a few days, and the lum ber and labor were provided by the people o f the village, whose responsibilities in the fields were tem ­ porarily over now th at the planting o r harvest was over. T he building o f the nogake butai involved a series o f brief purification rituals, understood to be preparing the site for the invited kami guests. T he day o f the perform ance, people w ould arrive around eight o r nine in the m orning and find their places. Each day w ould begin invariably w ith a perform ance o f Shiki Sanbasb. This perform ance served to purify the space and the village, bless the audience, and invite the kami o f the village to com e to the festival and enjoy the perform ances, for it was largely in their h o n o r th a t the entire event was taking place. I t has been suggested th at the elaborate series o f joruri perform ances lasting several days were secondary to the m ain purpose o f inviting and blessing the deity.76 T he next p art o f the festival, namely the day’s entertainm ent, took a variety o f form s, b u t was understood to be entertainm ent for b o th the hum an audience and the deity.77After the perform ance o f Sanbaso, one o f

the major joruri pieces would be presented. In a usual day, the theaters would present one piece in its entirety and perhaps two or three famous scenes (dan) or soliloquies (sawari) with which nearly everyone present would be well familiar. Some people in the audience frequently were amateur reciters o f joruri and would know the pieces by heart, listening intently to the performance to improve their technique. The presentation o f scenes rather than entire pieces raises an interesting issue for our understanding o f these puppetry festivals. Since a scene is by definition only a part o f a larger play, presenting only this segment means that the audience had to fill in the rest. Before and after the performance o f a single scene, I was told, nearly everyone present would reconstruct the rest o f the play, arguing occasionally over nuances or details. This arrangem ent provided a built-in opportunity for meanings in the puppet theater to be reinterpreted and experienced. These conversations before and after the presentation o f sawari should therefore be understood as an integral part o f the performance. Much like the N orom a ningyd discussed in chapter I, or the use o f kydgen in N oh, short light entertainm ent during intermissions made it possible for people to absorb the heavy tragedies o f the day’s joruri performances. There were two common forms o f short performance, both o f which dealt with the them e o f rapid transformation and im perma­ nence. The first was called hayagawari, (rapid changes). It involved a se­ ries of seven transformations (nana-bake) o f dolls and costumes and was considered to be very amusing. One puppeteer (not a puppet) would come onto the stage, and in a short period o f time he would strip away his outer costumes and reveal a different identity under each layer. The standard seven transformations were a fox, a princess who becomes a dem on, a zato (masseuse), a hanagusa (flower vendor), kam inari (a th u n ­ der and lightening dem on), a fox again, and two dolls, a joro (prostitute) and her yakko (customer), sometimes played by the puppeteer and a doll he would manipulate and often involving a gender switch on the part of the puppeteer.78 We have noted that the annual ningyo matsuri took place on temporary stages, constructed each year for the event. These stages were by no means simple; they had several levels, enabling backdrops to be changed and the area o f the stage to be brought closer toward or further away from the audience. The result was a stage with several depths. D uring the intermis­ sion, this device was p u t to another purpose—that o f dogugaeshi, a distinct genre o f performance is still presented in an abbreviated form at the Awaji Ningyd Joruri Kan.79 The inner stage was illuminated with a row o f can­ dles, and while the shamisen performer played a lively piece with a percus­ sion accompaniment o f wooden clappers, as many as eighty-eight scenes

w ould change on the stage in very rapid succession. The images varied depending on the num ber o f screens, but they usually depicted either a journey or m otifs associated w ith longevity, felicity, the seasons, and p u ­ rity, as well as animals and mythical beasts. These were usually presented during the evening interm ission, w hen the fading light w ould create a dreamlike impression. U m azum e M asaru, the director o f the theater and my chief source o f inform ation, suggests th at the dogugaeshi’s significance can be under­ stood as an expression both o f the status o f the particular theater group staging the perform ance and o f the m ore general sense o f illusion and fantasy that the w orld o f the puppet stage seeks to create. C oncerning the form er, he notes that the num ber o f scenes th at a particular theater stages (from fifteen or twenty to as m any as eighty-eight) reflects its prosperity. The screens are heavy and are carefully painted (traditionally on m ulberry paper). Consequently, having a large num ber o f screens is a vivid and staged display o f having the means to have them painted and hauled around from perform ance to perform ance. F urtherm ore, various theaters w ould advertise their dogugaeshi as one o f the marvels o f their perfor­ m ance. U m azum e recalls that as a child, he and his friends w ould go to the day-long perform ances mainly to see the changing screens, for this was the part o f the perform ance th at m ost appealed to children. U m azum e also suggests th at the changing screens express w hat is called “yugen no sekai.” While this phrase has a m ore precise m eaning in the w orld o f N oh aesthetics, in general usage it connotes the w orld o f pro fu n ­ dity and mystery always present at the edges o f everyday consciousness. T he rapid change o f screens took place behind a wall o f candles, which flickered w hen the stage props were m oved, creating the illusion o f a n ­ other w orld, grow ing deeper and deeper as one image was superim posed upon another, n o t unlike the w orld o f dream s or half-consciousness. The images form ed a series o f free associations betw een visual, m ythic, and poetic m otifs in Japanese culture— the plum or cherry tree, M o unt Fuji, various tem ples, the seven gods o f luck, tornadoes, dragons, the p h o e ­ nix— all presented one right after another accom panied by shamisen and w ooden clappers in a m ost dram atic ambiance. Like the hayagawari, the them e o f ddgugaeshi is transform ation, the m o tif o f the ningyo matsuri. O n Awaji, it was understood that the world o f puppetry was one o f inner transform ations, w hether the perform ances were rituals or dram atic pieces. T he puppets, as being once rem oved from the hum an realm, prom pted one to explore the life o f the spirit, an explo­ ration reflected in the plays and rituals and underscored by the rapidly changing scenes o f the dogugaeshi and the quick identity changes in the hayagawari.

The Last Puppet Show: Broken Puppets an d Sacred Parts In I9 6 0 the last m ajor pu p p etry festival to o k place on Awaji. It was orga­ nized by th e fledgling Awaji N ingyo Za, a reconstituted pu p p et troupe p u t to g eth e r in response to a request for a foreign tour. M ajor pieces were presented by people form erly associated w ith U em ura G en no Jo, once the largest and m ost prosperous th eater o n the island. T here had n o t been a m ajor ningyo m atsuri on Awaji for nearly tw o decades, and this was a charged and nostalgic event. I t was p h o to g rap h ed by the S um oto p h o to g ­ rapher M une T orasuke, and so a record o f the entire process has been preserved .80 M any people I spoke w ith rem em bered this festival. O ne m an to ld me th a t he had a great tim e there, m uch as one has a good tim e at a funeral b anquet, rem em bering a past life even while acknow ledging th at it is dead and gone. T he discussion o f ningyb m atsuri as a cultural perform ance show ed th at puppets form ed a nexus around w hich o th er significant actions turned. Rituals surro u n d in g the treatm en t o f puppets them selves— how they were handled, respected, and repaired— can also tell us som ething ab o u t the m eanings o f Awaji ningyo in people’s lives. T he puppets used in the Sanbasd rites were accorded a different status in the Awaji tradition th an o th er puppets. T hey were regarded as shinsei (sacred) and had to be stored in a designated location. A ningyo mawashi w ould keep them in a high place in his hom e, often even on the kam idana itself. M any people recalled seeing these puppets up on kam idana in th ea­ ters w ith offerings o f rice and sake before them . Even today at the N ingyd Joruri Kan in Fukura, the Sanbaso puppets are kept in a special case w ith offerings in front o f them . U nlike joru ri puppets, w hich could be taken apart after each perform ance, Sanbaso puppets rem ained assem bled at all tim es, except w hen they were being repaired. W hat happened to pu p p et kashira w hich could n o longer be used for som e reason and were beyond repair? U ntil the late n in eteen th century, they could n o t simply be throw n away o r recycled. And it was n o t suffi­ cient to p u t them in a rt collectors’ cases as examples o f folk art. P uppet heads and body parts o f th e Sanbaso puppets received preferential tre a t­ m ent: they were buried, and a kuyd service was perform ed for them . Kuyo, technically a B uddhist term , refers to a ritual practice th a t is at once a worship service, a form al apology and an expression o f gratitude, an appeasem ent rite, and a funeral. O ften recipients o f the rites are inani­ m ate objects such as needles, calligraphy brushes, household cleaning in ­ strum ents, and even underw ear. A kuyo is conducted w hen these objects are b ro u g h t to a tem ple o r shrine; the people w ho b ro u g h t them make a

series o f invocations and statem ents o f apology or gratitude; the objects are then ritually disposed o f in a way th at concretizes the feelings o f the people w ho used th em .81 Broken puppets routinely had kuy5 rites perform ed for them , and after the rite was concluded, they were buried in a cemetery. N ear the Sanjo O m ido on Awaji, there is an area which was once the pu p p et cemetery, called dekozam ma. T he practice o f burying these images carved “in the shape o f the h u m an ” suggests an awareness th a t while m atter and spirit may appear to be separate orders, once spirit has encountered a m aterial form , the latter cannot retu rn to m ere m atter b u t becom es m atter set apart. A dilapidated puppet— a head, arm s, perhaps a costum e, rattles, flutes and masks—will never again be m erely the sum o f its parts. Today, they are p u t in m useum s or glass cases, a practice th at worries m any older puppeteers.82

R itual Puppetry and th e Pow er o f Sym bolic A ction T he rituals described above raise a fundam ental issue th at occupied us in chapter I: W hy use puppets and n o t hum an actors? O n the one hand, this could be considered a m atter o f aesthetics and even fashion, as in both Ebisu and Sanbaso rituals around Japan, there are examples o f these ritu ­ als being presented w ith hum an actors. Even w ithin a relatively circum ­ scribed tradition such as Awaji ningyo, there is n o t a single answer to this question, since the decision to use puppets has several ramifications for the semiotics o f the rituals themselves. T here are a n um ber o f shifting, overlapping, interpenetrating, and in terdepen dent understandings o f the puppets in these rituals: ( I ) the puppet as a spirit vessel ( torimono) for sacred beings; (2) the pu p p et as a body substitute ( bunshin) for an absent (deceased) ritual specialist; (3) the puppet as a protective shield (tate) for the puppeteer; and (4) the puppet as a concrete m etaphorical self (daiyaku) for the puppeteer.83 T he m ost significant use o f th e pu p p et in b o th o f th e above rituals was as a to rim ono, a physical object th at served to draw and contain spirits sum m oned to the hum an com m unity in the ritual context. U sing a physi­ cal object as a receptacle for sacred forces is a com m on practice in Japanese shamanism . In h eren t in it are the dual ideas o f containm ent and p ro te c ­ tion. The physical object allows the sacred forces to be contained in a single location, and it allows the religious specialist w ho has sum m oned the force to be protected from the danger th at a superm undane force entails. To the extent th a t the puppets in these rituals were to rim o n o , they are typical o f a w ider pattern in Japanese religious practice. B ut this was n o t the only m eaning o f the puppet.

T he m yth which founds the Awaji tradition makes it very clear th at the pup p et is actually a ritual substitute for a deceased priest. I f we take this m yth and generalize the m eaning o f this single m otif, we can see th at the pup p et is a surrogate sacred specialist. W hen puppeteers challenged the authority o f the Nishinom iya shrine by setting o u t on their ow n and estab­ lishing their ow n center, the p u ppet, n o t the puppeteer, became the new locus o f authority. T he ritual device o f using a puppet allowed for the direct challenge to the Nishinom iya center while deflecting the pow er th ro u g h a substitute object. A third understanding o f the pup p et was its role as a protective barrier betw een the puppeteer and the liminal beings and noxious forces he sum ­ m oned. Because p art o f the role o f the puppeteer was to rem ove pollution from villages, the p u p p et stood betw een the puppeteer and the forces his perform ances attracted, serving as a sort o f a shield and pro tectin g the puppeteer from these forces. Finally, we notice th at as the ritual tradition developed, the puppets gradually becam e larger and larger, until by the end o f the nineteenth century they stoo d nearly half as tall as the puppeteers themselves. It is possible to argue th at at som e level, the puppeteers un d erstood these p u p ­ pets to be m etaphors for themselves. Just as puppets were the target o f pow erful and dangerous forces in society, so to o were the puppeteers as m em bers o f Japan’s m arginalized outcast group, a targ et for the projec­ tion o f negativity in society. Puppeteers m anipulated puppets to control dangerous spiritual forces ju st as the ruling powers m anipulated outcasts to control the danger o f peasant unrest. Perhaps the m ost im p o rtan t feature o f puppets as ritual m edia is their im itative quality. Puppets can never be equated w ith or reduced to that which it im itates. They are once rem oved from the hum an realm and ex­ press the awareness th a t although the spiritual w orld can be m etaphori­ cally described th ro u g h reference to the m aterial w orld, it can never be reduced to it. As U m azum e M asaru p u t it, “Awaji puppets aren’t secondclass hum an beings. They are first-class ning yo.”

6 Puppets and W hirlpools: Icons, N ostalgia, Regionalism , and Id entity in the Revival o f Awaji Ningyo

T h e M an w ith th e B icycle In Septem ber o f 1957, at the invitation o f the AU Japan Regional Per­ form ing Arts C onvention (Z enkoku kyodo gein5 taikai),1 a puppeteer from Awaji presented a brief dem onstration o f puppet m anipulation at the annual m eeting held in a Tokyo hotel. T he audience was predom inantly foreign, and the entire dem onstration took only ab o u t a half an hour. A ttending the event was the director o f the N ational M useum o f M oscow in th e Soviet U nion. H e was fascinated w ith the large puppets, w anted to offer an invitation for Awaji ningyo to perform in the Soviet U nion, and asked to which theater on the island he should direct his invitation. A t the tim e, only three theaters on Awaji still existed. T he war had been over for only twelve years, and the econom ic situation in rural Japan was still strained. Prewar puppeteers had either boxed up their puppets and stuck them in storehouses, sold them to collectors and foreigners, o r used them as firewood to heat their baths. M ost puppeteers had n o t picked up a p uppet since the late 1930s. A lthough three theaters still rem ained, they existed in nam e only, and none o f them had any puppeteers. T here were n o t enough people interested in Awaji puppets on the island to scrape to g eth er a troupe. It w ould be necessary, then, to com bine all the extant resources on the island to get to g eth er enough people for a p roper perfo r­ m ance, crossing lines o f prew ar tro u p e form ations. T he following year, a full tro u p e o f Awaji perform ers w ent to the S o­ viet U n io n and perform ed in M oscow and L eningrad. T he perform ers, ranging in age from sixteen to seventy-five, were widely acclaimed. To the audiences w ho saw them perform , they appeared to be a unified group w ho had worked to g eth er for many years. T he story o f this tro u p e ’s form ation, however, is m ore com plex and reveals the dynamics o f trad i­ tion reform ulation and invention, the showcasing o f “traditional culture” to foreigners, and the realities o f the postw ar fascination w ith a retu rn to roots th ro u g h w hat has come to be called m inzoku geino (folk per­ form ing arts).

At the tim e o f the Soviet invitation, U m azum e M asaru (the director o f the successful Awaji N ingyo Joruri Kan until the end o f 1994) was in his late tw enties. H e had grow n up in the village o f Sanjo, near the O m ido H achim an D aibosatsu, the cerem onial center o f Awaji puppetry. H is fa­ ther, a farm er by profession, had played the shamisen for local puppet troupes. H e rem em bered walking dow n the street in Sanjo-m ura as a boy to the Yoshida T roupe, the p ro m in en t th eater on the island, and w atching preparations for rehearsals and perform ances. H e had spent m uch o f his childhood hanging around the theaters listening to the puppeteers, ch an t­ ers, and musicians discuss and practice the various jo ru ri pieces. H e looked forw ard to the annual ningyo m atsuri from one year to the next. U m azum e him self had learned how to participate in the sacred perfor­ m ance o f Shiki Sanbasb as a reciter. B ut in 1957, Awaji perform ances no longer took place. Seasonal dokum bo mawashi did n o t come d o o r to door at N ew Year to offer rites of blessing and purification. T he puppet Sanbaso was n o t m ade to dance in front o f household altars or small shrines in rice fields, near w ater sources, or in the w orship halls o f H achim an and Ebisu shrines. People no longer congregated in one an o th er’s hom es in the evenings after a day’s w ork to practice jbruri recitation. Ebisu-mawashi no longer p erform ed on fishing boats or Susaki Island in Fukura Bay to ensure a good catch and a safe year at sea. Tw entyyears before, these events were com m onplace. It seems that the tradition o f Awaji puppetry, w ith both its ritual and joruri com ponents, once such an everyday p art o f the life on the island, had died a rather sudden death because o f the poverty o f the p o stw ar period, the censorship o f the occupying forces, the advent o f cinema (and later television) and the neglect and even hostility o f Awaji people themselves. Like m any perfor­ mance traditions th ro u g h o u t rural Japan, Awaji puppetry was defunct. W hen he heard o f the invitation to go to the Soviet U n io n , U m azum e recognized an o p p o rtu n ity to reorganize the older puppeteers on the is­ land in to one group, which w ould ensure th a t Awaji p uppetry w ould sur­ vive in at least som e form . T here was a lot o f talk ab o u t the invitation, but no one seem ed to do anything a b o u t it, so he to o k it u p o n him self to get everyone together. H e g o t on his bicycle, rode around Awaji, and took the ferry over to Shikoku. H e visited all the old puppeteers and chanters and tried to convince them to p u t to g eth e r a troupe just one m ore tim e. H e tracked dow n farm ers, fisherm en, and w orkers in prefectural offices w ho had once been puppeteers, frequently in te rru p tin g them at work. In the end, tw o o th e r perform ers agreed to com e to g eth e r to practice. A long w ith the puppeteer w ho gave the perform ance in Tokyo, th at m ade four, including U m azum e. A few others slowly joined up. In the m eantim e, the N ational T heater o f Japan had invited the group to Tokyo for a w arm -up perform ance and had given them a small loan,

b u t it was n o t enough to sup p o rt their practices. T hey needed stage props, new costum es for the puppets, and som e small rem uneration for the perform ers to make all the effort w orthw hile. They began to ask people on Awaji for contributions. W hen a small tro u p e was finally assem bled, one o f the recruits, Naniwa Kunie, sewed the costum es for the puppets, and U m azum e helped paint and decorate them . A practice hall was set up o n Awaji, and they m ade it to Tokyo for the perform ance at the N a ­ tional T heater. Soon thereafter, the troupe m ade its successful trip to the Soviet U nion. Back on Awaji after the tour, the newly form ed troupe m anaged to raise en o u g h m oney to pay o ff som e previous loans. It m oved first to a small building ow ned by a savings and loan firm and later to a small theater in Fukura above a tourist gift shop near the w aterfront. T here they gave brief perform ances to tourists several tim es a day. In 1987 they m oved to a new location on the cape overlooking th e N a ru to Straits, the narrow passage connecting Japan’s Inland Sea and the Pacific O cean. This new center is a m odern building housing a theater specially built for puppetry, w ith green room s, showcases for artifacts, an elaborate stage w ith a custom -m ade curtain, and a business office. T he com plex, a m inia­ ture mall, also has a gift shop, restaurant, and movie hall attached to a scientific exhibit hall explaining the natural forces th at create the N aru to whirlpools. O utside, a large parking lot can accom m odate dozens o f to u r­ ist buses. T he building o f this new theater and tourism center was m ade possible in p art w hen in 1976 the Awaji pu p p et tro u p e was granted the status o f an im p o rtan t intangible folk-cultural p roperty (juyo m ukei m inzoku bunkazm ), a form al designation from the Agency for C ultural Affairs (B unkacho), which provides form al preservation status and funds for the preservation o f this perform ance tradition. W hen I first encountered Awaji p uppetry in its form er location in the fall o f 1977, all the perform ers were essentially volunteers, and the small theater doubled as a m useum , displaying objects and old m anuscripts re ­ lating to Japanese ritual and joruri puppetry. Now , m ore th an thirty five years after the successful Soviet U nion foreign to u r, U m azum e, w ho had ridden his bicycle aro u n d the island and over to Shikoku to find enough perform ers, continues as the director o f the new theater in F ukura. H e told me why he w ent to such effort back in 1957 and over th e years since then to establish Awaji ningyo perform ances as a viable theater: “At th at tim e, I had a terrible sense o f loss. I felt like som ething o f Awaji was slip­ ping away. M y childhood. T he sounds. T he banners announcing the ningyo m atsuri. T he rehearsals. T he fancy dogugaeshi. I can’t really make light o f my reasons. They felt very serious to m e then. I still feel th a t way now. I t was painful to feel som ething so m eaningful sliding th ro u g h your fingers, and yet do n o th in g b u t talk about the good old days.”

H is bicycle ride around Awaji over th irty years ago set in m o tio n a process w hich m ust in som e sense be called a revival o f the tradition. Today, the Awaji N ingyo Joruri Kan operates w ith a tro u p e num bering over th irty full-tim e paid perform ers, an office staff, and a full-tim e direc­ tor. D u rin g holidays and the sum m er m onths, there may be as m any as eight perform ances daily, presenting works from the large repertoire o f joruri. T he perform ers are all paid by th e prefectural governm ent as civil servants, an irony n o t lost on the m em bers o f the tro u p e, given the “n o n h u m an ” status puppeteers suffered in the past. T he tro u p e makes num erous overseas tours and is frequently featured in the Japanese m edia and even international films as p ro o f th a t p erform ing arts can be rescued from extinction and m ade viable again.2 E verything ab o u t this revival in ­ dicates th at Awaji p u ppets have m ade a com eback, th a t th e dam age o f tim e has been u n d o n e. U m azum e M asaru (w ho now drives a car) is a frequent guest o n national talk shows. H e still lives in Sanjo, w here his family has lived for generations and w here his father once played th e sham isen for the local p u p p e t troupes. AU this success and activity raises a fundam ental question: W hat does this revival reveal— and conceal— about p eo p le’s experiences o f their own p u p p etry tradition? Clearly for U m azum e M asaru and o th er people on Awaji involved in this hard work, som ething ab o u t Awaji pup pets was p a rt o f their id en ­ tity— an identity held w ith m ixed em otions, since m any o f them have suf­ fered from the stigm a o f being the descendant o f a puppeteer. B ut this am bivalent relationship to the past is absent from discussions ab o u t reviv­ ing the tradition. M ost o f th e people involved speak o f the deep pain they felt w hen they rem em bered the prew ar period and realized th at this past was disappearing. A com m on response has been to devote oneself to p re ­ serving the tradition. T he Japanese term s m ost often used to d en o te this process are fu k k a tsu (revival, even resurrection), fu k k o (restoration), and saikai (reopening). Regardless o f the term used (and m ost people use them interchangeably), in h eren t in this process is an o rien tatio n tow ard the past as a tim e o f value and m eaning. W h at is authentic, in o th er w ords, is w hat is past. This charged o rientation tow ard th e past, as th e engine th a t drives the revival, suggests a n u m b er o f things a b o u t this interesting cultural p ro ­ cess. O n the one h and, such revival m ovem ents suggest th a t people can­ n o t Jive w ith o u t a past, even th o u g h restru ctu rin g it involves inventing large segm ents o f it, rew orking others to adapt to the contem porary situa­ tion , and m arketing it in ways and contexts th a t w ould previously have been o u t o f the question. T he revival o f a ritual tradition decades after its demise is therefore an excellent place to study how people develop narratives for their identities th ro u g h orientation to b o th ritual and the past. O n th e o th er hand, a revival m ovem ent becom es an o p p ortunity

for revision o f the past as well. In this case, revisionist revival may well have an anesthetic effect. As a defunct tradition is b ro u g h t back, new dim en­ sions o f the tradition can be invented and extant aspects o f the past can be enhanced and enlarged; meanwhile old and painful realities can be blurred or om itted and ironies o f sponsorship ignored. O u r discussion, th en , m ust allow for b o th the ideological and therapeutic aspects o f re ­ vival. Awaji puppetry reveals the relationship betw een the past and the present, nostalgia and identity, reality and fiction, creative rem em brance and selective forgetting. W hat was an Awaji puppeteer doing in Tokyo giving a lecture-dem onstration about “a living theater tradition” when in fact the tradition was already defunct? Why was the audience primarily non-Japanese? W hat idea o f cultural display was at work in this event, and how did it influence the next stage o f this process, assembling a reconstituted theater group uniting traditionally distinct troupes each insisting on its own style and aesthetics? F u rth er questions arise. W hat is the significance o f this m ajor com e­ back o f a perform ance tradition using puppets w hen supported by heavy governm ent subsidies? W hat role did nostalgia play in this process? H ow did this process appropriate a n d becom e appropriated by o th er discourses o f identity and nativism in Japan? H ow does the invocation o f foreign­ ness operate in this process o f retrieval? M ost im portant, w hat has hap­ pened to the ritual structure o f the kadozuke puppetry perform ances in this revival? This plethora o f questions requires discussion o f the following aspects o f the revival m ovem ent: ( I ) the participation o f Awaji revivalism in the larger discourse o f folk studies and folk society preservation; (2) the dis­ course o f nostalgia in contem porary Japan; (3) regionalism and the p ro ­ cess o f icon creation in the production o f Awaji as a regional identity, and the role o f puppetry as a focal icon; (4) the shift to a new “exoticism ” in the retrieval o f the ritual tradition o f Awaji puppetry (perform ances o f Sanbaso and E bisu); (5) the significance o f m oving these itinerant rituals into the theatrical semiotics o f stage perform ance; (6) the appearance o f new splinter groups o f revival activity on Awaji; and (7) the role o f the foreign scholar in this process o f revival.

S tages in th e P rocess o f R evival In the last several decades, the move to “ reclaim Japan’s past” has focused largely on the revival o f folk perform ing arts {m inzoku geino). T he use o f general term s to group together disparate ritual, aesthetic, and perfor­ m ance genres was necessitated th ro u g h the grow th o f the m inzokugaku (folk studies) m ovem ent, which gained m om entum in Japan at the end o f the Taisho period, largely under the charismatic leadership o f Yanagita

Kunio. This new m ovem ent needed to delineate its focus o f inquiry, and consequently it devoted m uch discussion n o t only to w hat constitutes “the folk” b u t also to w here and w hen they existed and to w hat kind o f value should be attached to the products o f “ folk c u ltu re.” 3 Before W orld W ar II, the term m in zokugeijutsu (folk arts) was fashion­ able, b u t as the term applies also to non p erfo rm in g arts, it proved too inclusive a category to be useful. T he term kyddo buyo (regional dances) also missed the o th er dram atic aspects o f the perform ance traditions, m any o f which did n o t privilege dance as the center o f perform ance. T he term kyddo ^ein o was the norm during m uch o f the 1950s, although kyddo implies a regionalism and rural quaintness th at m ade the term difficult to apply to urban traditions. N evertheless, the overtones o f w hat I will call “regionalism ” rem ain in the folk studies m ovem ent in Japan. In h e re n t in the idea o f regionalism is the view th a t although there is a com m on w orld th a t unites all phenom ena o f the folk, each region has its ow n unique version o f this larger reality. R ecreating these regional identities, fre­ quently as tourist destinations, has been a m ajor p art o f contem porary nostalgic discourse in Japan. It can be argued th a t Japanese nostalgia now constitutes som ething o f a space tim e-continuum : nostalgia is felt n o t only for the past b u t for a past tied to a geography, and both are produced th ro u g h discourse and th ro u g h regional developm ent campaigns. O ften, folk perform ing arts m ake good icons for such regionalism , and this was certainly the case w ith Awaji. It was H o n d a Yasuji w ho developed the criteria for determ ining w hat w ould be considered m inzoku geinb. H is schem a has becom e norm ative and is to a large extent uncritically accepted by scholars in Japan.4 For H o n d a and others, the problem has been one o f delineating the lines b e­ tw een th eater and ritual. Are m inzoku gein5 m erely theater traditions o f the countryside? H o n d a and others o f the m inzoku geino school make the distinction th at these perform ances, unlike the formal th eater traditions o f N o h , Kabuki, and B unraku, are tied to the seasonal religious life o f the people and m u st be in terp reted in th at context. H e insists th a t these ritu ­ als m ust be seen as p art o f the larger phen o m en o n o f w hat is term ed m inkan shinko ( “folk shinko” ). T he key term here is shinko, w hich is fre­ quently m istranslated in to English as simply “ b e lie f’ b u t conveys the n u ­ ances o f w orld view, praxis, and faith as well as belief.5 H o n d a w ould argue th a t Awaji puppetry m ust be understood as p art o f the everyday lives o f people on Awaji if it is to be u n d erstood at all. T he categorical schem a o f m inzoku geino, based on belief and insistent on a lim ited understanding o f context, becom es problem atic w hen revival implies an “in terv en tio n ” (to use Richard Schechner’s term ) th a t reshapes ritual according to the sensibilities and constraints o f stage perform ance. In w hat way is revival a kind o f cultural perform ance? H o n d a ’s agenda insists on a traditionalism th a t claims th at only the past is authentic.

In 1950, influenced b o th by the sense o f loss o f traditional culture th a t the postw ar era fostered and also by the grow ing popularity o f the m inzokugaku m ovem ent, the Japanese D iet passed legislation designed to p ro tect its cultural heritage. T he Bunkazai H o g o h o (C ultural Properties P rotection Law) raised the im p o rtan t issue o f w hat constitutes cultural property.6 W ho and w hat is w orth preserving and w ho and w hat can be allowed to becom e extinct? W hat constitutes a rt and perform ance? In the same year, as p art o f this larger agenda and th ro u g h su p p o rt from the M inistry o f E ducation, the AU Japan Regional Perform ing Arts C o nven­ tio n was form ed as an annual convention w here regional perform ing arts could be presented. Renam ed the AU Japan Folk P erform ing Arts M eeting (Z enkoku M inzoku G eino Kai) in 1958, these annual m eetings n o t only showcased “traditional Japan” for foreign scholars and even tourists, they also provided incentive for regional groups to get to g eth e r and practice. Perform ance form s th at had been defunct for decades were reinvented, based on recollections o f elders, scholarly inquiry, and im aginative re ­ gional com parison. Ritual perform ance traditions still barely alive were given the chance to gather strength in the face o f m odernity and find a new audience at a national level. Being a participant in these large public events was highly prestigious and had a cachet o f authenticity. Each convention served as a sort o f academic conference w here e th n o g ­ raphers, folklorists, and anthropologists could present their research on the various traditions being showcased. Since these events becam e im ­ p o rta n t m om ents in the building o f academ ic careers, it was in the best interests o f the academics th at these perform ances be as “ au thentic” as possible. Each o f these perform ance events represented an incredible am ount o f logistical planning, research, and retrieval activity. Som e claimed th a t only the version o f the past th a t they presented, th reatened by the current situ ­ ation, could be considered truly Japanese. Scholars in the m inzoku geino school claimed that various ritual perform ances around Japan represented at once the tru e and the vanishing. Efforts to preserve these rituals often had a decidedly religious zeal: the authentic Japan was being rescued from the onslaught o f m odernity. A t stake was a definition o f authenticity and nativism. W hat does it m ean to be truly and authentically Japanese? Veracity concerned m ore than merely determ ining w hat had happened before; it became an exercise in the staging o f value. T he presentations by scholars, often from the same stages w here the perform ances were to take place, becam e as m uch a part o f these events as the “ritual perfo rm ­ ances” themselves. T h ro u g h o u t the 1950s, the M inistry o f E ducation asked regional and prefectural boards o f education to identify local perform ance traditions. In n um erous villages th ro u g h o u t Japan, com m ittees were form ed to iden-

T he priest from the Kohyo shrine near U sa in O ita Prefecture dem onstrates the m anipulation o f a puppet.

tify the folk perform ances o f local shrines o r tem ples and carefully reco n ­ stru ct their histories, often filling in details and borrow ing dance, music, and costum ing from o th er areas to lend an air o f authenticity. Claim ing historical accuracy was o f ultim ate im portance, even if the perform ance tradition had been in te rru p ted by as m uch as eighty years. A stu n n in g example com es rig h t o u t o f the history o f ritual puppetry. In chapter 3 I n o ted the involvem ent o f ritual puppeteers in the H oj5-e o f the Usa H achim an shrine. Texts from th a t shrine claim puppeteers played a role in the defeat o f the H ayato by th e centralized governm ent in 720 C .E . U ntil the late Meiji period, puppeteers from tw o subsidiary shrines o f the U sa com plex, the Kohyo and Koyo shrines, presented a sum o m atch w ith puppets as p art o f this larger Usa rite. In spite o f this trad ition being defunct for over eighty years (long enough for firsthand inform ants to have died), today these rites have been revived, and the perform ances are incredibly detailed. T he Kohyo shrine has received large subsidies to rebuild its m ain structures, including a new fireproof struc-

212

CHAPTER 6

Musician puppets from the Koy6 shrine.

ture to house the ancient puppets. Basing their reconstruction of performances on written texts, the physical capabilities of the extant puppets they possess, and on the most general ideas of Shinto performance, puppets again perform the ritual appeasement ceremony every four years as part of the Usa H6j6-e, although its resemblance to the earlier tradition is dubious? The new puppet ritual is presented as an example of an ancient Japanese folk performing art. It appears, however, that only one part of this ancient rivalry between these two shrines is now represented: the other shrine that had been involved in this rite in the past, Koyo, received no subsidies and has not been able to do more than maintain its very old puppets. In the early postwar period, the Ministry of Education (Mombu Daijin) and Ministry of Cultural Mfairs (Bunkacho) also encouraged the formation of hozonkai ("preservation societies") for folk arts, especially those that were primarily performance based. People in Cultural Mfairs offered guidance and consultation to municipalities on how to organize a hozonkai and how to enable it to function most effectively. These societies

worked to ensure the continual transmission o f tradition, to safeguard the stages, props, texts, and other physical artifacts o f the tradition, to record the histories o f their development, and to see that opportunities were cre­ ated to keep those who were central to the tradition involved in it. Fre­ quently, however, the financial burden o f these activities was borne not by the government but by wealthy private patrons. This allowed for a certain freedom at the local level, but it also meant that personal bias and favor could be manipulated. W hat got preserved, in many cases, and how it got presented in a larger context was frequently left up to the personal tastes o f the patron. The revival o f Awaji puppetry is no exception. In 1976 the Cultural Protection Law was revised to provide tangible support for such theater traditions. Beginning in this year, awards were made recognizing certain traditions as im portant intangible folk cultural property. Recognition was granted based on recom mendation by a board o f scholars, most o f whom were folklorists from the minzoku geino school. As a result, to receive such a prize, ritual traditions had to conform to the aesthetics o f rusticity, regionalism, and rurality. As already noted, Awaji ningyo shibai was among the recipients o f this funding. For Awaji, efforts at preservation were started locally even before such organizations became fashionable and government subsidy was possible. As early as 1935, an organization calling itself the Awaji Puppetry Arts Revwal Society (Awaji Ningyo Geijutsu Fukko Ky5kai) was organized with Nakano Toraichi as its head. This organization was interrupted by the war, and in 1949 Fudo Saiichi (the father o f Fudo Satoshi, the old man with the photographs mentioned in the opening pages o f this book) founded the Awaji Ningyo H ozonkai.8 An employee at the local city hall, Fudo Saiichi wrote the small but extremely useful A w aji Ningyo Shibai no Turai (Origins o f Awaji puppetry), which marked the founding o f the new A w aji Ningyo Hozonkai organization. This book was followed by local efforts to collect docum ents and preserve artifacts, but the fact re­ mained that out o f the eighteen fully functioning troupes at the beginning o f the nineteenth century and the more than several hundred itinerant puppeteers living in Sanjo, hardly any troupes and no itinerant puppeteers were perform ing at the time. In 1907 the num ber o f troupes was twelve; in 1936, seven; and in 1951, only four, none o f which was really manag­ ing to survive. The revival movement had an obvious task, but in spite of its efforts immediately before and after the war, the trajectory toward ex­ tinction did not really slow down. It seems clear that nostalgia, for all its clout in the folklore studies movement, needs a constellation o f other factors to enable any movement to get o ff the ground. Nostalgia demands a focal point, an icon around which the currents o f feeling can flow.

N ostalgia as Experience, N ostalgia as Id eo lo g y Most people in Japan commonly use the adjective natsukashii, the verb natsukashimu, or the noun forms natsukashisa or natsukashimi to de­ scribe this vague sense that “things just aren’t what they used to be.” These are everyday words in Japanese, used to refer to events and sensa­ tions that evoke connections with the past and feelings o f hom e. For ex­ ample, a person when eating a good bowl o f soba after eating only those from machines for months may remark, “Aah! Natsukashii!” This term natsukashii has a variety o f referents, including “feeling,” “yearning for or missing som eone,” “becoming attached to som eone.” Alone as a noun, the written character means “ bosom, heart, breast, pocket,” implying something close to one’s heart and seat o f emotional attachment. Sometimes the character is translated simply as “nostalgia.” In Japanese these days, people familiar with contemporary discourse in literary critical studies may use the word nostarujii is used to refer to the general cultural phenom enon o f romanticizing the past. This im ported word (and example ofgairaigo, or “word coming from outside” ) conveys this general sense o f recollection and desire to return to a romanticized past. But because the word nostarujii has entered Japanese largely through the channels o f literary criticism, it is usually invoked only by scholars, current affairs commentators, and others familiar with metadis­ courses o f contemporary society. People on Awaji don’t use the word nostarujii to discuss what they are feeling and what motivates them to try to revive their puppetry tradition. They refer to the power o f natsukashisa. To use the term nostarujii is to make an imposition on the data, for the act o f “longing for the past” is understood within contemporary critical studies to mean participation in a hegemonic discourse: people using nostarujii in Japanese are likely to be referring to a cultural and political trend in which a fictionalized past is invoked to further a particu­ lar political agenda (usually a conservative one). This interpretation o f nostarujii lacks subtlety and does n o t do justice to the complexity o f “nos­ talgic activities” in Japan. In short, nostarujii and natsukashimu are not simply synonymous. While the nostalgic overtones o f the folklore move­ ment may veil a history o f violence and are undoubtedly great money makers, for many people, connecting with their past through the process of longing constitutes a healing process. In the case o f itinerant puppetry, remembrance means finding a way to situate a painful past within a pressent context. The manipulation o f nostalgic currents (whether conscious or unconscious) can be a way to redress a painful past w ithout simple revisionism.

A stro n g tone o f natsukashim i runs th ro u g h o u t the m otivations for reviving the Awaji p uppetry tradition. The pow er o f the past is a p o ten t force in contem porary Japan, and the surface fashion o f w hat one Japanese called “indigo and unpolished w o o d ” conceals a m ore fundam ental p ro ­ cess o f identity and hegem ony. T he governm ent, for its part, w ould prefer to cultivate and m anipulate a previously m arginalized and oppressed peo­ p le’s retu rn to an idealized past than deal w ith tendencies tow ard social revolution. Such a retu rn posits a shared experience, whereas in fact there was a system based on differences. A t stake in the nostalgic journey for people on Awaji is a re-form ing o f their identity. W hen governm ent m inistries take over and appropriate this discourse, it has a decidedly different face, m aking it necessary to ask, “W hose nostalgic discourse is this, anyway?”

From M edicalization to P oliticization O u r term nostalgia is derived from the G reek gerund nostos, which means simply “returning h o m e.” B ut added to this w ord is the ending with which m ost o f us are all to o familiar, “algia,” which refers to any painful con dition (for example, a neuralgia). So nostalgia refers n o t just to a sense o f pleasantly rem em bering the past b u t also to the charge o f pain which goes w ith it. The term was first used by the Swiss physician Johannes H ofer in the late seventeenth century to describe w hat he considered to be a clinical condition: Swiss soldiers fighting away from hom e becam e very hom esick, and suffered from “despondency, m elancholia, lability o f em o ­ tion, including p rofound bouts o f w eeping, anorexia, a generalized ‘w ast­ ing away,’ and, n o t infrequently, attem pts at suicide.” 9 O th e r clinicians and scientists o f the seventeenth and early eighteenth centuries concerned them selves w ith finding the ro o t and cure for w hat they understood to be a physiological condition. Originally created to label w hat was th o u g h t to be a medical condition, this term was later taken up by psychologists who, th o u g h n o t convinced o f its physiological ro o ts, w ere nevertheless given to regarding nostalgia as an ailm ent. A com m on thread ran th ro u g h u n ­ derstandings o f this condition, the experience o f in terru p tio n . As Fred Davis points o u t, “nearly all theories o f nostalgia, from the m ost m echa­ nistic and physiological to the m ost existential and psychological, draw on som e n o tio n o f sudden alteration, sharp transition, or m arked discontinu­ ity in life experience to explain the p h e n o m e n o n .” 10 Davis has rightly pointed o u t th at the history o f this term has u n d er­ gone transitions from a medical term to a term used largely in psychology and finally to an everyday w ord for b oth simple hom esickness and the

tendency to regard the past with special affection and to desire to return to it in some way or another.11 Davis argues that a shift in meaning o f the term nostalgia has occurred as the notion o f hom e as a physical place has eroded in contemporary society. This insight is particularly relevant in contemporary Japan, where a huge exodus from rural areas into the cities in the past forty years has left some villages empty o f all but the most elderly occupants. These villages have become symbols for a dying way o f life in Japan. “Homesickness” no longer evokes a place but a time. In essence, home has been temporalized; it now refers to a time in a person’s life, the past, rather than to a particular place. As Davis writes, “because . . . home as such can for so many no longer evoke the ‘remembrance o f things past’ it once did, it has fallen to other words, ‘nostalgia’ among them, to comprehend the sometimes pedestrian, sometimes disjunctive, and sometimes eerie sense we carry o f our own past and o f its meaning for present and future.” 12 In Japan, nostalgia now has a new dimension. It has become powerful political capital, capable o f being spent but never exhausted. The term used today refers to a general desire to “return to the good old days,” a pronounced aspect o f popular culture. Tourism, media, and advertising have waged a war on the urban present with destinations, dramas, and images o f a bucolic and thatch-roofed past. This spatial bias o f folklore studies and the nostalgic discourse o f which it is a part have been sum m a­ rized by Stephen Nussbaum: “This concerns the interchangeability o f geography and history, the notion that one may travel back in time by moving spatially toward the hinterland. According to this way o f thought the city is an im portant locus o f cultural change and the entire country becomes a living m useum .” 13 One o f the watershed academic works to directly address the dynamics o f nostalgia is Raymond Williams’s The Country and the City.u His work is an exploration o f the strong tendency in English literature and histori­ ography to posit a longing for a golden age in which the city (and the forms o f complex social organization it both produces and comes to rep­ resent symbolically) was not the dom inant venue o f hum an life. His study reveals that this bias toward the rural as the authentic can be seen as a dynamic throughout English history. Nostalgia for the rural he argues, is an ever present trope in English conceptions o f time and space. We can see a similar tendency in Japanese nostalgic ideologies as well. Recent scholarship in anthropology and sociology has begun to look at the role o f nostalgia as a means o f temporal (rather than spatial) orienta­ tion in ideologies o f value. Usually, nostalgic discourse is identified with the undercurrents o f fascism, because it creates a false sense o f a shared past that is better than the present, and therefore is able to draw current social problems into dramatic relief against a carefully constructed and

filtered past. W hat are the im plications o f the aesthetic appropriation o f the past, and how is the past o f an oppressed people appropriated by their oppressors in nostalgic discourse? Fredric Jam eson in his article “ Postm odernism , o r the C ultural Logic o f Late Capitalism ,” 15 an often cited w ork in the deconstruction o f nostal­ gic discourse, points to num erous examples in m odern art, film, and archi­ tecture in which various “signs” from the past are appropriated, w ith no concerted effort to find a com plete context for such appropriations. The aesthetic result he calls a pastiche, a patchw ork o f unrelated objects, set side by side as if to call atten tio n to (and, I w ould add, even fetishize) them selves. For Jam eson, this is the aesthetic m ode o f postm odernism . It is n o t m erely a question o f a particular aesthetic decision, one choice am ong many, b u t rather, an expression o f the tim e itself: “I cannot stress to o greatly the radical distinction betw een a view for which the p o stm o d ­ ern is one (optional) style am ong m any others available, and one which seeks to grasp it as the cultural dom inant o f the logic o f late capitalism: the tw o approaches in fact generate tw o very different ways o f conceptualizing the phenom enon as a w hole, on the one hand m oral judgm ents (about which it is indifferent w hether they are positive or negative), and on the o th e r a genuinely dialectical attem p t to think o u r present tim e in H is­ tory.” 16 This tendency to appropriate the past into a collage o f images is w hat Jam eson calls “the nostalgic m o d e.” We need to ask a question th at Jam eson only begins to address: Is n o s­ talgia as we see it today (in, for example, the revival o f the Awaji tradition) a unique feature o f this particular m om ent in history? I f so, does it reveal som ething about the late tw entieth century in Japan? I f n o t, w hat can it tell us? I f we read Jam eson correctly, it appears he w ould argue yes to the first question. T he “nostalgic m o d e” is “a desperate attem pt to appro­ priate a missing p a st.” Since this has been projected o n to a larger social m ovem ent (as in th e current conservative nostalgia for the 1950s in A m er­ ica, for example, o r in Japan’s nostalgia for the late E do period and the rural), it “is refracted th ro u g h the iron law o f fashion.” 17 N ostalgic m odes o f discourse are attem pts to reclaim th at w hich never really was. For Jam eson, and perhaps on this p o in t he can be said to speak for p o stm o d ­ ernists as a w hole, a nostalgic enterprise ultim ately reveals a crisis in our relationship to the entire issue o f tem porality. We have com e to realize th a t the past as a know able unit does n o t exist, b u t is forever subject to the herm eneutics o f suspicion. We are, simply, unable to rely on a logical tem poral construction as a m eans o f orientation. T he subject, Jam eson writes, “ has lost its capacity actively to extend its pro-tensions and re­ tensions across th e tem poral m anifold, and to organize its past and future in to coherent experience, it becom es difficult enough to see how the cultural productions o f such a subject could result in anything b u t ‘heaps

o f fragm ents’ and in a practice o f the random ly heterogeneous and frag­ m entary and the aleatory.” 18 Fred Davis said basically the same thing a num ber o f years earlier, w ith ­ o u t the “pastiche” o f semiotics and psychoanalysis: If, as I have m aintained, nostalgia is a distinctive way, th o u g h only one am ong several ways we have, o f relating o u r past to o u r present and future, it follows th a t nostalgia (like long term m em ory, like rem iniscence, like daydream ing) is deeply im plicated in the sense o f w ho we are, w hat we care ab o u t, and (th o u g h possibly w ith m uch less inner clarity) w hither we go. In sh o rt, nostalgia is one o f the means— or, better, one o f th e m ore readily accessible psychological lenses—we em ploy in the never ending w ork o f constructing, m aintaining, and reconstructing o u r identities. To carry the optical m etaphor a step further, it can be th o u g h t o f as a kind o f telep h o to lens o n life, w hich, while it magnifies and prettifies som e segm ents o f o u r past, sim ultaneously blurs and grays other segm ents, typically those closer to us in tim e.19

Jam eson sees pastiche in its nostalgic m ode as som ething o f a schizo­ phrenic endeavor. Basing his discussion o f schizophrenia on L acan’s ac­ coun t, he writes th a t w hen Lacan’s category is taken descriptively (rather than clinically), it is useful for understanding the postm odern m om ent. W hile n o t including all postm odern artists u nder the label o f schizo­ phrenic, he considers schizophrenia an illum inating aesthetic category. Very briefly, Lacan describes schizophrenia as a breakdow n in the signifying chain, th at is, the interlocking syntagm atic series o f signifiers which constitutes an utterance o r a m eaning. . . . T he connection betw een this kind o f linguistic m alfunction and the psyche o f the schizophrenic may then be grasped by way o f a tw o-fold proposition: first th e personal identity is itself the effect o f a cer­ tain tem poral unification o f past and future w ith the present before m e; and second, th at such active tem poral unification is itself a function o f language, or b etter still o f the sentence, as it moves along its herm eneutical circle through time. I f we are unable to unify the past, present and future in th e sentence, then we are similarly unable to unify the past, present and future o f o u r ow n b io ­ graphical experience or psychic life.20

Jam eson argues th at the prevalence o f nostalgic discourse is due to the crisis in the experience o f tim e and history as a way o f organizing m eaning, a crisis th at he refers to as “the breakdow n o f tem porality.” 21 W hile a person experiencing this sense o f crisis w ould probably find that expression alien to his or her experience, it suggests a m o m en t w hen the past seems alien in the present, creating a crisis concerning the sig­ nificance o f b o th the present and the future. I f I cannot identify or even locate the substance o f my “rem em brance o f things past,” how can I have an identity in the present? I f I have no identity, w hat can guarantee that

the future will n o t be even m ore chaotic? W hat, th en , can be extracted and even serve as icons from my past, icons th a t can com e to stand for my entire lost self) O ur case provides a field in which to explore the complexity o f w hat Jam eson refers to as “ the nostalgic m o d e ,” for unlike his cases in which pastiche is the d om inant cultural elem ent, we are looking at a case in which the revival o f a tradition grows o u t o f itself. The “ heaps o f frag­ m ents” o f the Awaji tradition are being carefully rebuilt into a new picture o f the trad itio n , and this construction can claim to be “retrieving” its ow n past. O n Awaji— unlike p ostm o d ern examples, drawn from m iddle A m er­ ica w here French bistro ads are placed side by side w ith G reek sculpture, Am erican beer signs, and English china— puppets are retrieved from an Awaji past and placed into an Awaji present, and frequently returned to the spatial contexts w here they once were com m on. P uppet perform ances o f Shiki Sanbaso and Ebisu-m ai in th e 1990s on docks and in rice paddies are good examples. T he pastiche, th en , is n o t com prised o f prom iscuous cultural borrow ing; it is a retrieval o f the past into a different context— the here and now. To place the past back into its “authentic” context implies th at tim e m ust rem ain static in o rd er to be authentic. T he revival o f Awaji puppets is n o t w ith o u t a serious dose o f kitsch. Elem ents o f the tradition have been singled o u t and decontextualized as icons for the entire past. In the theater gift shop, one can buy sake in a bottle shaped like a pu p p et head. Fans in the shape o f fam ous characters from the jo ru ri pieces are also on sale, and Awaji pu p p et telephone cards are big sellers. Tea towels depicting scenes from Keisei A w a no N aruto are popular gifts, and sweets stam ped w ith the raging head o f FIachiman Tar5 are the fastest m oving item s on the culinary side. B ut th e entirety o f this revival is n o t simply ab o u t m arket forces and m aking tourism viable as a base for regional economy. R ather, I suggest th at the semiotics are infinitely m ore m ultivalent. W ho owns this nostalgia? Can the nostalgic experience o f the young m an on the bicycle be described in the same way as th a t o f the ethnographer from Kanagawa w ho studies Awaji, o r the m iddle-aged m an from Tokyo w ho travels to Awaji to see the theater to reclaim his identity as a Japanese? W hen U m azum e M asaru reinvents Atvaji ningyo as the director o f the new theater, he is n o t random ly bor­ row ing the past in a collage o f unrelated images. H e is consciously select­ ing elem ents o f his tradition th a t he considers viable in a late-tw entiethcentury context, and which he sees can becom e icons for the entirety o f a lost tradition and focal points for its future. T he reader will recall an im p o rtan t distinction w ithin the Awaji trad i­ tio n betw een the joruri perform ances, intended for hum an audiences and having a “secular” agenda, and the kadozuke perform ances, presented on ritual occasions. In the process o f revival, a decision, at tim es only half-

conscious, was m ade to present the w hole o f the tradition as the joruri tradition. U m azum e explained th at was done largely because this p art o f the tradition was m ost likely to be viable. This decision seems rational. Just as no one w ould claim th at staging Shakespeare today constitutes a simple and naive “ retu rn to the past,” so to o the jo ru ri ballads contain m otifs, issues, and ideas which, th o u g h w rapped in the history o f their tim es, can speak to people o f any generation and even different cultures. T hem es o f love, divided loyalties, betrayal, jealousy, m urder, and incest are the stu ff o f timeless pieces o f dram a. T h at this tradition is being revital­ ized should n o t be a cause for puzzlem ent, since the joruri tradition has m aintained a continuing appeal, although at a m ore urban and formal level. T he presentation o f jbruri pieces on the Kabuki and Bunraku stages has continued with little in terru p tio n . Reviving these perform ances on Awaji affects the co n ten t o f the plays less than w ho perform s them , where they are perform ed, and w ith w hat aesthetic and theatrical sensibilities. T he key issue is this: the jbruri tradition initially was lim ited to one piece, and this single piece becam e an icon for a vanishing past. T he decision to make the perform ance tradition (joruri) o f Awaji puppets stand for the entire “ lost trad itio n ” in the early stages (1950s th ro u g h 1970s) and then to shift to the kadozuke aspect o f the tradition in the latter stages (late 1980s to the present) reflects the shifting hierar­ chy o f sym bolization, responding to subtle changes in the nativist dis­ course in contem porary Japan. T he elem ent o f “postm odern schizophre­ nia” in this retrieval activity becomes m ost apparent w hen we explore how the Awaji puppet tradition is being placed w ithin a larger context o f to u r­ ism on the island. Let us look closely at just which parts o f the tradition bore this burden o f representing the entire tradition and how those as­ pects o f tradition get w edded to o th er regional features to represent “Awaji as a state o f m in d .”

Puppets and Whirlpools: Icons of the Vanishing In the process o f revival, Awaji puppets have been the w edded to a larger process o f regionalism. In the 1970s and 1980s, the tourist industry in Japan launched two very successful campaigns to p rom ote rural Japan as exotic and yet a place w here people could discover som ething authentic in their identity as a Japanese.22 Jisukahaa Japan (Discover Japan) and Ekusochikku Japan (Exotic Japan), b o th products o f the Japan T ourist Bureau, cast the retu rn to a mythical past in rural Japan— an escape to the roots, as it were— in foreign w ords, a linguistic decision designed to u n ­ derscore the possibility o f an experience o f the exotic and even the ethnic right here at hom e. Posters show, for example, a young lady standing in a

dense forest near a m ountain tem ple; she watches while an ascetic walks away into the hill. Im ages o f m asked perform ances and traditional arts m ade room in the posters for the solitary (usually female) traveler in tim e and space. O ne could “get real” and never leave Japan. W hile it is clear these cam paigns have created a discourse o f their own a b o u t exoticism , I argue th at they reflect a deeper cultural process at work, a regionalism which has its roots in the folklore m ovem ent o f Yanagita K unio. T he success or failure o f any project to pro m o te regionalism depends on the p ro d u ctio n o f a visible icon th at can sym bolize a reg io n ’s past by standing for its vanishing, endangered, or lost realities and values, and th at advertises the region and its com plex system o f m eanings. By icon, I refer to a single im age serving this m ultiple capacity— flashy, eye catching, readily rem em bered, and easily reproduced. H ere, I exam ine the creation o f the “iconography” o f Awaji regionalism and the role o f pu p p etry in th at process. In the early stages o f the revival o f Awaji puppets (and later as well), the pilgrim age scene from Keisei A w a no N a ru to 1 a m ere tw enty-eight m in­ utes long, was presented as th e only surviving piece o f th e Awaji reper­ toire. This play com bined w ith elem ents o f the Awaji landscape to assume the role o f an icon in the process o f revival. D uring the 1960s the stru g ­ gling theater above the gift shop did n o t even have a regular chanter, so som etim es the piece was presented accom panied by a recording o f a tayu chanting this tragic joruri tale o f a m other and daughter encountering one another after a separation o f several years. This single scene has been p re ­ sented several tim es every day for decades, and until quite recently this was all th at Awaji pu p p etry as a revived theater consisted of. I suggest th at this piece is best u n d ersto o d as an icon n o t only for th e revival m ovem ent b u t for the larger m ovem ent to create Awaji as a new M editerranean in Japan, th at is, as a carefree place far from th e m adding crow d o f the city, the new vacation spot for urban Japan. B ut why is this piece such a pow erful icon linking Awaji puppets w ith Awaji regionalism? In the scene, a w om an is sitting in her hom e w hen she hears a pilgrim arrive at her door. From the h at the young pilgrim is carrying, we know she is on the Saikoku Pilgrim age.23 T he pilgrim , it turns o u t, is a m ere child. T he w om an, surprised to see such a young child alone on a pilgrim age, asks her w here her parents are. T h e child replies, “I am going in search o f my parents. I am from the province o f Awa, from Tokushim a, and was separated from my parents at the age o f three w hen I was left to be raised by my grandm other. Now, I have com e looking for my parents. I w ant to m eet them . I w ant to see th e m .” (In Tokugawa Japan, w hen m obility was highly restricted, people frequently used pilgrim age as the pretext to conduct business— family or otherw ise— which required traveling.)

T he m o th e r in ICeisei A w a no N aru to looks longingly for her child after she has sent h er away. C ourtesy o f F ujim oto Yasur5.

T h e m other realizes th a t this is her ow n daughter, w hom she and her husband had to abandon several years before w hen they learned o f the danger they were in due to her h u sb an d ’s political affiliations. N ow , her ow n child has com e begging at her door. T he m other struggles over w hether or n o t she should reveal her identity to the child, b u t, realizing th at to do so w ould endanger the girl, she decides against it. N evertheless, she pulls her inside th e d o o r to get a closer look at her. She looks for, and finds, a telltale mark on the child’s forehead identifying her as her d a u g h ­ ter. T he rem ainder o f the play is a tragic enactm ent o f this m o th er’s agony as she holds her vulnerable child close to her and hears how the child is being abused on her travels. It ends w hen she tries to give th e child m oney and sends her away. The m other is left alone on stage and lam ents the

separation and the painful realities that have made this necessary. The scene closes as the m other decides to go looking for the child. In the next act (which is rarely perform ed, and is never presented at the Awaji theater in everyday performances) after leaving the person who is her m other, the child encounters her father on the road. Unaware that this is his daughter, he tries to rob her. She tries to get away from his grasp, and in the ensuing struggle he accidentally kills her. N ot recogniz­ ing his own child, he brings the body to his home and lays it o u t on the tatami. His wife arrives hom e and tells him that their child has come. She describes a young pilgrim, the person he has just robbed and killed. When he realizes what he has done, he tries to convince his wife that this dead form to his right is actually only sleeping and begs her not to wake the child. W hen the wife discovers her child is dead, she goes crazy. H er hus­ band delivers a famous soliloquy in joruri lamenting his fate as the m ur­ derer o f his own child and the strange twists o f fate that have led him to this tragic mom ent. This act gives the play its real power as theater. Since the entire piece is so deeply affecting, why has this latter part o f the play been left out o f the revival process?24 Quite simply, the sheer weight o f this tragic scene makes it too heavy to bear the burden o f the iconographic process. The new image o f Awaji regionalism is o f a carefree, rural place, lighthearted and perfect for a vacation. Awaji regionalism is decidedly n o t about myths and epics o f Oedipal proportions. The gravity o f Awaji puppetry as drama is thus made subordinate to the revival process, a fact not lost on many eld­ erly people on Awaji who feel that the tradition is n o t being retrieved but merely fetishized into one short piece. A num ber o f logistical reasons justify the decision to use the pilgrim­ age scene from Keisei A w a no N aruto as the sole performance. First, the piece requires only two puppets, one o f whom is a child, which means that it can be manipulated by two people (even one in a pinch) rather than three. In all, that means that five puppeteers can present the piece. Sec­ ond, the scene can stand as a playlet in its own right, and people not familiar with the longer play can be deeply moved by the pathos o f the tragedy o f a m other and child. Third, it is famous, and people have often heard something about it, even if they know little about the classical joruri tradition. And fourth, since there are only two voices and both o f them are female, it does not require a chanter with an extensive range to prepare the piece and therefore serves as a starter piece for chanters to learn the art o f recitation. But a deeper logic was at work in selecting Keisei A w a no Naruto as the lifeline rescuing the tradition from extinction. The piece participates in the Japanese nostalgic discourse o f regionalism. In spite o f the fact that

regional variation in language, food, ritual performance styles, and cus­ tom have been eroded dramatically through rapid urbanization and mass media, the nostalgic discourse insists that if one retreats from the cities to the rural areas, one will find these regional distinctions still intact. But the regionalism must be made so apparent that it will jum p out at you from a rapidly moving tour bus. Keisei A w a no Naruto works as an icon for the Awaji tradition because it is a regional piece. People will recognize the place-name in the title and associate it with Awaji. The title refers to the N aruto Straits, a location literally within a stone’s throw o f the theater. But the title also suggests that human life (in Tokugawa Japan at least) is n o t unlike the geological squeeze created by the proximity o f Shikoku and Awaji. After watching The Tragic Straits o f N aruto, one could board a boat and go out to see the whirlpools o f N aruto firsthand. The play, then, was one part o f a packaged deal. Awaji regionalism meant puppets and whirlpools, and when the two could be superimposed upon one another, all the better. The theater over the gift shop at Fukura Bay was just across a narrow street from the harbor where one boarded the boat to go and see the whirlpools. In fact, one could buy a ticket for both the boat and the theater from the same little booth. While the whirlpools captured the imagination because they ap­ pear so dangerous (and actually are), the puppets also appeal to the h u n ­ ger for the rustic (and implicitly, the wildness that rusticity implies). Awaji puppets were seen as the rustic but real origins o f the more refined Osaka Bunraku puppets. Both the puppets and the whirlpools, the analogical construction seems to suggest, have a natural innate wildness. The two icons reinforce one another. When I went back to Awaji in 1984,1 was surprised to discover that the pilgrimage scene o f Keisei Aw a no N aruto, was the only piece being per­ formed. Still in its location over the gift shop, the troupe was starting to attract young members from around Awaji who were trained in high school clubs. The troupe had just returned from a trip to H olland and Belgium. Awaji puppets were starting to make a comeback, but even in their local setting their performances were limited to this one scene. In 1987 when I returned to Awaji to conduct extensive field work over a two-year period, the troupe had moved to its present location on the cape overlooking the N aruto Straits, and even though it now had govern­ m ent subsidy and a fancy theater, the pilgrimage scene o f Keisei A w a no Naruto was still the troupe’s only offering. The new theater is in a larger building called the N aruto Kaikan (N aruto Hall), in which the icons o f Awaji regionalism are carefully displayed. A small movie and natural sci­ ence display hall explains the whirlpools o f the N aruto straits. After look­ ing at the natural science exhibit, one enters a movie hall, where one can watch (with 3-D glasses) a film o f the whirlpools and get the feeling o f

actually being in them . Fish pulled from the w hirlpools seem to jum p rig h t into your face. O th e r seasonal events from the Awaji ritual calendar pop o ff the screen at you in three-dim ensional clarity. The N a ru to Kaikan also has a gift shop selling sweets w ith th e m otifs o f w hirlpools and puppets as well as regional delicacies such as N a ru to wakam e (seaweed). In the pu p p et theater, a fancy brocade curtain decorates the stage. O n it are depicted (it should com e as n o surprise) th e m o th e r and dau g h ter from Keisei A w a no N a ru to against a backdrop o f w hirlpools. T he semiotics had n o t changed, b u t the packaging had becom e m ore slick. Instead o f w hirlpools accessible by boat, they are visually accessible. O ne can climb up to the viewing deck and look over the N a ru to Straits th ro u g h b in­ oculars (for 100 yen). T he iconography is the same: puppets and w hirl­ pools, b o th standing for Awaji as a w hole, as a rural place, a p a rt o f Japan’s vanishing past. In 1 9 8 7 ,1 asked m em bers o f the tro u p e how they felt ab o u t p e rfo rm ­ ing th e pilgrim age scene from Keisei A w a no N aruto several tim es a day, day in and day out. Every one o f them expressed feelings ranging from b oredom to rage at th e redundancy. N o one denied it was a pow erful piece, b u t they resented the fact th a t the th eater had to stick to this one piece for public perform ances. T he tro u p e was actively involved in practic­ ing o th er pieces, and even presented them occasionally at local events a round Awaji, b u t for the m ost p a rt the daily fare at the Awaji N ingyd Joruri Kan was Keisei A w a no N aruto. I asked U m azum e-san why they did n o t expand the perform ance repertoire. H is answer was telling: “W hen we first started to present th e plays at th e previous location, we w anted som e­ th in g people liked. A lthough m any people w ere familiar w ith b o th the larger piece from w hich this scene is taken and also the larger repertoire o f joruri pieces, this was a favorite, and a regional piece referring to lower Awaji and Tokushim a. B ut eventually, it g o t to the p o in t w here people w ould only com e to see th at piece. It is the only piece o f th eater they associate w ith Awaji ningyo. Awaji ningyo m eans Keisei A w a no N aruto these days. I f a to u r bus calls to book a show ing, they request th a t piece. W henever I suggest th a t perhaps the perform ance m ight change, the to u r­ ist in dustry becom es uncom fortable.” 25 T he decision to make Keisei A w a no N aruto the iconographic center­ piece o f the revival m ovem ent was an act o f inclusion. M eanw hile there was a parallel act o f exclusion. D u rin g my fieldwork, I was interested in the kadozuke aspect o f the Awaji tradition, decidedly absent from the new, “iconographically correct” situation. I was interested in how the reli­ gious past o f the tradition w ould be understo od, ignored, or appropriated as the tradition o f Awaji ningyo was being revived. T he puppets for the Sanbaso perform ance were kept in a small case to the left o f the stage, on display, b u t the perform ance was rarely presented and hardly ever dis-

cussed. It was clear this ritual piece could not be used as an icon for the Awaji tradition (even though it is a highly distinctive aspect o f it, in fact more typical o f Awaji ningyo’s past than the joruri tradition), so it was relegated to a glass case. Even within the revived theater, some aspects o f the tradition are acknowledged as part o f the past, while others (puppets and whirlpools in Keisei Awa no N aruto) are part o f its present and future. As o f the early 1980s, ritual puppetry had quite simply lost its context as performance and become a museum object. The glass case, it can be sug­ gested, was the symbolic space in which kadozuke puppets still made sense—as relics o f the past. Perhaps another dynamic was at work in this tendency to de-emphasize the kadozuke tradition and showcase the joruri aspects o f the tradition. Renato Rosaldo suggests that our assumption that nostalgia is something com m on to human beings at all times is highly suspect. Using Davis as his sole source for the history o f the term , he argues that because the term itself has undergone such variation since its invention, nostalgia can­ n ot be seen as a constant: “the changing meanings o f ‘nostalgia’ in W est­ ern Europe (not to m ention that some cultures have no such concept at all) indicate that ‘o u r’ feelings o f tender yearning are neither as natural nor as pan-hum an, and therefore not necessarily as innocent, as one m ight imagine.”26 Rosaldo notes the role o f nostalgia in the writing o f ethnogra­ phies (and as it appears in contem porary films about the colonial period), and claims many ethnographic enterprises are the result o f the yearning by certain people (ethnographers, missionaries) for the “traditional” cul­ tures they have destroyed. H e calls this “imperialist nostalgia,” and de­ scribes it as follows: Curiously enough, agents o f colonialism— officials, constabulary officers, mis­ sionaries, and other figures from whom anthropologists ritually dissociate themselves— often display nostalgia for the colonized culture as it was “tradi­ tionally” (that is, when they first encountered it). The peculiarity o f their yearning, o f course, is that agents o f colonialism long for the very forms they intentionally altered or destroyed. Therefore, my concern resides with a partic­ ular kind o f nostalgia, often found under imperialism, where people mourn the passing o f what they themselves have transformed. Imperialist nostalgia re­ volves around a paradox: A person kills somebody, then mourns the victim. In any o f its versions, imperialist nostalgia uses a pose o f “innocent yearning” both to capture people's imaginations and to conceal its complicity with often brutal dom ination.27

Few ethnographers in Japan would step forward and confess to having m urdered folk performing arts in Japan, and arguing such would be per­ haps taking Rosaldo further than he would want to go. Nevertheless, this insight into the use o f nostalgia to veil violence and dom ination applies to

our case in an interesting way. The nostalgic enterprise on Awaji, which is rapidly creating a m odern context for ritual puppetry as a showcased cul­ tural perform ance, has almost com pletely co-opted the participation o f children o f ritual puppeteers in the burakum in civil rights movement. As I have suggested, a decision seems to have been made in the early stages o f the revival m ovem ent o f Awaji puppets that the jbruri aspect o f the Awaji tradition should be m aintained and resurrected, while the kadozuke origins should be noted b u t n o t fostered. In the 1970s and early 80s, kadozuke materials were showcased in such a way th a t they appeared to be a part o f Awaji ningyri’s very distant past, b u t the m ore recent exam­ ples o f ritual activity were neglected altogether. For evidence o f this, we can cite b u t a few examples. First, joruri material only was presented on all foreign tours. Sanbaso ritual was n o t presented even w ithin Japan, where the ritual tradition stood some chance o f being recognizable to audiences. Second, in spite o f the fact th at Sanbaso puppet heads were the m ost com ­ m on o f kashira (because there were so many itinerant perform ers), m ore o f these kashira than joruri heads were actually lost or destroyed. Third, while younger puppeteers were trained in the recitation o f joruri pieces, it was n o t until the late 1980s th a t these puppeteers even to o k an interest in learning the Sanbaso pieces. As late as 1987, none o f the younger pu p p e­ teers in the Awaji theater knew the story o f H yakudayu, the “founder” o f ritual puppetry, although they were well versed in the histories o f the var­ ious dramatic troupes on the island. I frequently was in the embarrassing situation o f know ing m ore about the ritual history o f Awaji ningyo than the young perform ers in the theater, w ho were often am azed to learn that itinerant perform ers did m ore than present “skits” for entertainm ent. The irony o f this de-emphasis on the ritual tradition is th at while the joruri tradition is based on w ritten texts and is relatively easy to transm it from one generation to the next, th e kadozuke tradition is m ost vulnerable in the transmission process because so m uch o f it is based on oral transm is­ sion. Many o f the variations in kadozuke perform ance were simply never recorded and have been lost forever just in the past few years, w hen they could have been preserved. This fact led me to conclude th at som ething beyond a task o f retrieval was at work in this revival m ovem ent. I f preser­ vation and retrieval were th e task, why n o t immediately focus on that which is m ost vulnerable to time? These simple points suggest th at in the early stages o f revival, the ka­ dozuke tradition w ith its ritual perform ances was deem ed w anting as an appropriate icon for a lost past. A nd as a result, an im portant thread in the history o f this tradition was eclipsed. This decision to ignore the kadozuke tradition was largely based on the perception th a t this itinerant perfor­ mance was som ehow an em barrassm ent to a m odern Japan. It looked like begging and was inextricably linked to popular religion. Furtherm ore, it

was the very itinerancy o f the perform ance th at had co n trib u ted to the outcast status o f the perform ers. As we have seen in past chapters, ritual puppeteers were regarded as b o th polluted and outcast in society before the tw entieth century. Today the struggle continues against this hereditary stigm a. P uppet theaters w hich presented joruri also suffered from this social stigm a because they were in the perform ing arts and were itinerant. In the postw ar period, and particularly during the occupation, as theaters began to reopen around Japan, puppetry troupes were often the last to receive funds and perm is­ sion to reopen their doors. D iscrim ination against these groups co n tin ­ ued, and people were encouraged to discard the practices o f ritual p er­ form ances, because they were an em barrassm ent to Japan th a t was trying to p u t on a m odern face and distance itself from any overt signs o f “super­ stitio n ” and “folk religiosity.” In the postw ar period, m oreover, the con­ text for popular “folk” perform ance form s, p u p p etry included, was dra­ matically eroded by three factors: the censorship o f the occupation forces over the presentation o f plays with feudal them es ( a category in to which m any o f the joruri ballad pieces fell), the w idespread appearance o f movie theaters around Japan, and a rapid urbanization o f the population. By the end o f the 1950s television was added to this list o f factors which had a severe effect on the survival o f popular perform ance forms. This ban on “ the traditional” was n o t unilateral. Tea cerem ony, for example, a highly elite a rt form practiced only by a select g roup o f people w ith significant access to w ealth, was held up to th e w orld (and the occu­ pying forces) in postw ar Japan as p ro o f o f the cultivation o f the Japanese people. A carefully calculated decision was m ade a b o u t just w ho and w hat could represent Japanese cultural arts, and outcast arts were deem ed an em barrassm ent. B oth fashion and funding underscored this attitude, and as a result, ritual p uppetry quickly came to an end. T he reader will recall th a t at the end o f ou r discussion o f kadozuke perform ance in chapter 2 we n o ted th at m any o f the postw ar cam paigns to wipe o u t itinerant perform ance actually came from the people w ithin those perform ance traditions. T heir reasons im plicated them in the cycle o f oppression and were probably in p art the result o f som e kind o f inter­ nalized oppression. Because the practice o f p u p p e try had m ade it im possi­ ble for people to escape from hereditary pariah status during the Tokugawa period, any indications th at this practice m ight continue in to the postw ar period needed to be wiped o u t, lest they com prom ise th e hope for a viable liberation m ovem ent. By showcasing the joruri tradition on Awaji, it was possible for Awaji puppets to participate in a “geography o f value” in Japan th a t suggested th at the rural is the authentic, the “folk” is the real. This authenticism is best captured by the claim (possibly tru e) on Awaji th at Bunraku in Osaka

was started by an Awaji p u p p eteer nam ed U em u ra B unrakuken. Awaji p u p p ets, th e n , are to B unraku p u p p ets w hat ancestors are to those o f us in the m o d ern p eriod— o u r tru e selves, less sophisticated, dead, b u t partici­ patin g in a reality th a t we m oderns can only approxim ate. Regionalism for Awaji has m ean t a careful assessm ent o fA w aji’s history as a cen ter o f ritual puppetry. O u tcast puppeteers p resenting ritual purifi­ cation rites and perform ances for a tem p estu o u s and even dangerous deity w ho som etim es causes epidem ics d o n o t m ake good icons. M otherd a u g h te r tragedies do (while fath er-d au g h ter m u rd e r stories do n o t). B ut if this set o f priorities w ere to be tu rn e d on its head, as it has been in the past five years, w hat w ould th a t m ean and w hat w ould it say a b o u t the local Awaji ap p ropriation o f nostalgic discourse and regionalism for its ow n ends?

R eviving Kadozuke: From Ritual to Stage A ro u n d 1990, th ere was a shift in th e revival m ovem ent to h ighlight the ritual trad itio n o f th e Ebisu and Sanbaso perform ances as th e distinctive features o f the Awaji trad itio n . Perhaps this shift was due to the fact th a t the revived trad itio n o f Awaji pu p p etry was now stable en o u g h to begin enlarging its rep erto ire and retrieving larger aspects o f its past. B ut I argue th a t the sudden in terest in the kadozuke tradition was in p a rt a response to a new thread in nativist and nostalgic discourse in Japan, and had a redressive function in th e social dram a o f extinction, in terv en tio n , revival and renew al o f Awaji puppetry. A ccording to this subtle shift in nativist discourse, real capital for revival was now found in those traditions th at could in som e way lay claim to exoticism . T h e m ore esoteric and e n ig ­ m atic a tradition was, the greater its value as a cultural com m odity. R itu ­ als, by virtue o f th eir heavy use o f m ultivalent (and frequently u n intelligi­ ble) sym bols, are alm ost by definition exotic and enigm atic. M arilyn Ivy, in h e r study o f nostalgic discourse in c o n tem p o rary Japan, rightly points o u t th a t a p a rt o f this process o f a “ re tu rn ” to traditional Japan lies in m aking Japan seem o th e r to itself.28 This insight raises the qu estion o f how the exotic and strange can be m anipulated as b o th a sym bol and a com m odity in th e process o f revitalization. H o w can Awaji puppets be m ade to appear o th e r while still rem aining Japanese? O th ­ erness, as we have seen, is so m eth in g ritual puppeteers know a lo t about. B ut in this new discourse o f exoticism , otherness does n o t place ritual p u p p e try outside o f Japanese m ainstream society. O therness is w hat m akes it tru ly Japanese! T h e Sanbasb rite (to a larger degree than the Ebisu rite) was suddenly m ade th e central concern o f people in th e th e a ­ ter. W hile the daily perform ances co n tin u ed to present a slightly w ider

repertoire o f jdruri pieces, scholars and m em bers o f the th eater w ere su d ­ denly obsessed w ith retrieving the Sanbaso perform ance. My research on this subject was suddenly o f great interest. T he irony is th at if retrieval alone had been the actual goal o f this revival m ovem ent, the attem p t came ab o u t ten to fifteen years to o late, even th o u g h in the 1970s people had been well aware o f how vulnerable this part o f the tra d i­ tio n was to the vicissitudes o f tim e. A t th at tim e, it simply had n o t seem ed th at im p o rtan t or valuable. Basically, the kadozuke tradition becam e the focus o f “reen actm en t” activities, and was also the centerpiece o f a series o f sum m it m eetings for the revival o f Japanese puppetry, organized by U m azum e M asaru and d e ­ signed to further the revival m ovem ents o f p u p p et theaters all over Japan.

ccA Past of Things Present;” Reenacting R itu a l Puppetry in its Traditional Contexts In July o f 1990, tw o self-consciously “ historic” perform ances to o k place on Awaji, ab o u t tw o days apart.29 T he puppeteers from the Awaji N ingyb Joruri Kan did som ething th a t had n o t been done in over forty years: they perform ed the rite o f Ebisu on the small island called Susaki in Fukura, and the Sanbaso rite in the precincts o f th e O m ido H achim an shrine, the cerem onial center o f the Awaji p u p p et tradition. Now, m em bers o f the new and revived theater on the island, the Awaji N ingyb Joruri Kan, were attem pting to recreate these rituals by retu rn in g to the sites w here they had often been p resented in the prew ar period. These perform ances m arked the beginning o f a new dim ension in the revitalization m ovem ent. Ritual perform ances, and n o t ju st joruri pieces, w ere now being consid­ ered as objects o f retrieval. T he central activity o f these perform ances is “reen actm en t.” Examples o f reenacted rituals, frequently d o cto red up and greatly transform ed to respond to the dem ands o f new m edia (film o r video, stage, television), new audiences (anthropologists, foreign tourists), and new aesthetic sen­ sibilities are com m on in the history o f anthropological discourse. M any o f the docum ents we have for the study o f o th e r cultures are based on the careful staging and scripting by scholars concerned to make a certain point. T he portraits o f native Am ericans in cerem onial garb (and the w om en frequently h alf nude) by Edw ard Curtiss are am ong o u r earliest examples. C urtiss’s case reveals how the camera lens constructs an image o f the o th er th at reinforces and legitim ates the violence o f dom ination, particularly when a do m in an t (and in the case o f the conquest o f America, genocidal) pow er holds the camera. This creation o f noble savages is p art and parcel o f reenactm ent as a trope o f the anthropological enterprise.

The transform ation in reenactm ent responding to the new presence o f the camera lens is n o t always inaugurated by foreigners. G regory Bateson and M argaret M ead’s 1938 film Trance a n d Dance in Bali underscores how people w ithin a tradition can feel the need to transform a ritual for the sake o f a new audience, even while rejecting w hat could be considered authentic w ithin their own tradition for the purposes o f the new context. T he Ballet Folklorico in M exico w ith its stage presentation o f th e deer dance is another example o f rituals being made into stage performances, and eventually com ing to have lives o f their ow n as pieces o f theater di­ vorced from a ritual context. Richard Schechner, in Between Theater a n d Anthropology, explores the process o f retrieval, reconstruction, and staging o f rituals, often wrenched from their traditional context and perform ed for tourists mainly for money. Some o f the cases he cites are quite tragic, as is the case o f the “m udm en” dancers from Papua New Guinea, w ho invented dances for tourist buses, b u t saw only 10 percent o f the m oney they generated.30 W hen Schechner views such cases from the standpoint o f a theater direc­ tor (which he is as well as a scholar), he is inclined to accept this trans­ form ation o f rituals into theater as part o f the m odernization process, implying that a shift from ritual to theater is the natural progression o f all perform ance events. T he dual im pact o f intervention (by the camera lens, the anthropologist, or the postm odern situation itself) and invention (o f tradition) is simply an event in the w orld o f perform ance: I see nothing amiss in restorations o f behavior [like Bharatanatyam and Purulia Chhau]. Arts and rituals . . . are always developing, and restoration is one means o f change. What happened [in Bharatanatyam and Chhau] is analogous to what the French dramatists o f the seventeenth century did when they conform ed to what they thou ght were ancient rules o f Greek tragedy. The dramatists had at hand Aristotle, H orace, the Greek and Latin playtexts, archi­ tectural ruins, pottery, but they did n ot have the actual behaviors o f the an­ cient Athenians. The restorers [ o f Bharatanatyam and Chhau] had living arts they had presumed were vestiges o f older, more classical arts. They also had ancient texts, sculptings, and their ow n deep know ledge o f Hindu traditions.” 31

The decision to enact ritual puppetry as stage perform ance could be seen as an example o f w hat Schechner calls “restored behavior.” Trans­ form ing a ritual o n to the confines o f a stage by definition implies an audi­ ence, a new feature ofperform ance from the perspective ofitinerant ritual. T hough Schechner claims th at there may be “noth in g amiss” in this phenom enon, this does n o t lessen our need to analyze the retrieval and staging process and determ ine how im portant issues in a critical history o f ritual puppetry are addressed, ignored, or even enhanced.

In th e case o f Awaji puppets, we can explore the sem antics o f reenact­ m ent by exam ining w hat happens to the following aspects o f the ritual: ( I ) the m arking o f the perform er as an outsider, an issue o f central m ean­ ing in these rites; (2) the creation o f ritual space and the relationship o f th a t perform ance space to the unknow n space o f the road, to which the p u ppeteer m ust by definition re tu rn after his perform ance; (3) the rela­ tionship o f the audience to the perform ers and vice versa; and (4) the im p o rta n t com m itm ent o f the audience to th e “o u tco m e ” o f th e perfo r­ m ance. H o w does the idea o f the “ agent” o f a perform ance shift? W hat is th e new nexus o f the performance? F o r a num ber o f reasons, reenacted rituals are obviously transform ed rituals. A reenacted ritual presents us w ith a m ajor sem antic shift. T he contents o f the sem iotic process may appear to be unchanged: the same masks, puppets, and cranes, the same chants, possibly even the same p er­ form ers. Elsew here, I have n o ted th at all rituals have w hat I call a nexus, a central feature around w hich all o ther disparate elem ents in the rite m ust o rien t themselves. In the traditional context, this nexus was invariably the pu p p et body and the process by w hich it “com es alive.” 32 W hat has b e ­ com e the nexus in the perform ance reenactm ent is n o t the co n ten t o f the original ritual, b u t time, itself. R eenactm ent has m ade tem porality, and the existentially disturbing experience o f a ru p tu re o f its continuous nature, the nexus o f the rite. R eenactm ents are o u r new rituals to symbolically confront this ru p tu re , to try to u n d o tim e. R eenactm ents are ritual cele­ brations o f nostalgia. To explore these issues, let us examine tw o cases o f reenacted ritual perform ances and see how they came about.

R eenactm ent I: E b isu -m a i on Susaki Islan d Susaki Island in Fukura Bay, w hich m easures only a few h u n d red m eters in length, form s a natural barrier betw een the quiet waters o f Fukura h ar­ bo r and the Pacific O cean, and is now ow ned by the Fukura fishing collec­ tive. T he place-nam e Susaki is often used along the coasts o f Japan, as the term refers to small islands or stretches o f land created by alluvial action or slow m ovem ents o f tides in and o u t o f bays. O n Susaki in Fukura Bay, there is a small shrine dedicated to E bisu, and it was before this worship site th at the puppeteers presented their perform ance in 1 9 90.33 Before the early postw ar period, itinerant Ebisu-kaki w ould be requested to present their rituals on this island, for it literally m arked the gate betw een the fishing harbor (a safe space) and the unknow n ocean (a dangerous, u n ­ charted space w here anything could happen to a fisherm an in a small boat). Fisherm en w ould often stop at this shrine o n their way o u t to sea if they had som e im p o rtan t request o r wished to ensure their safety. This

T he small w orship hall o n Susaki serves as th e backdrop for a ree n ac tm en t o f a ritual Ebisu p erfo rm an ce in July 1990.

small shrine was thus an im portant location for worship, m ediating safety and danger, the known and the unknow n. Ebisu was the deity who re­ sided at this im portant passage. The shrine, which still exists, has a haiden (worship space) with a small stagelike structure, a honden (inner sanctum ) with a stone effigy o f Ebisu, and a concrete torii dem arcating the space. O ther than that, the island is mostly empty, except for a few warehouses where fishing equipm ent is kept. The island is artificially stayed by webs o f enorm ous entangled concrete shoreline reinforcements. The decision to perform Ebisu on this site (and Sanbaso at the Sanjo site a few days later) appears to have been in p art the result o f the “ o b ­ server effect” o f my own fieldwork. O n this particular research trip, I had brought along a student from the U nited States and video equipm ent to docum ent some o f the perform ances. My permission to film these per­ formances coincided with repeated requests from scholars on Awaji inter­ ested in the history o f Awaji ningyb to record the perform ances in their traditional settings. There was a real sense th at having video footage o f these “reenactm ents” was o f great value, and there was quite a bit o f com ­ petition and behind-the-scenes jockeying for position to get tripod loca­ tions at these performances. While these perform ances were n o t staged

solely for my benefit, it was clear that my presence (and th a t o f my assis­ tant) was a catalyst, and th at the possibility o f foreign students viewing Awaji ritual puppetry held a great deal o f appeal to the people involved. These perform ances were also intended as som ething o f a gift to m e. Al­ th o u g h I had worked for several years studying the past o f Awaji puppetry, I had n o t yet had the pleasure o f experiencing the perform ances in their historical contexts. Prior to the perform ance on Susaki in 1990— the first o n the island in nearly forty years— perm ission had to be obtained from the proper a u ­ thorities and from the fishing collective to land on the island and stage the event. Local people were invited along as an audience. O n the afternoon o f the perform ance, m em bers o f the th eater m ade their way to the docks right after the theater closed. The day was cool and clear, and a slight wind was picking up. T heater m em bers began to unload their equipm ent from mini-vans and o n to boats: all the makings o f a com ­ plete stage, a box o f puppets, and musical instrum ents. My assistant and I w ent over by boat, carrying o u r video equipm ent. The fisherm en took their ow n boats, and several trips were necessary to get everyone from the theater across. By the tim e we arrived, a num ber o f local ethnographers had already set up their video and recording equipm ent. They did not tu rn their cameras o n until the actual perform ance began, whereas my research assistant and I were interested in recording the entire fram ing o f the event as a reenactm ent. O nce on the island, while the theater m em bers were setting up their perform ance space, each person, film crew, theater m em ber, and local per­ son from th e audience presented him self or herself before the Ebisu shrine honden. Inside this h onden, which was opened up for the perform ance, there is a small stone-shaped shintai, the spirit vessel for the deity Ebisu, in the typical shape o f a fat old smiling m an holding a sea bream under his arm. Taking a bottle o f sake and pouring it over the head o f Ebisu, each person then closed hands in prayer. A t this p o in t in the proceedings, it became clear th a t while people were viewing the ritual as a “reenactm ent,” they also realized th at proper behavior was necessary for those w ho w or­ shipped at the Ebisu site. This p art o f the day seem ed unscripted and n o t a reenactm ent. People were n o t acting, n o t pretending to be a traditional audience. They were just being themselves. W ith the prayers com pleted, the perform ance began. Fisherm en sat around under the stone torii, m any with their grandchildren sitting in their laps o r by their sides, and tw o o th er camera crews set up under the torii recorded the event. The co n ten t o f the perform ance was the same as th at translated in the previous chapter: a hum orous play, in which Ebisu appears at the d o o r o f a wealthy hom eow ner, announces his arrival, and asks for sake. As he be-

PUPPETS AND WHIRLPOOLS

235

The audience at this Ebisu performance on Susaki sat under the torii demarcating the shrine.

comes intoxicated and dances, he tells of his miraculous birth, how he is to be worshipped, and what blessings he bestows on those who pay attention to him. At the end, he catches a live fish, a sea bream. At the close of the performance, this fish was presented to the head of the fishing collective, who was in the audience. It was a gift for having allowed the performance to take place. After the performance, I was able to interview members of the audience and the theater. To the last person, the overriding emotion was one of a strange sadness, which can only remind us of the original Sense of nos talgia as "a pain to return home." Everyone remarked that seeing the ritual was "natsukashii," and many of the men recalled coming with their parents and grandparents as children over to the island once a year to watch the performance. Many wondered why this performance was not pre-

A scene from the Ebisu perform ance.

sented at this site m ore often. A few o f them directed their questions to U m azum e-san, w ho jokingly responded th a t “you co u ld n ’t afford us!” H e then reiterated his claim th at the puppeteers at the Awaji N ingyo Joruri Kan are professionals (for which he used the gairaigo term purofueshona-ru). A lthough said in jest, the com m ent had an air o f tru th ab o u t it because for the theater to survive, it had to continue to present joruri pieces to tourists. Awaji puppetry could no longer afford to be an example o f a local cultural perform ance. W hen we consider this perform ance in light o f the four aspects o f ritual reenactm ent listed above, the semantic shift in this new perform ance is dramatically apparent. First, the marking o f perform ers as outsiders appears to be alm ost entirely subsum ed under the activity o f theatrical costume. Both perform ers and audience arrived at Susaki together, and both live in the same area in Fukura. In their everyday lives, they all shop at the same Daiei superm arket, send their children to the same schools, and even have their family graves in the same temples. T he difference betw een them here, then, was the tem porary difference o f the conventional, m odern idea of theater, in which actors and audience are delineated th ro u g h the creation o f a stage. T he perform ers and audience were both outsiders to the island ofSusaki, and it was only when the perform ers donned their costum es that there was any distinction. The m arking o f the perform er as an outsider was no longer an issue in this reenactm ent in the way it had been. N ow accom ­ plished through costum e, the marking served to delineate actors from audi-

ence, rather than th e sacred realm o f the specialist (and the otherness o f the road) from the public, dom estic w orld o f the householder. T h e issue o f th e creation o f ritual space was m ore com plex. A t three levels, this perform ance was in a space rem oved from everyday life: it was on an island w here people never spent any tim e, it was w ithin th e precincts o f a shrine, and it was w ithin a tem porarily co n stru cted stage. R eg ard ­ less o f th e m o d ern ity and even artificiality o f this perfo rm an ce, this layer­ ing o f differentiated space was an im p o rta n t feature. A lth o u g h different from th e ritual space o f itin era n t perform ances in th e past, it was ritual space nonetheless. T h e highly m ixed audience for this event consisted o f eth n o g rap h ers, fisherm en, th eir g ran d ch ild ren , and th e staff from th e theater. M y assis­ ta n t and I, along w ith th e o th e r e th n o g ra p h ers, may have had th e g reatest investm ent in th e perform ance. M uch as past p atro n s d e p e n d ed o n the efficacy o f these perform ances for th eir prosperity, we n eeded a com plete and accurate reco rd o f th e perform ance for the substance o f o u r academ ic careers. T h e o th e r audience m em b ers’ stake in th e p erform ance was o n e o f d is­ ta n t attach m en t. T h e o ld er fisherm en and sen io r m em bers o f the th eater w ere particip atin g in a rem em b ran ce, b u t n e ith e r h ad any special stake in the o u tco m e o f th e perform ance. F o r th e audience, it was a reen actm en t o f th e p ast evoking tim es before th e war. F o r th e th ea ter, it was an experi­ m en t in ree n a c tm e n t and, as well, a p o ten tia l ad d itio n to its rep erto ire and a source o f prestige. A nd at a very basic level, we w ere all playing w ith nostalgia and having a g o o d tim e.

Reenactm ent 2: Sanbasd Performance a t the O m ido H achim an Daibosatsu in Sanjo A few days later th e perform ers from th e Awaji N ingyo Jdruri Kan staged a sim ilar event. T his tim e, the Sanbasd rite was p resen ted at O m id o H achim an D aibosatsu, the w orship hall o f th e cerem onial cen ter o f Awaji p u p p e try in Sanjo. T h e location o f fre q u e n t p u p p e t festivals for h u n d red s o f years, this shrine has a large op en and flat space w here audiences used to sit. Flanked by huge C ryptom eria tress, the area is shaded and cool, even o n a h o t, July m orn in g . Inside the precincts, th e regular w orship hall consists o f a stagelike haiden a n d beh in d it a sm aller in n er sanctum (hon den) d edicated to H a c h im a n . T o th e left o f this w orship hall is a smaller subsidiary w orship site (hokora), w here th e legendary figures o f the fo u n d ­ ing o f Awaji p u p p e try are enshrined in shrine statuary: D d k u m b o , H yakudayu, Ebisu, and a w om an said to be H y akudayu’s wife b u t m ost probably th e vestige o f a sham anic figure.34

Traditionally, itinerant D 5k u m b 5 mawashi setting o u t from Sanjo d u r­ ing the N ew Year period w ould stop at this shrine and present a perfor­ m ance before this worship hall. O n the first day o f the year, all the local pu p p etry troupes living in Sanjo w ould take turns perform ing Sanbaso before this shrine. Even today, this rough and hum ble site is the cerem o­ nial center o f Awaji puppetry. O n the m o rning o f the perform ance in July o f 1990, tw o perform ers from the Awaji theater arrived at the shrine dressed in blue jeans and shirts and shoes. They had driven directly from hom e in the th ea ter’s van, which one o f them had driven hom e the night before. They b ro u g h t their cos­ tum es with them in a case, and all the puppets, paraphernalia, and eq u ip ­ m ent in boxes. T he shrine, located on a country road near Sanjo, was quiet and deserted as usual. T h e worship hall o f the H achim an shrine next to the small hokora for p uppetry served as a dressing room w here they changed into the ritual perfo rm er’s costum e. After dressing, they spread a small straw m at o n the g ro u n d before the w orship hall and arranged the box o f puppets and instrum ents in front o f them . O th e r ritual item s were b ro u g h t in and placed before their perform ing space. As they were setting up, the camera crews (including my research assis­ tan t) proceeded to set up their tripods on the actual worship hall itself, so th a t they could film the perform ance from the perspective o f the “orig i­ nal” intended audience: the sacred patrons o f puppetry inside the h o ­ kora.35 W hen b o th puppeteers and camera people were ready, the p u p ­ peteers sat dow n on their perform ance space and reenacted the type o f perform ance th at w ould have taken place in fro nt o f the small worship hall in years gone by. B ut this tim e, the eye o f the kami was replaced by the lens o f a cam era, which w ould relay this perform ance to students o f an­ thropology, theater, and religion around Japan and the U n ited States. T h e Sanbaso perform ance was the one outlined in the previous chapter, w hich approxim ates those o f itinerant perform ers before the postw ar p e ­ riod. A t the end o f th e perform ance, th e puppeteers repeated the perfor­ m ance again, so th a t it could be viewed from a different angle. T he rep eti­ tion also allowed the editing o u t o f an erro r in the first perform ance, w hen the mask o f one o f the puppets slipped o ff its face. W hile dealing w ith such unexpected things is always p art o f the challenge o f a ritual event, in this case, it could simply be edited o u t o f existence. Repeat th e ritual until it happens correctly. In the process o f this reenactm ent, o th er interesting and unexpected things happened.36 First, w hen the puppeteers were preparing by w arm ing up on the flute and small han d drum used in the Sanbaso perform ance, som e children riding by on bicycles stopped and looked into the shrine from the torii. U m azum e-san called o u t to th em , “C om e w atch the p u p ­ pets,” and w ith great reluctance, they slowly approached the perform ance space. As soon as they arrived, Naniwa-san and her d au g h ter dragged

A Sanbaso p erform ance at O m id o H ach im an before th e war.

A Sanbaso p erform ance at O m id o H ach im an in 1990.

benches over from along the edge o f the gateball co u rt so the children could sit dow n. T he children, slightly nervous, sat there and chatted w ith U m azum e-san and am ong them selves. We treated them to sodas th at we had b ro u g h t along. A few m inutes later (the event was a bit behind sched­ ule) the elderly gateball brigade arrived for their m idm orning game and discovered, m uch to their delight (for m any o f them rem em bered p u p ­ petry perform ances from their y outh) th at a Sanbaso perform ance was ab o u t to begin. Usually, their gateball game begins as each o f th em pray before the shrine, b u t this m orning, they w ent straight for the area where we were perform ing. T he old folks joined the children on the benches, and the “ ritual” began. After the perform ance, the perform ers from the theater, U m azum e-san in particular, stayed in the shrine precincts for ab o u t forty-five m inutes playing gateball w ith the elderly neighbors. U m azum e-san was quite fa­ miliar w ith them , since he lives nearby and grew up in this small village. In the past rituals were frequently followed by perform ers and m em bers o f the audience playing to g eth e r or having a meal. We stayed at the shrine until it g o t to o warm w hen we loaded everything into the van and headed back to the theater. Unlike the Susaki case, which required an intentional trip to see the perform ance, this reenactm ent was very easily reabsorbed into daily events. Everyone living near Sanjo knows th at this site is the center o f ritual puppetry, and it seem ed natural to see the pup p et group perform ing there. A revival, it becam e clear, m ust be exam ined b o th as it is u n d er­ stood by people outside the tradition (who are reclaim ing it as som ething their ow n, but w ho in fact have n o relationship to it) and as it is seen by people w ith a living relationship to it—in this case the seniors w ho play gateball o r the children w ho play hide-and-seek and ride their bikes on the grounds o f this cerem onial center Let us retu rn to ou r four categories for analyzing the sem antic shift o f reenactm ent: the m arking o f perform er as outsider, creation o f ritual space, relationship betw een audience and perform ers, and th e com m it­ m en t o f the audience to the perform ance. H e re , the m arking o f the per­ form er as outsider is enacted th ro u g h costum ing rather than a highly m arked arrival and departure from the w orld o u t there, th a t is, the road. T he ritual space in this case w ould come to include the haiden o f the H achim an worship hall, for it is here th a t this transform ation o f an actor in to a ritual specialist takes place. T he status o f the audience in this perform ance is com plex. Again, we have several groups: staff from the theater, ethnographers, the cameras (which alone faced the perform ance directly while spectators viewed it from the side), and an accidental audience o f young children and elderly gateball enthusiasts. O riguchi S h inobu’s category o f the uninvited guest

(th a t is, h u m an beings at a ritual event) seems to have been reinvented, a lth o u g h in this case the prim ary audience consisted o f cam eras, and n o t a kami. These events in the sum m er o f 1990 were early stages in the process o f reviving the kadozuke aspects o f th e Awaji trad itio n . A lth o u g h these perform ances w ere p resen ted from tim e to tim e on o th e r occasions, the 1990 reenactm ents m arked a new m o m en t in the revival process. After this p o in t, n o t only did kadozuke perform ances becom e objects o f re ­ trieval, they becam e a central concern o f scholars attached to th e Awaji trad itio n and to th e m em bers o f th e theater, for they represented the legitim acy o f Awaji p u p p e try ’s history. Itin e ra n t perform ance was how Awaji ningyo spread, and it is w hat gives Awaji the rig h t to proclaim itself the “ cradle o f ningyo shibai.” T h e em bracing o f th e kadozuke tradition was a strategic step in ensuring th at Awaji p u p p e try rem ained the prem ier exam ple o f rural p u p p e try in th e face o f m any im postors en terin g the scene and riding th e waves o f the dentdβ είη δ b u m u (traditional folk p er­ form ing arts b o o m ). W hile Awaji p uppets have certainly been the m ost celebrated case o f a rural p u p p e try trad itio n being m ade viable an d even lucrative as a theater, there have been o th e r efforts aro u n d Japan to follow in the footsteps o f the Awaji tra d itio n , even w hen this m eant fabricating a history o f p u p ­ petry. In this clim ate o f “ reclaim ing th e vanishing,” ritual p u p p etry and p u p p et tro u p e s presenting ballad dram as have been revived and even “ dis­ covered” th ro u g h o u t Japan. In a few cases, w hich will here rem ain n a m e ­ less, villages w ith no history w hatsoever o f puppetry, or those w here one p u p p eteer m ay have lived briefly, have d ecided th a t p u p p e try is g o o d busi­ ness—jo ru ri perform ances in th e countryside m ake pleasant to u rist o u t­ ings, and hence are g o o d for th e to u rist trade. T hese ningyd no fu ru sa to (hom etow ns o f pu p p etry ) have sp ru n g up aro u n d Japan as a result o f this veritable boom . We tu rn now to the next stage in this revival process, w hich p u t ka­ dozuke fro n t and center: the sum m it m eetings o f 1991, 1 9 9 2 , and 1993.

T h e P o litics o f th e C enter: Aw aji and th e Z en k ok u N in g y d Shibai S am itto In the sum m er o f 1 9 9 1 ,1 was privileged to w itness th e next stage in this revival process, th e developm ent o f th e Z enkoku N ingyo Shibai Sam itto (AU Japan P u p p etry S um m it), a forum for enhancing th e revival o f p u p ­ p e try all over Japan, designed and organized by U m azum e M asaru. By this tim e, Awaji p u p p e try could safely be considered stable in its process o f revival. U m azu m e, w ho genuinely loves Awaji ningyo and

wants it to rem ain viable as a theater form , developed an arena w herein the experiences, struggles, and successes o f the Awaji tro u p e could be used to strengthen the revival m ovem ents o f p uppetry around Japan. T he Awaji tradition had already confronted the issues o f funding, recruiting new m em bers to the theater, preserving the related arts such as puppet-head carving and costum e creation, and advertising and going m ainstream with th e tourist industry. It is clear th a t while the AU Japan R egional P erform ­ ing Arts C onvention, the activities o f preservation societies, and the new M inzoku G eino Taikai (Folk Perform ing Arts C onvention) have been a popular success in the m edia, they have n o t dealt effectively w ith the co n ­ text and com plexity o f the revival o f rituals and the perform ing arts. A different kind o f forum , one less geared tow ard the public w ith little know ledge o f theater, was in order. T he sum m it is a new m odel o f m ee t­ ing for the retrieval and preservation o f folk perform ances. Prior to the early 1990s m ost public settings for scattered folk perform ances were an ­ nual m eetings at w hich diverse perform ances w ould be presented. T he sum m it focuses on one genre o f perform ance (puppetry, in this case, or regional Kabuki, Kagura, D en g ak u ) and brings to g eth er specialists and people directly involved in th e nuts and bolts o f reviving theater. This m odel recognizes the need for a m ore critical look at the current state o f “ traditional” theater in Japan. U m azu m e’s idea was straightforw ard: bring all fledgling p uppetry pres­ ervation societies (ningyo hozonkai) to Awaji for several days to m eet, dis­ cuss the problem s they are facing w ith the logistics o f revival, and see how they can help one another. Follow ing a tren d in o th er folk perform ance groups o f using the “sum m it” m odel, he decided to convene such a su m ­ m it for puppetry.37 T h e sum m its, to continu e for several years, w ould begin th a t fall and w ould take place on Awaji. In h ere n t in this m ove was a m odern version o f an older reality: Awaji w ould still be the center o f puppetry. Because o f Awaji’s unique status in the history o f Japanese p u p ­ petry, coupled w ith its recent success in the process o f revival, it was the natural setting for such a sum m it. Preparations for these sum m its consisted largely o f inviting the heads o f p u p p et groups around Japan to attend. To this end, U m azum e decided th a t it w ould be a g ood idea to travel to those groups in w estern Japan and m eet w ith them individually. T h at sum m er, I traveled w ith U m azum e and a scholar o f the Awaji tradition, Nakanishi H id eo (a high school teacher and th e director o f the high school p uppetry club), th ro u g h o u t Shikoku and Kyushu for several days while we visited various p uppetry groups and invited them to the sum m it th at fall.38 D uring this trip, we visited a n um ber o f small pu p p etry groups, some no longer active, som e being revived th ro u g h school clubs and student groups, and others attem pting to em ulate the Awaji m odel by building

m o d ern theaters. A t each th eater, U m azum e-san was involved in inviting the tro u p e s to a tte n d the Awaji sum m it. A lth o u g h I learned a g reat deal a b o u t ritual p u p p e try and th e histories o f a n u m b e r o f small tro u p es aro u n d Japan, it rapidly becam e clear to m e th a t th e in te restin g a n th ro p o ­ logical event was n o t h isto ry o r even th e p u p p e try groups b u t th e process o f tra d itio n b u ilding in th e p o stm o d e rn era. T he Awaji scholars and d irec­ tors w ere co n stru c tin g a revival, and th e language and action o f this process w ere heavily co d ed w ith created m eanings and th e staging o f id e n ­ tities, pow er, hierarchy, m oney, and politics o f a very real kind. In sh o rt, the p erfo rm an ce h ere was n o t th e one th a t had taken place in the past; it was th e in teractio n betw een th e directors o f th e various th e a te r groups; the “ acto rs” w ere n o lo n g er p u p p ets b u t th o se w h o m anipulated their m anipulators— tro u p e m anagers and th e financiers o f th e m any hozonkai th ro u g h o u t Japan. T h ere can be n o d o u b t th a t U m a z u m e M asaru is genuinely in terested in seeing p u p p e try p erform ance, particularly jo ru ri, becom e a living p e r­ form ing a rt in Japan again, and n o t solely in th e expensive national th e a ­ ters in O saka and Tokyo. F o r several h u n d re d years in th e h istory o f Japa­ nese p erfo rm in g arts, as th e classical trad itio n s o f B unraku and Kabuki w ere solidifying th eir aesthetics, highly inventive an d tale n te d p e rfo r­ m ance artists w ere m ain tain in g th eaters th ro u g h o u t Japan. T hese theaters thrived w ith o u t th e patro n ag e o f th e u rb an centers, an d they developed im p o rta n t p erform ance styles th a t urban theaters o ften copied. U m azum e was con cern ed to preserve these trad itio n s n o t m erely because they w ere p opulist, n o r was he m otivated solely by econom ic im pulses. H e felt th a t th e them es and aesthetics o f Awaji ningyb shibai are deeply significant, and he genuinely w an ted this th e a te r form to be h a n d e d dow n to fu tu re g enerations. H e often talked a b o u t h o w th e range o f em o tio n s and senses b ro u g h t a b o u t by w orking in th e th e a te r to u c h e d him at a deep and exis­ tential level, and p rovided a narrative for his ow n life. S o m eth in g a b o u t th e p u p p ets, th e plays, and th e tra d itio n was, to b o rro w T ennessee W il­ liams'’ expression, em otionally autobiographical for him . I was in terested to w atch o n this trip h o w th e directors o f th e theaters in Johnny-com e-lately h o m eto w n s o f p u p p e try tried to cu rry favor w ith the Awaji tra d itio n , and h o w at th e sam e tim e U m azu m e c o n tro lled access to these sum m its in such a way as to assure th e role o f Awaji as th e center o f this new m o v em en t in nostalgia and th e in v en tio n o f trad itio n . H e decided w hich th eaters w o u ld be considered actual historical th eaters and w hich w o u ld be allow ed to participate as “ m o d ern in v en tio n s.” U nlike o th e r festivals, such as th e p o p u lar Folk P e rfo rm in g A rts C o n v en tio n and th e o ld er AU Japan R egional P e rfo rm in g A rts C o n v e n tio n , w here a u th e n ­ ticity and an air o f th e ancient w ere im p o rta n t, in this new fo ru m it was m eaningful and even desirable to have new p u p p e try clubs p a rtic ip a te .39

From U m azum e’s perspective, reviving puppetry as a theatrical m e­ dium around Japan can only be good for the Awaji tradition. B ut the very success o f these sum m its created problem s in the local setting. After the third sum m it, covered widely in the national m edia, tensions arose am ong people on Awaji concerning w hat was perceived to be a dilution o f Awaji’s uniqueness as o th er pu p p et theaters around Japan becam e viable again. As the sum m it drew to a close and the Awaji N ingyo Joruri Kan staff col­ lapsed in exhaustion, one o f the m ajor financial backers o f Awaji puppetry came to the theater to com plain th at this new form at threatened th e m o ­ m entum th at had been built on Awaji. U m azum e was severely chastised for taking a national (zenkoku) perspective in his attem pts to revitalize puppetry. T h o u g h the sum m it has been a great success, U m azum e ad m it­ ted that he had created a difficult political problem for him self at the local level. H e was unable to conceal his frustration and disappointm ent th at his intentions were n o t being fully recognized by local people.

From R itu a l to Stage: The Problems of Context in the Preservation Movement T he Z enkoku N ingyo Shibai AU Japan Puppetry Sum m its represented a new era in the revival m ovem ent o f Awaji ningyo. Since th en , there has been a self-conscious attem p t to retrieve the ritual tradition and transpose it o n to the stage. As dem onstrated in previous chapters, ritual p uppetry is d ependent on a highly constructed context to be experienced as effective. T he ritual depends on a shared com m itm ent to the perform ance on the p art o f perform er and ritual sponsor in w hose house (or on whose bo at or dock) the rite is being presented. W hat happens w hen the ritual context and the relationship betw een the perform er and the sponsor shift dram at­ ically? W hat happens w hen ritual hits the stage? In this chapter I am suggesting th at part o f the dynamic involved in reviving the ritual dim ension o f puppetry has been a response to a c o n ­ structed discourse o f nostalgia, in which the quest for the truly Japanese becom es the quest to reclaim the vanishing and the exotic. T he outw ard expression o f this shifting discourse, I m aintain, has been the initial n e ­ glect o f the kadozuke rituals o f the Awaji tradition followed by the sudden and intense interest in retrieving these rituals and casting them as stage perform ances. T here are scholars in Japan and the U nited States w ho claim th a t this tendency to retrieve rituals by m aking them stage en tertainm ent makes it impossible for the theater in Japan to develop, since these icons o f the past inhibit the dynamism o f the theater as a reflection o f and response to the p resent.40 Sasahara, one o f the m ore outspoken critics o f this

“stage m o v e m e n t,” observes th a t the tim e and space o f the stage are a rti­ ficial, so th a t at folk perform ings arts conventions, a sum m er bon-odori, a cherry blossom view ing dance, and a fall dew K agura are stacked one on to p o f a n o th e r, all d eracinated o n th e stage. H e decries th e im plicit ideology o f “ o rig in s,” “ th e a n c ie n t,” and th e c o n stru c te d aesthetics o f “loveliness” and “ beauty,” and he d en o u n ces these new “staged ritu als” as sham ritu a ls.41 T h e pro b lem w ith th e critique o f Sasahara and o th ers is th a t they have restricted th eir definition o f th e cultural p erfo rm an ce to m erely w hat is p resen ted on stage. T hey are prim arily concerned w ith identifying w hat these p erform ances m ay m ean for o u tside audiences. F ro m my research, it is clear th a t th e actual staging o f th e rituals is b u t one m o m e n t in a larger series o f cultural perform ances w hich intersect o n th e stage. People on Awaji have o n e cultural perform ance in m ind w hen they see these revived rituals. People from Tokyo have an other. A nd scholars from ab ro ad have yet another. W hile Sasahara and o th ers correctly identify th e m anipulation o f nostalgia in this process, they have focused only o n th e nostalgia o f the ou tside consum ers o f these rituals. F o r people on Awaji, attem p ts to stage these rituals are n o t sim ply statem ents a b o u t th e past. T hey are openen d ed questions a b o u t it, ways o f exploring w h at these rituals m eant and w hat they m ig h t still m ean. P eople are actively p o sin g th e q u estio n , D o these rituals still m ean anything at all? A nd they are decidely open to the answ er th a t perhaps the rituals will be only rem n an ts from th e past, r e ­ m em bered onstage. In th e latter stages o f m y fieldw ork, I n o tic e d a m arked a ttitu d e o f ex p erim entation and play at w ork in the theater. People w ere genuinely en gaged in exploring th e q u estio n o f w hat these rituals m ig h t m ean again in the m o d ern context. O n ce, w hile we w ere sittin g in a back ro o m d is­ cussing th e m eanings o f reviving Sanbaso, I n o tic e d th a t th e Sanbasb p u p ­ pets w ere sim ply h a n g in g on stands, a practice w hich w ould have been considered disrespectful fifty years ago. I n o te d this, and U m azu m e r e ­ sp o n d ed w ith a laugh, “ I th in k Sanbasb is g e ttin g used to all these changes, and can to lerate ju st a b o u t an y th in g these days.” Perhaps for people from elsew here, retrieving kadozuke rituals was p a rt o f a discourse o f nostalgia and exoticism . F o r people on Awaji, it was a b o u t exploring a eclipsed p a rt o f th eir historical experience. Suddenly it was okay to talk a b o u t itin eran t p erform ance and to explore its sem iotics, its textures, and its constraints and freedom s. T h e decision to transform kadozuke perform ances in to stage pieces p ro p e r is an ideal case study for stu d y in g w h at happens w hen ritual b e ­ com es theater. T h e decisions here have been m ade consciously, w ith the in te n tio n o f retrieving and show casing th e past, w hile fully aw are th a t the c o n tex t has b een lost.

I have already indicated th a t the first step in this process was the unsuc­ cessful retu rn o f puppet rituals to original contexts— on Susaki and in the historical center o f ritual p uppetry in Sanjo. These perform ances failed to bring back these rituals, for they m erely served to confirm w hat people already knew: at best the audiences for these reenactm ents w ould be fabri­ cations, people b ro u g h t o n to the scene artificially to play the role o f audi­ ence. At a deeper level, these reenactm ents were hau n ted by the sense th a t perhaps the rituals in these contexts w ere m eaningful only for ghosts o f times now past. N ext, the pieces were cast as stage perform ances. H o w this dram atic shift in context affected the ritual process is best seen the staging o f the Sanbaso rite. T he four versions o f the Sanbaso ritual m ost com m only p er­ form ed as stage entertain m en t reveal the recognition and acceptance o f the constraints and liberties o f this new m edium , particularly for the a n o n ­ ym ous, disinterested, and frequently ig norant audience. Ironically, the new context constrains the audience. W hile it w ould be perfectly accept­ able to drink sake o r beer, or to belch, sm oke cigarettes, and stretch ou t, or even to talk, get up, and walk around during a Sanbaso ritual in a shrine, anything sh o rt o f com plete quiet and feigned rapt a tten tio n while in the audience at a staged perform ance is unacceptable. C onsequently, such perform ances are exhausting, as they were usually n o t designed to be w atched w ith such an atten tio n to detail. A Japanese m an I know w ho was a student in the U nited States in the early days o f television said th at w atching a Sanbaso ritual o n the stage was like staring at the picture o f the Indian that decorated the T V screen w hen the station had gone o ff the air. I will review each o f these four Sanbaso versions w ith atten tio n to ( I ) the nature o f the p u p p eteer’s role as perform er, (2) the significance o f the box o f ritual im plem ents carried by itinerant perform ers in the ritual context, (3) the m eaning o f the perform ance space, and (4) th e relation­ ship o f the audience to the perform ance. These four com ponents reveal the creative response to the loss o f th e ritual context as a new semiotics com es in to play. Each com ponent o f the Sanbaso rite assumes a new level o f signification as it is introduced in to the perform ance space o f an in d o o r theater stage. T he Russian folklorist Petr Bogatyrev has argued th at “the stage radically transform s all objects and bodies defined w ithin it, bestow ­ ing upon them an overriding signifying pow er which they lack.” 42 This “ m anifesto” (as Elam calls it) o f th e Prague school m aintains th a t “AU th a t is on the stage is a sign.”43 As show n in previous chapters, the role o f the pu p p eteer as outsider to the com m unity was a key elem ent in the efficacy o f the ritual. T he p u p p e­ teer and his puppets em bodied many opposites: p u p p et vs. hum an, p u p p et vs. sacred, hum an vs. sacred, and itinerant puppeteer (h in in ) vs. agrarian settled person ( ryom in). M uch o f th e perform ance was a b o u t the m anipu-

Jation and resolution o f these relationships o f alterity. The puppeteer was marked as an outsider by being dressed in special ritual attire and carrying a box slung over his shoulder. Special footgear, usually straw sandals, indi­ cated the puppeteer’s low social status, as did the fact th at his head was always uncovered. Prior to the perform ance, a call on the flute served to announce the arrival o f an outsider. But when the ritual piece is trans­ posed to the stage, how will these signs be carried over? Because the p u p ­ peteer is now som eone employed at the theater, the real outsiders are members o f the audience. It is they who arrive at a theater and then leave, unlike the previous context where a puppeteer arrived at a hom e and then left. Is a theater puppeteer w ho presents itinerant ritual as stage en tertain ­ m ent a new kind o f actor? Arrival and departure are still essential to the event, but the roles are shifted.

Version I: Low-Level Contextual S hifting Sanbaso as Stage Performance T he earliest attem pts to present Sanbaso as a stage perform ance began in about 1987. I first saw this version o f Sanbaso on stage at the Fukura theater on New Year’s Day in 1988. W hat struck me about the perfor­ mance was that attem pts to adapt the perform ance o f a ritual to the constraints o f the stage were minimal. T he perform ance to o k place on a small raised platform outside the reg ­ ular theater at the N aruto Kaikan. I t was about eleven o ’clock in the m orning, and as the event had been announced in the local paper, a few people from the tow n had come up to the theater with children and grandchildren. A crowd o f about thirty or forty people gathered. The perform ance itself was brief, w ith little attention to arrival and d ep artu re. The perform ers simply came o u t from the back o f the theater carrying their puppets onto the platform (which had been prepared in advance w ith a m at), set them in place, and then began the perform ance with the musical instrum ents. W hen the perform ance was over, they picked everything up, w ent back inside and left the stage for an im prom ptu a p ­ pearance by another outsider: the mayor o f Seidancho announced th at a foreign scholar was visiting and asked me to ascend the platform and com m ent on the perform ance. Tem pted as I was to com m ent on “myste­ rious visitors at New Year,” I refrained from irony and discussed th e ritual significance o f this perform ance o f Sanbaso at New Year again after a hia­ tus o f several decades. This perform ance o f Sanbasb, unself-consciously and presented on the spur o f the m om ent in a rather m atter-of-fact way, was the first step in the revival o f the Sanbasb rite as part o f the New Year’s events at the theater.

After the perform ance, I spoke w ith people in the theater and asked them how they felt ab o u t it. N o one seem ed very m oved one way o r an ­ other, b u t there was a concern th at the perform ance had been flat. S om e­ th ing had n o t clicked. This perform ance, I surm ise, initiated a conscious­ ness th at Sanbasd ritual was w o rth the efforts o f preservation b u t th at the process w ould be tricky. T he subsequent perform ances on stage (co m ­ bined w ith the attem pts discussed above to retu rn the ritual to its original context) were levels o f discovery in this process. People at the th eater and elsewhere were actively asking the question, Can we bring this back, and if so, how?

Version 2: Kadozuke on Stage U ntil very recently, w hen Shiki Sanbasd was presented on stage, it was always an exact replica o f the kadozuke perform ance, follow ing early p o st­ war filmings o f the rites in th a t context. T he tw o puppeteers walked o n to the stage and all their instrum ents and puppets were set o u t for them . T here was little concern o r reflection about effecting a shift from ritual to theater. In the perform ance, the th eater used a curtain w ith a pine tree painted on it, the same backdrop used in N o h for the dance o f Okina. T he m ost significant aspect o f the perform ance was th a t the offerings m ade to the deities in the beginning o f the rite were o m itted completely, suggest­ ing th at for a ritual to becom e theater, it m ust be divested o f its overtly religious content. This was the standard version o f Shiki Sanbaso on stage th at I saw re ­ peatedly during my fieldwork. F or people familiar w ith the piece, the dance, m usic, and chanting make it an evocative presentation. B ut the consensus at the theater was th at som ehow this piece lacked the depth th a t w ould have given it a transform ative power.

Version 3: New Actors in the Shiki-Sanbaso Performance At the th ird pu p p et sum m it, the Awaji N ingyo Joruri Kan prem iered its latest version o f the Shiki Sanbaso rite. It represents a new stage in the shift from ritual to theater. In it, the puppeteers are actors, w ho play the role o f itinerant puppeteers. W hile the greater p art o f the perform ance is copied from the perform ances o f postw ar kadozuke perform ers and is identical to the previous perform ance, this subtle shift in the perform ance changes its m eaning entirely. H ere is how it is accom plished: At the beginning o f the perform ance, the curtain rises to reveal a stage decorated w ith the masks used in the perform ance. They are placed on an

offering stand in full view o f th e audience. A fter a few m o m en ts, from stage left, tw o m en dressed in black kaori hakam a enter. O ne p u p p eteer carries o n his back a box w rapped in a black furoshiki (cloth) and the o th e r person carries a bag. T hey sto p , face the masks (and the audience), and th en set up th eir perform ance space, carefully spreading a cloth o n the g ro u n d and unpacking all th eir ritual in stru m en ts and puppets. T his part o f th e perform ance takes a b o u t th re e o r four m inutes. T h e perform ance th en proceeds exactly as the o th e r Sanbaso rites described above, except th a t w hen it is over, th e p uppeteers stop and pack up th eir p u ppets and in stru m e n ts in full view o f th e audience, and th e n get up and walk offstage. O nly th en does th e curtain com e dow n. It was in terestin g to w atch the audience reaction to this sta te m e n t th a t the p u p p eteers are now actors playing the p a rt o f itin eran t perform ers. W hen th e last p u p p e t was p u t d o w n, som e people in th e audience felt free to get u p , m ove a ro u n d , and talk to one another. O th ers in th e audience politely asked th em to sit do w n , as the actors w ere still o n stage and the perform ance was n o t over. N o cu rtain had com e dow n yet. O n e m an, w ho by the end o f the su m m it had earned him self th e nicknam e o f “ G ekijo no F udo ” (ferocious guardian o f th e th ea ter), was quite upset and began call­ ing o u t to people to sit dow n until th e curtain fell. Since several Sanbaso perform ances from a ro u n d Japan w ere being p re ­ sented at this sum m it, I asked a n u m b er o f people w hat they th o u g h t o f this idea th a t th e tran sitio n from ritual to stage m ean t th a t th e puppeteers w ere now actors playing th e role o f itin eran t p uppeteers and th e puppets props for th e p erform ance, rath e r th a n m edia for ritual. M ost people ini­ tially found this to be a very d istu rb in g idea, b u t after som e th o u g h t n o ted th a t in d eed this n o tio n o f acting was apparent.

Version 4: Noh Aesthetics on a Puppet Stage T he fo u rth version o f Sanbaso I have seen o n stage was also prem iered at th e 1993 p u p p e try sum m it. T h e piece com bined tw o features n o t seen in any o f th e o th e r versions. First, ra th e r th an having th e puppeteers recite the c h a n t, it was chanted by a ch o ru s seated to the rig h t o f the stage in th e c h a n te r’s box. M em bers o f th e chorus held small d ru m s, and their postu res and style w ere clearly an im itation o f th e N o h theater. Second, it devoted m uch m o re a tte n tio n to dance and m ovem ent on th e stage, show ing m uch m ore o f a family relationship w ith the extant Awaji lineage Sanbaso perform ances found in o th e r parts o f Japan. It is clear th a t this piece reflects a synthesis o f tw o different strategies for b ridging the worlds o f ritual and theater. O n th e one h an d , it retrieves a p art o f the p erfo r­ m ance from an o th e r location and rein tro d u ces it o n Awaji, a m ove th a t

argues for a simple return to the past. On the other hand, the piece com ­ pletely avoids the dilemma by appropriating N oh aesthetics, a move sug­ gesting that the project o f bringing ritual onto the stage as ritualhad been abandoned. These four versions o f Sanbaso performance reflect the dynamics in­ volved in crossing over from ritual to theater. People at the theater ex­ pressed some frustration with the process, but everyone concurred that the symbolism in Shiki Sanbaso was coherent and powerful enough to communicate something to people regardless o f its framing, and for this reason, they are continuing to play with the piece to see how to best pre­ sent it in the new context.

O ther Voices o f Revival: A m ateur P u ppetry o n Awaji Thus far, our attention to the revival o f Awaji puppetry has focused on the activities o f the only professional theater on the island, the Awaji Ningyo Joruri Kan, under the direction o f Umazume Masaru. We have seen how this theater has had to work with the demands o f a nativist discourse re­ quiring that it be part o f a packaging o f Awaji as the rural and exotic. The professional theater on the cape overlooking the Great N aruto Bridge is not the only future for Awaji puppetry. Awaji was considered “the island o f the puppets” for several centuries not simply because there were numerous professional theaters and ritual performers on the island, but because people living on Awaji took an avid interest in joruri recita­ tion, puppet manipulation, and even puppet kashira creation. A story I was often told while I was on Awaji (though no one remembered exactly when or where the events occurred) reflects the pride people had in their identity as “the island of the puppets.” A chanter from the Osaka Bunrakuza (the major puppet theater in Osaka at one time) came to Awaji as a “guest” reciter. Apparently he underestimated his rural audience and was tired and bored as he chanted the jbruri. People in the audience were outraged that his performance was so poor, and they began to yell and demand that he stop. Finally he got down off the stage, and a farmer from the audience jumped up and finished the performance. Did this really happen? W ho knows? But I was told this story many times to underscore the conviction o f people on Awaji that at one tim e, nonprofessional reciters from the island were better than a chanter from Osaka on an off day. This amateur tradition on Awaji, which suffered the same fate as pro­ fessional theater in the postwar period, has also been revived in recent years. In Mihara in particular, another revival movement is in process,

one that self-consciously presents itself as “the real thing, done by real people and n o t professionals.” A num ber o f times during my fieldwork, I was contacted by people from this rival preservation society and encour­ aged to give up my studies o f the “artificial stage puppets o f the th eater” and come “back to the hom etow n o f puppetry,” where real people were working to keep this tradition alive. This split between the am ateur and the professional has been a dynamic as old as Awaji ningyo, it seems, and it should come as no surprise th at it appears again in this stage o f the revival m ovem ent. A lthough they do n o t have the formal theater or the professional staff, this group o f people in M ihara is nonetheless reviving puppetry. They get together to chant jdruri, an activity which has always been a part o f life on Awaji,44 learn to carve and create kashira, prepare puppet costum es, and practice puppet manipulation. This group does n o t perceive itself as p ro ­ fessional but rather as a local group interested in their local art form. They meet in the civic hall and at senior centers and attend lectures on the history o f Awaji puppetry given by local scholars. M any o f them are study­ ing the arts o f their parents and grandparents. M ost are over sixty-five and many recall puppetry from before and immediately after the war. This group also has made a concerted effort to revive the related art o f carving and painting kashira. M aking a kashira th at can easily be m anipu­ lated is quite difficult to master. O e M inosuke, advanced in years, has a studio in his house across the bridge on Shikoku in the tow n o f N aruto, where he carves kashira for the Bunraku stage. O n Awaji, the grandson o f the famous puppet-head carver Urakame has taken his grandfather’s arts name and learned the art o f making kashira. H e hopes to revive this essen­ tial aspect o f puppetry on Awaji, O n Awaji in Sanjo, a num ber o f people have taken up this problem in earnest and produce beautiful (though perhaps n o t very practical) kashira. I have interviewed a num ber o f them , and they all stress their concern that this im portant part o f puppetry is dying out. M any indicated th at the ka­ shira they were carving could never actually be used as real puppet heads, because they could n o t ever expect to master the construction o f the com ­ plex karakuri (inner workings) o f th e kashira. At the 1993 sum m it, in the lobby o f the theater while performances were going on, Urakame sat on a small platform dem onstrating the art o f making a puppet head. H e was there for both days the sum m it m et at the theater, working from m orning until evening. The purpose o f “staging” this aspect o f the tradition as an ongoing perform ance in the lobby was to draw attention to the plight o f puppetry as an art form and to make a very im portant point: I f no one knows how to carve puppet heads, how can ningyo joruri continue?45

O e M inosuke, the m aster p u ppet-head carver, in his w orkshop in N aruto-shi.

T his g ro u p also has its ow n scholar, a M r. F u jin o , w ho is an energetic retired schoolteacher and an avid cam era and video enthusiast. H e is also very in terested in learning to carve kashira and w rites and self-publishes detailed and exhaustively researched articles on the histo ry o f various as­ pects o f Awaji ningyo. Self-publishing has becom e an im p o rta n t m eans o f expression for p e o ­ ple c o n n ected w ith this am ateu r revival g ro u p . T his expresses th e lack o f access they have to m ainstream academ ic p u b lishing houses, b u t to an even greater ex ten t, it is a fu n ctio n o f th eir in te n d e d audience. W riting chiefly for them selves, they p ro d u ce e n o u g h copies o f th eir books and articles to share w ith o th e r people involved in this process. M r. F u d o Satoshi, the old m an w ith the p h o to g ra p h s I m en tio n e d at th e sta rt o f this bo o k , has follow ed in his fath e r’s footsteps an d has w ritten his version o f the revival o f Awaji p uppetry. T he b o o k contains copies o f prim ary source do cu m en ts for studying Awaji along w ith recollections a b o u t th e revival process. These self-published articles and books should n o t be assessed only on an academ ic level, needless to say; th ey should be seen as p a rt o f a larger cultural perform ance o f id en tity c o n stru c tio n and p resen tatio n on Awaji today. T h e re is som e tension betw een this loosely affiliated g ro u p and the for­ mal th ea ter, since the am ateurs often p o in t o u t anything th a t appears to be n ew o r invented in the professional theater. In spite o f the tensions, w hen national events have focused o n Awaji, these tw o groups— th e p r o ­ fessional p u p p eteers o f the N ingyo Jo ru ri Kan and the inform al M ihara g ro u p — have chosen to p resent a u n ite d fro n t. T h e first day o f th e th ird Z enkoku N ingyo Shibai sum m it in O c to b e r 1993 was convened at the local city hall near Sanjo in M ihara-cho, and while perform ers from the professional th ea ter in F ukura rep resen ted Awaji o n the stage, display halls w ere set up th ro u g h o u t th e b uilding show ing th e h istory o f Awaji p u p ­ p etry and displaying th e new kashira and costum es being m ade by local people from th e local group.

Homegrown E bisu-m ai: A m ateur R evival O n Awaji in th e su m m er o f 1 9 90, I had an o p p o rtu n ity to see h o w this am ateur process com pared w ith o th e r revival activities o f ritual p u p p e try at th e professional theater. O n e elderly m an by th e nam e o f Inai H aruichi has revived th e practice o f itin era n t Ebisu perfo rm in g , and travels ab o u t Awaji in his car giving E bisu perform ances in local schools, senior citizens centers, and local festivals and shrines. H e only uses p u ppets he has m ade com pletely him self, w ith the help o f his wife w h o sews th e costum es. T he

254

CHAPTER 6

Inai Hacuichi with his homemade puppet ofEbisu .

kashira of the puppet he uses is decidedly rustic, and the paint job on the face is not in the tradition of adding ground seashell to the paint to give a special patina to the "skin." His kashira and costume are quite bright, and the puppet body has hands and feet, like all Ebisu puppets. Although Mr. Inai admits that his puppet is soboku (rough-hewn), he works hard at manipulating it well. He performs for free, and seems to have a great time doing so . He carries his stage materials in a box in the back of his car. A native of Sanjo, he has recently become a wealthy man, the owner of a large house surrounded by farmland on the site of his childhood home. Now nearly eighty, he says he hasn't had so much fun in years. Most interesting about his activities is the fact that he seems to be naturally recreating a niche for Ebisu-mai as a relevant performance, and not merely a "reenactment."

W hen he discovered th a t I was studying ningyo, M r. Inai cam e to F ukura and p rese n ted a p erfo rm an ce in th e zashiki o f m y h o st fam ily’s house. I t to o k him a few m o m en ts to set up his tem p o rary stage, and because it was the first tim e h e had p erfo rm ed for a foreigner, he was quite nervous. In to th e basic sto ry o f th e E bisu perform ance he wove elem ents in te n d e d to appeal to m e, and he added com m ents and blessings for my research, m y m arriage, my fu tu re as a m o th e r, and even hopes th a t I w ould w rite g o o d books a b o u t Awaji puppetry. H is p erform ance to o k full advantage o f th e flexibility b uilt in to th e E bisu p erform ance to m ake it relevant to the situation. Because the house w here I lived was op en to the narrow stre e t, people passing by sto p p ed in w hile th e perform ance was in progress, and a few children from th e n e ig h b o rh o o d m ade a lo t o f noise and tried to ru n up b eh in d th e screen w hile he was p erform ing. M r. Inai dealt w ith these unexpected events by w eaving th e m in to th e p erfo r­ m ance. E bisu, q u ite d ru n k , chastised th e children in local dialect, and everyone laughed. L ater th a t w eek, he perfo rm ed Ebisu-m ai o n stage at th e local Ebisu festival in S an jo . It was evening, and everyone from th e village h ad assem ­ bled for th e festivities. Stalls sold sweets and grilled co rn and toys for children. E veryone p resent m ade a trip to th e m ain w orship hall o f the Ebisu shrine to pray and drink a small glass o f sake w ith th e deity before crossing th e stre e t to w here m o st o f th e festivities w ere taking place. A stage had been set up w ith chairs scattered a b o u t, and m em bers o f the co m m u n ity w ere taking tu rn at karaoke, singing w ords to fam ous songs backed up by an “ em pty o rch e stra .” Because I was at th e festival th a t evening, I was asked to in tro d u ce M r. In a i. H e p resen ted his perform ance on th e stage rath er th an in fro n t o f th e shrine and in his recitatio n o f the chant m ade several h u m o ro u s references to th e festival. T h e style o f M r. Inai differs greatly from th a t o f the theater. W hen perform ers from th e th e a te r p resen t th e E bisu p erform ance, they are c o n ­ cerned w ith th e “re c o n stru c tio n ” o f an a u th en tic p ast and m ain tain in g the perfo rm ance as a p eriod piece. T h e E bisu p erform ance is now a standard p a rt o f the rep e rto ire o f Awaji ningyo w hen they perform ab ro ad , and the universally recognizable situation o f so m eo n e in toxicated being acted o u t by a p u p p e t is always very popular. T h eir style and ch a n tin g are beautiful, and th e p u p p e t m an ip u latio n is highly skillful. By co n trast, w hen M r. Inai presents E bisu, he also m akes th e perform an ce h u m o ro u s by en hancing th e d ru n k en aspects o f E bisu, b u t his concern is w ith m aking th e p erfo r­ m ance as relevant as possible. F o r exam ple, Ebisu n o w asks for sake by b ran d nam e, w hich draw s m u ch lau g h ter, and references to blessings are specific to th e au d ien ce.46

Friendships, Play, and B lu e Eyes: T h e R o le o f a Foreigner in the Process o f Revival I t rem ains to be asked w hat m ost anthropologists until recently asked and answ ered only w ithin the private spaces o f their own journals: H o w did my presence influence the process w hich took place before me? H o w did my ow n experience influence w hat happened? W hy did I ask the questions I did? In the course o f doing this research, I becam e very attached to the people w orking on Awaji to revive puppetry. T heir m otivations and dedi­ cation to their tradition, to trying to w ork o u t w hat Awaji p uppetry m eant to their ow n identities, m oved m e, deeply at tim es. F or people like U m azum e M asaru, F udo Satoshi, N aniw a K unie, Inai H aruichi, Nakanishi H ideo, and many others, reviving Awaji p u p p etry was a m ajor p a rt o f the stories they are constructing and telling a b o u t their lives. As I w orked and traveled w ith them , and helped so rt th ro u g h the fragm ented evidence o f Awaji ningyo’s past, I became a small p a rt o f the story, and their stories becam e a p art o f my story. H ere, I am interested in how these stories start to interconnect. It is n o t possible for m e to discuss the revival o f the Awaji tradition and the m eanings o f nostalgia w ith o u t im plicating m yself in this process. My awareness th a t I had already been im plicated came one afternoon in the sum m er o f 1988 while I was reading in the Awaji H istorical Archives in Sum oto. In a volum e on Awaji puppet heads, I read an article w ritten by U m azum e M asaru. In this article, U m azum e writes th at the relationship betw een p uppetry and religion is an im p o rtan t one, and “even a blue-eyed girl from the University o f C hicago is investigating th is.”47 T he next day at the theater, I asked U m azum e-san ab o u t the com m ent. (I was p a rticu ­ larly concerned th a t he had g o tte n my eye color w rong.) H is response was interesting: “ Saying som eone has blue eyes is simply a way o f saying the person is a foreigner. T he relationship betw een religion and Awaji p u p ­ petry is one area th a t has been a problem to study.48 W hen I m ention th at a scholar from abroad is interested in this subject, it makes it okay for us to talk ab o u t it.” In the process o f w riting this chapter, this com m ent has com e to have a new significance to my study, for it implicates me in the revival o f Awaji p uppetry and its grow ing inclusion o f religious dim en ­ sions in the constructed narrative ab o u t w hat Awaji pu p p etry means. U m azum e him self has had a deep interest in the kadozuke tradition o f Awaji puppetry. T he first tim e I m et him in 1984, he drove m e around Awaji to show me the geography o f Awaji ningyo. We w ent to the shrine in Sanjo, walked along country roads, and visited sites w here Sanbasb per­ form ances had been com m on. H e was especially anim ated w hen he dis­ cussed the ritual dim ensions o f puppetry, and he described the Sanbasb

p erform ances w ith grea t a tte n tio n to detail. I t becam e clear, how ever, th a t as a th e a te r directo r he was finding it difficult to develop his keen interest in kadozuke. O ver th e years since 1 9 8 4 , U m azu m e and I have w orked closely, and m any o f the view points I have taken in my research have been suggested by him . In m any ways, I have been so m e th in g o f a foil for him . Because I am a foreigner, it is possible for him to use m y presence to ask a lo t o f q u estio n s th a t w ould otherw ise rem ain unasked. M anipulation o f this for­ eignness has been an im p o rta n t p a rt o f m y particip atio n in th e revival o f this trad itio n . T h e re was also a certain playful quality th a t was possible d u rin g my fieldwork. M uch o f the w ork I did w ith U m azum e-san had an academ ic purpose b u t was also ju st plain fun. D riving a ro u n d in th e co untryside looking at inscriptions in d eserted shrines, g o in g o n w alking to u rs o f Sanjo and talking to elderly people a b o u t their recollections, taking boats o u t to d eserted islands to look at am azoku to m b s and telling stories o f retrib u tio n s from m alevolent spirits to scare each o th er, traveling aro u n d Kyushu m ee tin g people and lo o k in g at th eir p u p p e try collections w hile we invited th em to the ningyo sum m it m eetings, o p en in g the in n er sanctum o f a child h o o d shrine— all these experiences afforded U m azum e-san a cer­ tain freedom his jo b as directo r o f th e th e a te r w o u ld n o t have m ade possi­ ble h ad he n o t had th e responsibility o f taking care o f a fo reign scholar. F o r m any historians o f religions and an th ro p o lo g ists, this aspect o f play in the fieldw ork experience and deep relationships th a t develop from it b e ­ com e an im p o rta n t p a rt o f o u r h id d en interpretive schem a and the m o ti­ vation in o u r w ork. O u r nostalgia for th e “g o o d old days” w hen we did fieldwork o ften inspires us in o u r teaching and w riting, and in how we design o u r fu tu re research. B eing a p articip an t ob serv er m ay be less a b o u t try in g to live o th e r p eo p le’s lives th an a b o u t th e natural h u m an activity o f play as a m eans o f self-discovery an d o f u n d e rsta n d in g others. M y presence as a foreign scholar studying p u ppets was also m an ip u ­ lated consciously and unconsciously in m ore form al ways. M o st o f the form al research trips I c o n d u c te d on Awaji w ere d irected tow ard studying th e p ast o f this tra d itio n , tracking dow n elderly people, old p h o to g ra p h s, p u p p e t heads, costum es and paraphernalia, an d old d o cum ents. Initially, as an h istorian o f religions, my in terest lay in discerning th e stru ctu res and in te n tio n s o f th e ritual tra d itio n o f p u p p e try as an exam ple o f w idespread religious practices and beliefs at various po in ts in Japanese history. D u rin g each o f these periods o f research, I was o ften taken to o th e r parts o f Japan, from K yushu to N agano. B u t th re e events, one in 19 8 8 , a n o th e r in 1991, and a th ird d u rin g my m o st recen t trip in 1 9 93, stand ap art from the rest o f m y research experiences. O n these trips o ff Awaji, o r w henever Aw aji’s place in a larger Japanese c o n tex t was central (as in th e su m m it m eetings),

w hat I was doing on Awaji was always an im p o rtan t issue. A whole new arena for perform ance was opened up, and the focus o f my study u n d er­ w ent a dram atic shift. This enabled me to ask a different series o f questions ab o u t the m eaning o f trad itio n in contem porary Japan, and to explore the pow er o f an age-old malaise, nostalgia, to m otivate, m anipulate, and m a­ neuver activity in rem arkable ways.

Event I In A ugust 1988, I accom panied the tw o leading scholars o f th e Awaji puppet tradition, Nakanishi H id eo and Takeda Shin’ichi, to N agano P re ­ fecture. B oth o f these m en are high school teachers on Awaji and were born and raised on the island. M r. Nakanishi is in his m idforties and is the director o f the M ihara-gun H igh School P u ppetry C lub, an organization w hich trains young people in th e a rt o f joruri recitation, puppetry, kashira carving and costum e creation, and sham isen, and w hich supplies artists for the professional th eater on th e island. Mr. Takeda is in his early sixties, and before retirem ent he tau g h t early Japanese history. B oth publish ex­ tensively in Japanese journals and are great aficionados o f the ningyo joruri trad itio n .49 The purpose o f our trip to N agano was ostensibly to attend the In te r­ national P uppetry Festival staged each year in Iida-shi. This event, sp o n ­ sored by U N IM A ,50 attracts puppetry groups from Australia, E urope, V ietnam , Japan, and even the Americas. We w ent by car, leaving Awaji at m idnight and catching the first ferry to Kobe so th a t we could arrive in N agano by m idm orning. T he real purpose o f this trip, I soon discovered, was to track dow n the graves and rem nants o f a Tokugawa puppeteer from Awaji w ho had settled in N agano and started a puppetry group there. T he m an had died in the 1700s, and m y colleagues w anted to find o u t ab o u t his life and the influence he had on the p uppetry traditions in N agano. We knew th e nam e o f th e tem ple w here his m ortuary tablets were kept and the cem etery w here his bones are said to have been buried. In the heat o f an A ugust afternoon, we spent several hours in a m ountain cem etery trying to find the actual gravestone o f this Awaji puppeteer. O n the surface, finding his grave was a way o f proving th e connection betw een Awaji puppetry and N agano puppetry. Situating Awaji puppets as the center o f ritual p uppetry in Japan is an im p o rtan t dynam ic o f the process o f revival. At an o th er level, it becam e clear th at for these scholars, this was a visit to the past, and the effort to find this grave was m otivated by m ore than academ ic or “political” reasons. Finding this grave was an attem p t to

make some sort o f contact with this dead puppeteer. W hen they actually found the grave, both o f the m en solemnly cleaned o ff the stone and made offerings. W orshipping and visiting a grave in Japan entails m ore than paying one’s respects. It says som ething about the relationship o f the w or­ shipper to the deceased; it implies a tie. The visit to the grave was about forging a bond between the dead and the living, between the past and the present. And this involved no small am ount o f sweat and inconvenience. Searching for and w orshipping at a rem ote m ountain grave became the central drama o f the trip to N agano, rather than the U N IM A festival. For Nakanishi and Takeda, the visit to the grave bro u g h t the story o f this puppeteer into their lives. W hen we w ent to N agano in 1988, I was in the m idst o f my doctoral field research, and I was presented in all m eetings and gatherings simply as “a Ph.D . student from the University o f C hicago” writing a dissertation on the Awaji tradition. M eetings with leading perform ers and scholars o f ritual puppetry around Japan were enhanced by having a foreign scholar studying the Awaji tradition. My presence gave Awaji ningyo an extra prestige w ithin the w orld o f ritual traditions. D uring research trips o ff the island o f Awaji, however, I quickly learned that different replies were re ­ quired o f me when asked why I was doing a doctoral research on Awaji ningyo. The proper answer depended on the audience. I f I were asked this question by people from established theaters in urban areas (o f which the N ational T heater is the best example) or by a m em ber o f the M inistry o f Education, I remarked how this tradition was an excellent example o f “true Japanese ritual tradition,” how the plays were m ost poignant, and how the rural versions o f the theaters had a rough and authentic feel to th e m . I was to respond w ith the formulaic expression that a retu rn to the rural and the marginal was in fact a retu rn to the authentic. It was also acceptable to m ention th a t marginalized groups which had been ignored should now be seen for w hat they are— true Japanese performance. This interpretation seemed to have been agreed upon as the reason a great deal o f money was being channeled into revival efforts by various organizations and governm ent agencies. If, however, my audience consisted o f oth er ritual puppeteers o r people directly connected with the various traditions around Japan, this previous answer merely drew smiles and knowing winks. T he answer required by this audience was to point o u t the centrality o f the Awaji tradition as a unique nexus and source tradition for m ost ritual puppetry traditions around Japan (which is in fact correct). Both o f these responses rely on the hidden prestige o f Awaji ningyo being perceived as im portant by an outsider. The only difference in the answers is w hat that foreigner is perceiving.

Event 2 A nother case from my field notes indicates the prestige o f outsider appre­ ciation. A few years later, Ph.D . in hand and hired at an Am erican univer­ sity, I had a different status vis-a-vis the traditio n I studied. I was traveling arou nd Kyushu during the preparations for the ningyb sum m it. A lthough I was essentially extraneous to the discussions (since they largely centered around m oney, funding for travel, tim ing), at each theater I was in tro ­ duced as the Am erican scholar w ith a recent Ph.D . (now a professor at an Ivy League school) w ho had done a dissertation on the Awaji theater. A lthough the in tro d u ctio n was usually followed w ith such hum ility m ark­ ers as “They give P h .D .’s for things like th a t in Am erica, od d , I know, b u t anyway she did her Ph.D . on us,” it was clear th a t I was an im p o rtan t stage prop in this perform ance. Since a m ajor co m p o n en t o f these trips was to solidify Awaji as the center o f p uppetry in Japan (aside from the classical Bunraku stage), my presence (and the extra clout o f my status as an aca­ dem ic) was im p o rtan t political capital.

Event 3 At the m ost recent trip to Japan, I was keynote speaker at the th ird annual ningyb sum m it. This event, unlike the previous tw o, was to draw signifi­ cant m edia attention. This three-day event focused on the “revival” o f the Sanbasb perform ance, and several puppetry groups from around Japan presented their perform ances o f Sanbaso side by side for com parison. The m edia billed it as “ a battle o f the Sanbasbs,” and the th eater was packed. T he ratio o f photographers to regular theatergoers was unusually high, and film crews rivaled one another for the best angles in the theater. It was difficult to see and even m ore difficult to hear the perform ances, and it was clear this was a “ m edia event” and a rare p h o to o p p o rtu n ity for the p upp etry groups to be on national news. A foreign professor as the key­ n o te speaker lent a fu rth er air o f gravity to the event. Again, it becam e clear th a t my foreignness was an im p o rtan t co m p o n en t in this three-day m edia event. This was m ost apparent w hen I looked at the m edia coverage o f the sum m it on national news. In spite o f th e fact th a t there had been a n um ber o f rare perform ances (in the m odel o f the M inzoku G einb Taikai th a t Sasahara lam ents), m ost snippets on the national news gave as m uch coverage to my keynote address as they did to the rem arkable p e rfo rm ­ ances, a regrettable reflection o f w hat makes news in Japan even in these days o f k o k u sa ik a (“internationalization,” the recent buzzw ord o f the M inistry o f E ducation).

U nlike the previous events w here I had a heavily scripted role to play, at this sum m it I was given free rein to speak o n w hatever topic I chose. (I spoke o n th e m eanings o f revival and th e significance o f nostalgia as a m o tiv atin g force.) W hile on one level m y com m ents c ould be seen as “d e ­ c o n stru c tin g ” the very event itself, th e atm osphere was o pen, and th e im ­ plicit m essage was th a t Awaji was now so clearly the cen ter o f this revival m ovem ent th a t it could risk th e unscripted and even critical com m ents o f a foreign scholar. T h ro u g h o u t these th ree experiences, people in Awaji used my foreigner status to fu rth e r th eir ow n revitalization m ovem ent. T his experience could n o t p o in t o u t m ore directly ho w th e presence o f th e observer affects the o u tco m e o f events. W hile it w ould be grandiose and a rro g a n t to claim any credit for the revival o f this trad itio n as a viable th ea ter, it is clear my presence— as an invisible and c o n stitu te d O th e r— gave stre n g th to p a rtici­ pants in this process.

Fragm ents to th e L ig h t I t w ould seem sim ple e n o u g h to reveal h o w h id d en m eanings in the Awaji p u p p e try revival are c o n stru c te d and m anipulated. O nce we have u n d e r­ sto o d these d eep er ideological functions, we can be c o n te n t th a t we have u n d e rsto o d the process itself. T he pro b lem w ith such an approach is th a t it ignores a very im p o rta n t p a rt o f this entire process: th e sense o f identity o f th e people involved. F o r U m azum e-san, as well as m any o f m y o th e r inform ants, th e larger social and political c o n tex t o f th e d e n to geino b u n iu — the boom in folk arts in Japan— c a n n o t exhaust the d eep er m ean ­ ings o f th eir actions. Revival, it seem s obvious e n o u g h , encodes ideologies ab o u t value, taste, aesthetics, and class, and it appeals to an easily m anipulated aspect o f us all— the desire to u n d o tim e and re tu rn to th e past. D eciding ju st w hat will be retrieved, w hat will be left o n th e du st heap o f history, and how this past shall be p resen ted constitutes a hierarchy o f m eaning. Revival also appeals to people at th e level at w hich we are least inclined to be suspi­ cious— the past. In reflecting on how th e Awaji trad itio n has been revived, I have tried to identify h o w certain aspects o f th e tra d itio n have been req u ired to b e ­ com e sym bols for th e entirety o f th e “vanishing p ast.” F or different groups o f people, criteria for those decisions differ, at tim es dramatically. T h e process o f sym bolizing th e past is always in m o tio n , resp o n d in g to subtle changes in larger discourses o f id entity in Japan. T he uses o f nostalgia in th e revival o f Awaji p u ppets c an n o t be u n d e r­ sto o d sim ply as a uniform ideological appropriation o f the past, for the

p uppetry tradition has had different pasts for different people, depending on their relationships to b o th the tradition and one an o th er in various pow er structures. This revival is a vibrant example o f an ever developing discourse on w hat “ being Japanese” m eans to different people, appropri­ ated in any num ber o f ways. This river o f nostalgic discourse has had three currents, at tim es confluent: ( I ) people living on Awaji w ho as children o r young adults w itnessed (and even helped to bring about) the end o f the “island o f the p u p p ets,” and for w hom the associations w ith puppetry have been sim ultaneously a source o f regional pride and social stigm a; (2) “outsiders” in Japanese m edia, arts, and academ ia, for w hom this revival provides a “living m useum ” from w hich they can draw m aterials for the business o f m arketing the past o r furth erin g their academic careers; and (3) people in governm ent, m ost specifically the M inistry o f E ducation and Bureau o f C ultural Affairs, for w hom the celebration o f Awaji pu p p etry as a folk perform ing art makes it possible b o th to posit a m ythology o f a com m on past shared by all Japanese people th ro u g h an appeal to the idea o f kokumin (national folk) and to camouflage a history o f oppression o f Japanese outcast groups. In this chapter, I have discussed how the Awaji tradition b o th partici­ pates in and moves beyond the significance o f this larger m ovem ent in Japanese society in the latter p art o f the tw entieth century. F or the revival m ovem ent o f Awaji ningyo to succeed, a delicate balance has to be struck. O n one level, the revival is ab o u t going “m ainstream ” with a tradition which has in the past been the ro o t o f m arginalization. O n another level, reviving Awaji puppetry helps create a local to urist trade and a new id en ­ tity for Awaji as Japan’s “ M editerranean.” Revitalization encompasses a larger ideological construct o f finally becom ing p a rt o f m ainstream Japa­ nese folk a rt culture, albeit under the rubric o f a som ew hat fictitious category. Revitalization encom passes such issues as identity, liberation, autonom y, and, o f course, econom ics. It is also ab o u t finally claiming an identity and a relationship w ith o n e ’s ancestors. I have tried to show how th e nostalgic discourse in Japan in the last several decades has played an im p o rta n t (and at tim es even dom inant) role in the revival o f Awaji puppetry. B ut another, som etim es ironic and partly unconscious m ovem ent has been afoot. To dismiss the entire com plex process o f recasting ritual perform ances as stage pieces solely as a p art o f a nostalgic enterprise misses an im p o rtan t point. Mr. U m azum e and o th er people o n Awaji are cognizant o f the fact th a t the contexts and intentions o f kadozuke ningyo are no longer extant in Japan. Ostensibly, according to the larger nativist discourse pro m o ted by the Bureau o f C ulture and the M inistry o f E ducation, the rituals are to be kept alive for tw o reasons: they open a w indow to the past and they serve as a source for creative new perform ances.

U m azum e is fully aware o f the lim itations o f keeping these ritual p e r­ form ances alive by casting th em in an artificial context, and he realizes th at they are no lo n g er rituals b u t are now in fact theater. B ut this, he says, is preferable to lettin g th e a rt form s be lost altogether. M r. U m azu m e, like m any oth ers o n Awaji, feels th a t the actual sym bolic c o n te n t o f these r itu ­ als— even w hen rem oved from th eir contexts— m ay be p o te n t e n o u g h to transform people. T h o u g h th e distance created by th e form ality o f th e stage may dilute th e p o ten tial for a transform ative experience, the rites’ pow er to affect people persists. A t a deeper level, th e fact th a t ou tcast arts are suddenly being heralded as “ authentically Japanese” has a healing effect. M any people on Awaji w ere deeply attached to th eir p u p p e try tra d itio n , and w hen in th e tw e n ti­ eth cen tu ry it was reg ard ed as so m eth in g no lo n g er useful o r necessary b u t m erely a source o f em barrassm ent and stigm a, people felt g reat conflict a b o u t lettin g it die (o r even actively killing it). T his new era, for all its problem s, has allow ed m any people to claim w h a t th eir parents, gran d p ar­ ents, and even great- and g reat-g reat-g ran d p aren ts did as so m eth in g m eaningful, beautiful, and even Japanese. In the s h o rt p eriod o f tim e I have had c o n ta c t w ith people o n Awaji (from 1 9 7 7 u n til now — a little u n d er tw enty years), people have gone from denying th a t they had any co n n ectio n w ith D o k u m b o m awashi in th e past to show ing m e w ith great pride the rem nants o f th eir family p u p p e t collections o r paraphernalia. People w ho in th e early 1980s to ld m e th eir ancestors had been farm ers have since com e to m e and said, “ D id I say my grandfather was a farmer? W ell, he was, b u t he also m anipulated puppets. I guess I fo rg o t to m en ­ tion th a t.” Perhaps a larger discourse o f exoticism and nativism created the c o n tex t for th e revival o f kadozuke n in g y o . Perhaps th e revival is an awkw ard and hopeless false sta rt w hich will never overcom e th e lim ita­ tions o f casting ritual o n to th e stage. M aybe a lo t o f it is a b o u t Awaji regionalism and th e to u rist industry. A nd perhaps the larger nativist dis­ course will once again reabsorb this new found affirm ation o f th e history o f ritual p u p p eteers and tu rn it in to so m ething else, so m eth in g divisive. B ut for th e tim e b eing, a lo t o f people o n Awaji are able to situate th e m ­ selves in relationship to th e lives and livelihood o f th eir parents, g randpar­ ents, and great-g ran d p aren ts. T hey have been able to pick up th e frag­ m ents o f evidence from these nearly invisible p u p p eteers’ difficult and often painful yet tale n te d lives, h o ld th em up to th e light, and ask, “ H o w is this p a rt o f my story?”

U m azum e M asaru at O m ido H achim an in 1990.

E p ilo g u e

S i n c e th e research for and w riting o f this b o o k w ere com pleted, a m ajor

earthquake hit H y d g o p refecture in Japan. Awaji was th e epicenter o f the quake. W hile n o n e o f th e people c o n n ected w ith th e th e a te r lost their lives o r hom es, th e dam age o n Awaji was considerable. T h e m ain highw ay w hich traverses th e island was repaired only in M arch 19 9 6 , a full fo u rteen m o n th s after th e quake. As a result, th e business at th e th e a te r has been very bad, and people are again w orried th a t Awaji p u p p e try will be n o th in g m ore th an a passing to u rist attractio n . W hen I spoke w ith U m azum e-san in M ay 1 9 95, he was pessim istic a b o u t th e revival being anything o th e r th an a short-lived dream . W ith th e rec o n stru c tio n o f the highw ay and th e so o n -to -b e c o m ­ pleted bridge betw een H o n sh u and Awaji, people are hopeful th a t Awaji puppets will draw larger Japanese audiences th an before. O th ers en te rta in serious d o u b ts th a t th e th e a te r will survive th e setback caused by the earthquake. In 1 9 95, U m azum e-san retired as d irecto r o f th e th eater, and his suc­ cessor was chosen from am ong th e y o u n g er puppeteers. B ando S entard was th irty -o n e years old w hen he assum ed leadership o f th e theater, the same age as U m azum e-san w hen he began try in g to revive th e trad itio n . This is considered to be a lucky coincidence by people in th e theater. I wish him well.

Notes

In tro d u ctio n : O f Stories and Fragm ents 1. I t is co m m o n in Japanese to use th e term jd r u r i to refer to p u p p e t theater, a lth o u g h th e te rm literally m eans “ pure lapis lazu li.” T he co n n ectio n betw een p u p p e try an d this nam e com es from th e fam ous love story o f Lady Joruri and U shiw aka (th e b o y h o o d nam e o f Y oshitsune), w hich m ade this genre o f ballad recitation fam ous. See C. J. D u n n , The E arly Japanese Puppet D ra m a (L o ndon: L uzac, 1 9 6 6 ), pp. 7—13. In Japanese, see th e exhaustive study by W akatsuki Yasuji, “ Joruri him e m o n o g a ta ri,” chap. 2 in Joruri-shi no kenkyil (Tokyo, 1943). 2. Jane M arie Sw anberg-Law , “ P u p p ets o f th e R oad: R itual Perform ance in Japanese F olk R eligion” (P h .D . diss., U niversity o f C hicago, 1990). 3. F ro m 1 9 8 7 th ro u g h 19 8 9 , I was a researcher at T sukuba U niversity in the In stitu te o fP h ilo so ph y, fu n d ed by a research scholarship from the Japanese M inis­ try o f E d u catio n . D u rin g these tw o years I sp en t large segm ents o f tim e o n Awaji d o in g research. 4. F o r a discussion o f narrative know ledge, see Jean Franyoisc L y o tard , The Postmodern C ondition: A R ep o rt on Knowledge, trans. G eo ff B en n in g to n and Brian M assum i (M inneapolis: U niversity o fM in n e so ta Press, 1 9 8 3 ), pp. 18—23. 5. H o ri Ich iro , Folk R eligion in Japan: C o n tin u ity a n d Change, ed. Joseph M . Kitagaw a and Alan M. M iller (C hicago: U n iv e rsity o f C hicago Press, 1 9 6 8 ), p. 10. 6. Ib id ., pp. 1 0 -1 1 . 7. Joseph M . Kitagaw a, “ ‘A Past o f T hings P resen t’: M ajor M otifs o f Early Japanese R eligions,” in O n U nderstanding Japanese R eligion (P rinceton: P rince­ to n U niversity Press, 1 9 8 7 ), p. 48. 8. Ib id ., pp. 4 3 -5 8 . 9. Ib id ., p. 70. 10. D u rin g th e years th a t I stu d ied w ith Professor Kitagawa, h e o ften m ade th e p o in t th a t th e future o f Japanese scholarship lay in finding ways to bring the issue o f “ o th e rs” to th e fo reg ro u n d . It was P rofessor Kitagawa w h o , in n u m erous conversations we h ad n ear th e en d o f his life, en couraged m e in my study o f Japanese o utcasts and ritual p u p p eteers, precisely because he saw the value o f w hat he considered to be a “ corrective” to th e unitary m eaning stru c tu re approach to Japanese religions. 11. H . D . H a ro o tu n ia n , Things Seen a n d Unseen: Discourse a n d Ideology in Tokugawa N a tiv ism (C hicago: U niversity o f C hicago Press, 1988), pp. 4 1 9 -4 2 0 . 12. L y o tard , The Postmodern C ondition, p p . 8 1 -8 2 . C hapter I : In th e Shape o f a P erson I . M o st studies o f Japanese p u p p e try focus alm ost exclusively o n m eth o d s o f m anip u latio n , to th e n eglect o f o th e r aspects o f puppetry. An excellent and infor­ m ative exam ple o f this ap proach is D u n n , Early Japanese Puppet D ra m a , w hich

explores th e early joruri tradition. H e m entions ritual puppetry and religion only in passing on pp. 2 1 -2 6 . 2. This co m m en t was m ade to m e on D ecem ber 2 3 ,1 9 8 8 , w hen I interview ed Professor N agata at the nursing hom e in Yudawara where he was living. H e was then ninety-nine years old. 3. E. G o rd o n C raig, “T h e A ctor and the tJb erm ario n e tte ,” in E. T. Kirby, Total Theater (N ew York: D u tto n , 1969), pp. 5 4 -5 6 . This essay was originally printed in The Mask I , no. I (April 1908). It also appears in !Michael J. W alton, C raig on Theater (L ondon: M eth u en , 1983). I have relied on the latter transcrip­ tion o f this fam ous essay. 4. Bil Baird, The A r t o f the Puppet Theater (New York: M acM illan, 1965), p. 13. 5 . In M argareta N iculescu, ed ., The Puppet Theater o f the Modern World, trans. EwaId Osers and Elisabeth Strick (B oston: Plays, Inc., 1967), p. 58. 6. This is a con ten tio u s topic in Japan. It seems m ost likely th at the use o f puppets arose o u t o f a need to dissim ulate artistic responsibility and risk. 7. Ibid., p. 35. 8 . W ith m inor variations, this story is reco u n ted in m ost o f the m ajor works on C hinese shadow theater. See Rene Sim m en, The World o f Puppets (N ew York: T hom as Y. Crow ell, 1972), p. 79; also Genevieve W im satt, Chinese Shadow Shows (C am bridge: H arvard U niversity Press, 1936), pp. ix-x. 9. Craig, “ A ctor and tJb e rm a rio n e tte ,” pp. 3 8 -5 0 . 10. N iculescu, Puppet Theater, pp. 1 8 -1 9 . 1 1 . Paraphrased from L o tte R einiger, Shadow Theatre an d Shadow Films (L ondon: B. T. Batsford, 1970), pp. 2 0 -2 1 . 12. N agata Kokichi, Nihon no ningyo shibai (Tokyo: Kinseisha, 1974), p. 2. 13. C arm en Blacker, am ong others, has suggested th a t the female clay effigies found in these tom bs are actually the representations o f miko (ancient female sham an). To su p p o rt this quite probable hypothesis, she points to the headgear, posture, and ornam entatio n o f the figures. O n e such haniwa, discovered in O kaw am ura in G um m a Prefecture wears a perfectly flat headboard, and is sitting in a m ost unusually up rig h t position w ith a focused and peaceful expression, sug­ gesting the posture o f one in an in ten t trance. A round her neck, she wears beads, and from her waist hangs a mirror. A n o th er figure, similarly adorned and in the same posture, wears m a g a ta m a , the com m a-shaped jewels believed to have spiritual pow er in ancient Japan. T h e beads, similar to those found in tom bs th ro u g h o u t Japan, surely had a pow erful magical role as spirit lures, inviting the kami to dwell in th e body o f th e sham an. T hese m agatam a ornam ents were origi­ nally m ade o u t o f either b one or animal teeth and eventually were carved o u t o f stone and sem i-precious stones. I t is possible th a t the origins o f their shapes stems from a belief in the pow er o f animal spirit helpers, w hose teeth gave shamans special pow ers in their rituals. Blacker and others have com pared this ritual attire to th at o f the Siberian sham an. C arm en Blacker, The C atalpa Bow: A Study o f Sham anic Practices in Japan (L ondon: G eorge Allen and U nw in, 1975), pp. 1 0 6 -1 0 9 . T ungusic sham ans in n o rth ern M ongolia used copper m irrors as a de-

vice to see the soul a dead person. C oncerning this, Eliade writes that such copper mirrors are clearly Sino-M anchurian in origin, and while their meanings differ from tribe to tribe, “ the m irror is said to help the shaman ‘see the w orld’ (that is, to concentrate), or to ‘place the spirits,’ or to reflect the needs o f m ankind, and so o n .” Mircea Eliade, Shamanism: Archaic Techniques o f Ecstasy, trans. Willard R. Trask, Bollingen Series, no. 76 (Princeton: Princeton University Press, 1964), pp. 153—155. Ryuzo Torii also maintains that northeast Asian shamans suspend mirrors from their waists, which give o ff light and jangle as they dance, giving the mystical, synesthetic effect o f dance, light, and sound. In the case o f the mikos, the female shaman, this sound and light may have served to invite the kami to enter her body. The history o f Japanese perform ing arts and the Japanese religious conception o f possession {kam i-gakari or tsukimono) certainly support this hy­ pothesis. M irrors also have a wider m eaning in Japanese religious practice. Often it is a m irror that serves as the vessel (shintai, yorishiro, or mitamashiro) for the kami during a rite. 14. A standard interpretation o f this term has been that there were seven rivers, but since the character used refers to shallows within a river, it appears more likely that there were seven parts o f a particular river in which these rites could take place, places where the current slowed enough to create a safe ritual site. These places along the river were also understood to be sacred sites. It is no t known which river was the actual site for these rites, although Yamagami Izum o suggests that it may have originally been the Kamo River. See “Nanase harai no genryu,” Kodai b u n k a 2 2 , no. 5 (August 1970): 126-127. 15. For a discussion o f this practice, and a transcription o f the kam bun text describing it, see Kitamura Tetsuro, “N ingyo,” Nihon no bijutsu, no. 11 (March 1967): 1 -2 9 ; also Nagata, Nihon no ningyo Shibai, pp. 19-25. Tsunoda Ichiro also m entions the relationship between puppets and the removal o f pollution. See his Ningyo geki no seiritsu ni kansuru kenkyii (Osaka: Kuroya Shuppan, 1964), p. 248. 16. Yamagami, “Nanase no harai no genryu,” p. 130. 17. Yamagami Izum o, Miko no rekishi (Tokyo: Yuzankaku, 1984), p. 26. 18. Kitamura, “N ingyo,” p. 30. 19. Nagata Kokichi, “Ningyo shibai no kigen shiryo,” in Geino ronshU ed. HondaYasuji (Tokyo: Kinseisha, 1976), pp. 635-6 3 7 . This article by Nagata also presents some clear photographs o f extant hitogatas from various archaeological museums around Japan. 20. A photograph o f this figure appears in ibid., p. 30. 21. See Nagata, Nihon no ningyd shibai, pp. 25-28. 22. Murasaki Shikibu, The Tale o f Genji, trans. A rthur Waley (London: George Allen and U nw in, 1935), p. 366. Waley writes, “The Heavenly Children were dolls which were intended to attract evil influences and so save the child from harm ” (p. 366n). 23. “Ise ke hisho tanj5 no ki” is cited by Kitamura, who argues that it was believed that if during pregnancy the doll grows hair, the child will be a girl and if not, a boy. I have found no sources to support this claim. Kitamura, “ N ingyo,” pp. 3 0 -31.

24. Effigies at th e N ishinom iya Ebisu shrine, discussed in detail in chapter 3, were venerated as capable o f p ro tectin g tiny infants from smallpox. 25. T he ten types take th eir nam es from villages o r resorts in th e T o h o k u re ­ gion: Tsuchiyu, T o g atta, N ag u ro , Yajiro, Yam agata-Sakunam i, Zao-Takayu, Hijiori, N am bu (H anam aki), Kijiyama, and T sugaru. F or a clear and w ell-illustrated description o f the ten , see C hizuko Takeuchi and R oberta Stephens, A n In v ita ­ tion to Kokeshi Dolls (Sendai: T sugaru S h o b o , 1982). 26. William LaFleur, L iq u id Life: A bortion a n d Buddhism in Contemporary Japan (Princeton: P rinceton U niversity Press, 1992), pp. 9 9 -1 0 7 . L aFieur’s dis­ cussion o f ab o rtio n and infanticide in Japan is the m ost extensive study available in English. 27. This com m en t was by the T okugaw a C onfucianist Sato N o b u h iro (1 7 6 9 1850). Ib id ., p. 107. 28. O ne way th a t these “ erased children” w ere allow ed a presence in the lives o f people in T ohoku was in the form o f the zashikiwarashi (child o f the zashiki), also called zashiki boko o r zashiki bokko, a small and mischievous spirit ab o u t which stories w ould be to ld , and w hich was said to be always present in a house, playfully hau n tin g its inhabitants. 29. See LaFleur, L iq u id L ife, pp. 9 7 -1 0 2 . 30. T h e leading auth o rity o n Sado p uppetry is the elderly Y am am oto ShQnosuke. H is authoritative book is the only w ork in Japanese devoted wholly to p u p ­ petry on this island. See Yam am oto Shflnosuke Sado no ningyd shibai (N iigataken, S ado-gun, M ano-cho, Shinm achi: Sado ICyodo Kenkyukai, 1976). In the sum m er o f 19 9 1 , w ith a faculty research gran t from the C ornell U niversity East Asia Program , I co n d u cted research on th e island o f Sado and interview ed Mr. Yamamoto. In his hom e he has one o f the finest collections o f Sado puppets, including rare puppets n o t found in m useum s. 3 1 . Because Sado p u p p etry was n o t central to my research, I only spent a sh o rt period o f tim e on Sado in th e sum m er o f 1 9 9 1 . Since I did n o t know anyone well, it is likely th a t th e d earth o f inform ation I received a b o u t these sex puppets re ­ flected the em barrassm ent m any people felt in talking ab o u t these puppets w ith a stranger, let alone a foreign w om an. 32. Y am am oto, Sado no ningyd shibai, p. 126. 33. D onald L. Philippi, trans., K ojiki (Tokyo: U niversity o f Tokyo Press,

1968), pp. 81-86. 34. Ib id ., p. 84f. 35. Ibid. 36. Yam am oto goes on to say th a t a p u p p et exactly like this Kinosuke, w ho urinates on the audience, can be fo u n d in Korea. 37. This com m en t was m ade to me by the m aster ningyd-shi, O e M inosuke. I interview ed O e M inosuke at his hom e in N aruto-shi, T okushim a-ken, in A ugust 1988. Masks in th e N o h th eater were also said to be “well b o rn ” if they held spirits. 38. In th e joruri trad itio n , a pup p eteer is responsible for gettin g and assem­ bling his ow n puppet. This process is referred to as “helping the pu p p et find its body.”

C h a p te r 2: K adozuke 1. This description is based on interviews, ethnographic films and descriptions, and old photographs I collected during my fieldwork. For the purposes o f this discussion, I have com posed a general description o f a puppeteer presenting a Sanbaso perform ance, the m ost com m on type o f ritual perform ance using puppets during the New Year. I discuss the actual content o f this type o f performance in detail in chapter 5. 2. To a certain extent, these names referred to different styles o f puppetry. 3. Bernard H arrison in his Form and C ontent (New York: H arper and Row, 1973) discusses what he calls the “discrimination thesis,” namely, “the m ost fun­ damental level o f discourse about experience is the level at which we state facts, so far as they are known to us, about our own and others’ ability to discriminate between particular types o f present stimuli. If we ask w hether there is no t some more fundamental level at which we can describe the phenom enal character o f experiences between which we discriminate, the reply m ust be that there is not. Experience talk boils down ultimately to discrimination talk” (p. 4). 4. Michael Theunissen, The Other: Studies in the Social Ontology o f Husserl, Heidegger; Sartre a nd Buber, trans. C. M acann (Cam bridge: M IT Press, 1984), p. I. 5. For this reason, it seems that the standpoint o f history o f religions is no t incompatible with contem porary critiques o f ideological systems. 6. Early in the history o f contem porary anthropology, studies o f non-W estern peoples as “primitive” often isolated the category o f the stranger and the fear o f strangers as a unique feature o f “the primitive, savage m ind.” Consider the fol­ lowing quote from the Encyclopedia o f Religion and Ethics (New York: Charles Scribner’s Sons, 1934), s. v. “strangers” : “We have seen that, to the savage, the world which lies beyond the com m unity to which he belongs— i.e. beyond his group and the groups associated with it on terms which are friendly rather than hostile—is a world strange and mysterious, peopled with beings whom he hates or fears as his deadly foes. H e thinks o f them as belonging to an order other than his own, as less or, it may be, as more than hum an; and he looks upon them as abso­ lutely rightless; for the sphere o f rights is codeterm inous with the sphere within which he him self lives. As regards himself, life is possible for him only within the little circle o f his com m unity.” 7. F or a thorough analysis o f this phenom enon, see Edward W. Said, Oriental­ ism (New York: Pantheon Books, 1978). 8. See H arootunian, Things Seen and Unseen. 9. See Ernst van Alphen, “The O ther W ithin,” in Alterity, Identity, Image: Selves and Others in Society and Scholarship, ed. Raymond Corbey and Joep Leersen (Amsterdam: Rodopi, 1991), pp. 1-16. 10. Michel Foucault, Madness and Civilization: A History o f Insanity in the Age o f Reason (New York: Pantheon Books, 1965), pp. 8 -9 . O ne doubts w hether being mentally ill and homeless could have been “easy,” regardless o f the freedom this provided.

11. H ere, D ouglas is in line with T u rn er in his discussion o f liminal states in ritual processes. See V ictor T u rn er, The R itu a l Process: Structure an d A n ti-S tru c ­ ture (Ithaca: C ornell U niversity Press, 196 9 ), pp. 9 4 -1 3 0 . 1 2 . M ary D ouglas, P urity a n d D anger: A n Analysis o f the Concepts o f Pollution a n d Taboo (L o n d o n : R outledge and Kegan Paul, 1966), pp. 1 1 4 -1 1 5 . 13. Ib id ., p. 122. 14. C om prised o f elaborate rules and ritual behaviors for dealing w ith the natural functions and processes o f th e h um an body and its p ro d u cts, ritual purity systems are a d o m in an t feature in m any religious systems th ro u g h o u t th e w orld. Im p o rtan t studies responsible for setting o u t th e general categories o f in terp reta­ tion are Louis D u m o n t’s study o f the H in d u caste system, H om o Hierarchicus: A n Essay on the Caste System (C hicago, U niversity o f C hicago Press, 1966), espe­ cially chapter 2 , and D ouglas, P urity a n d Danger. O th e r recent studies include Em iko O hnuki-T ierney, Illness a n d C ulture in C ontem porary Japan (C am bridge: C am bridge U niversity Press, 1984), R o b ert Parker, M iasma: Pollution a n d P u ­ rification in Early Greek Religion (O xford: O xford U niversity Press, 1983), and H ow ard E ilberg Schw artz, The Savage in Judaism : A n Anthropology o f Israelite Religion a n d A n c ie n t Ju d a ism (B loom ington: Indiana U niversity Press, 1990). 15. Yokoi Kiyoshi, Chiisei minshii to seikatsu bunka (Tokyo: Tokyo D aigaku Shuppan Kai, 1975), p. 267. O hnuki-T ierney (Illness a n d C ulture, pp. 2 1 -7 4 ) explores how notio ns o f purity and pollu tio n , insider and outsider, and above and below continue to operate as essential distinctions in Japanese society. 16. A recent explanation for th e rebuilding o f the shrine concerns the need for builders to be trained in th e co rrect techniques o f construction. Since rebuilding m ust be done every tw enty years, there will always be persons alive w ho know how to rebuild the shrine. 17. O hnuki-Tierney, Illness a n d C ultu re, p. 37. 18. Ib id ., p. 269. 19. See N am ihira E m iko, K egareno kozo(Tokyo: Seidosha, 19 9 2 ), particularly chapters 4 and 5. See also Yokoi, ChUsei minshii to seikatsu bunka, pp. 2 6 7 -2 9 4 . 20. Ib id ., p. 267. 21. Yokoi, Chiisei minshii to seikatsu bunka, p. 269. 22. Q u o te d in Yamagami Izu m o , Miko no rekishi, p. 27. 23. Yokoi, Chiisei minshii to seikatsu bunka, p. 280. 24. Philippi, K ojiki 1:1. T he literal sentence “they hid th eir bodies” is tran s­ lated as “their forms were n o t visible” by Philippi. 25. Awaji is considered to be th e location o f O nogorojim a, and even today there is a sacred sp ot o n the island w here it is said th e heavenly jew eled spear first created th e self-curdling island. I discuss this in m ore detail in chapter 4. T he nearby island o f N ushim a (o ff th e so u th ern coast o f Awaji, also has a shrine said to be th e site o f O nogorojim a. 26. It has been suggested th at this is also a reference to the island o f Awaji. 27. This deity is also called H i no Kagu Tsuchi no Kami. 28. K ojiki 1:7, 18—22. 29. F or a sum m ary o f these in terpretations, see K ojikJ additional note 5. See also D aniel C. H o lto m , The N a tio n a l Faith o f Japan: A Study in M odern Shinto (1938; reprint, N ew York: Paragon Books, 1965), pp. 9 3 -1 2 1 . H o lto m

argues against a euhem eristic in te rp re ta tio n and prefers to see these m yths as expressions o f a “socialization o f experience w ith n a tu re ,” particularly th e w eather (p. 121). 30. K o jiki 1:7, 2 6 -2 8 . 31. Ibid. 1:8, 1 -5 . 32. Ib id . 1:8, 6 - 1 4 . 33. T h e obvious parallel w ith th e O rp h eu s an d E urydice m yth is striking. 34. K o jik i 1:9, 6 - 1 5 . It is in terestin g th a t th e corpse o f the m o th e r is divided in to th e sam e body p arts as th o se o f h e r so n , th e fire deity, after he is m u rdered by th e father. 35. Ib id . 1:11, I. 36. Ibid. 1:11, 1 -2 5 . 37. D u m o n t, H om o H ierarchicus, p. 47 . D ouglas has also p o in ted o u t this co m m o n m isu n d erstan d in g o f ritual purity systems. D ouglas, P u rity a n d D anger, pp. 2 9 - 4 0 , 73. 38. Yokoi, Chiisei MinshU no Seikatsu B u n k a , pp. 2 8 3 -2 8 7 . 39. L ouis D u m o n t, H om o Hierarchicus. 4 0 . Several surveys o f th e tran sfo rm atio n s o f th e m eanings associated w ith these specialists o f im p u rity are available in E nglish. O n e is the 1933 study by N inom iya Shigeaki “An In q u iry C o n cern in g th e O rigins, D ev elo p m en t, and P res­ e n t S itu atio n o f th e E ta in R elation to th e H isto ry o f Social Classes in Japan,” Transactions o f the A sia tic Society o f Ja p a n , 2 d ser., 10 (D ecem ber 1933): 4 7 -1 5 2 . In this early and o ften cited w o rk , N inom iya presents an overview o f th e m ajor theo ries p u t fo rth to acco u n t for th e existence o f an o u tcast g ro u p in Japan (called in 1933 by th e d e ro g a to ry w o rd eta, w ritten w ith the characters “ m uch Iiegare"), an d his discussion evaluates th e various theo ries, from the ridiculous to th e cred i­ ble. H e p roposes th a t hered itary o ccu p atio n led to th e institu tio n alizatio n o f the pariah status in Japanese society. A n o th e r is a b rie f discussion (sum m arizing Japa­ nese sources and draw ing o n N inom iya) in E m iko O hnuki-T ierney, Monkey as M irror: Symbolic Transform ation in Japanese H istory a n d R itu a l (P rinceton: P rin ceto n U niversity Press, 1987). O h n u k i-T iern ey has o u tlin e d th e tran sfo rm a­ tions in statu s o f these “specialists in im p u rity ” in Japanese history, w hom she refers to as “ special status p eo p le.” She show s n o t only how different g roups o f people fall in to this category for a variety o f reasons b u t also how the m eanings attach ed to o u tcast people as a w hole change dram atically th ro u g h o u t Japanese history. T h e general g ro u p o f p eo p le she calls special status people is actually com prised o f a w ide variety o f p eople, including those involved in im perial funer­ als, ritual purifiers, itin eran t perfo rm ers, m onkey trainers, puppeteers, pro stitu tes, m idw ives, latrine cleaners, stre e t sw eepers, m ausoleum guards, tatam i m akers, leath er tan n ers, falconers, sandal weavers, ink m akers, and indigo dyers am ong others. W h at all o f these people have in co m m o n is th a t th e ir professions were u n d ersto o d to be p o llu tin g according to the ritual purity system ou tlin ed above. T h e transform ations in th e ir status th ro u g h o u t Japanese history are due to tran s­ fo rm atio n s in th e im p o rtan ce an d m eanings associated w ith ritual purity itself, a p o in t th a t is tied directly to nativism and nationalism . A new w ork, M ichele M arra’s excellent study Representations o f Power: The Literary Politics o f M edieval Ja p a n (H o n o lu lu : U niversity o f H aw aii Press, 1 9 9 4 ), also discusses the issues o f

exclusion in ritual efficacy and provides valuable sources concerning the history o f Japanese outcasts. N one o f these excellent studies, however, focuses directly on the im portance o f th e im agination o f th e hum an body as a symbolic language o f society and th e cosm os alth o u g h the data clearly p o in t to the hum an body— its physiology, its representation, and its handling, as the central co m p o n en t o f the ritual purity system. R ecent studies by Japanese scholars have been critical o f the view th at outcasts arose o u t o f an anonym ous historical process. They argue th at the Tokugaw a gov­ ern m en t intentionally created and m anipulated th e statuses o f outcasts. O utcasts were m aintained in th eir liminal state th ro u g h carefully designed laws so th at they could serve as guards, executioners, and policem en to suppress peasant riots d u r­ ing th e dictatorial Tokugaw a shogunate. F or an overview o f this approach to Japan’s outcasts, see N agahara Keiji, N ihon no Chiisei Shakai (Tokyo: Iw anam i S hoten, 1973). In English see N agahara, “T he M edieval O rigins o f the E taH i n i n f Jo u rn a l o f Japanese Studies 5, no. 2 (1 9 7 9 ); Ian Neary, Political Protest a n d Social Control in Pre-W ar Japan: The O rigins o f B u ra ku Liberation (Atlantic H ighlands, N .J.: H um anities Press International, 1989), pp. 12—30. 41 . T hese people have at certain tim es in Japanese history had a very high status. I choose to call th em outcasts, however. I am fully aware th at this general term can n o t cover th e wide range o f historical transform ations th a t Japan’s pariah groups have u n dergone. I do so for several reasons: First O hnuki-T ierney’s term “ special status p eople” is to o euphem istic; it softens th e hard tru th . Second “o u t­ cast” draws atten tio n to th e im p o rtan t dynamics o f this case. 42. F o r a 1935 m ap w ith populations o f b u ra ku m in per thousand by prefec­ tu re, see Neary, Political Protest, p. 6. 43 . See M ihara Rekishi Kenkyukai, M ih a ra g u n shi, pp. 4 8 0 —488. 44 . Yoshida T eigo, “T he S tranger as G od: T h e Place o f the O utsider in Folk R eligion,” Ethnology 20 , no. 2 (1964): 88. 45 . T h e term w aza originally carried th e co n n o tatio n o f “unrevealed divine will” and cam e to include the idea th a t divine will could be responsible for scourges, bad om ens, and disasters. 46. C om bs w ere traditionally m ade o f to rto ise shell, so persons in this occupa­ tion were involved in th e killing o f animals. Killing turtles was also a symbolic violation (although th e products o f dead turtles w ere highly prized): a turtle was the symbol o f ten th o usan d years. 47 . Yokoi, Chiisei m inshu no seikatsu bunka, p. 335. K aw ara no mono literally means “ people w ho live along riverbeds.” 48 . See Jonathan Z. Sm ith, “T he Influence o f Symbols upon Social Change: A Place o n W hich to S tand,” in M ap is N o t Territory: Studies in the H istory o f Religions (Leiden: E. J. Brill, 1978), pp. 1 2 9 -1 4 6 . 49. Yokoi, Chusei m inshu to seikatsu bunka, pp. 3 3 7 -3 3 9 . 50. N agahara, N ihon no chiisei shakai, p. 217. 5 1 . Hayashiya T atsusaburo, Chiiseigeino shi no kenkyii: K odai kara no keisho to sozo (Tokyo: Iw anam i S h o ten , 1975), pp. 2 6 9 -2 8 0 . T he ChUyiiki text fragm ent is presented on p. 271. 52. Kita Teikichi, “Sanjo hoshi ko: Kodai shakai soshiki no kenkyu,” M inzoku to rekishi I, no. 3 (1921): 125.

53. Ib id ., pp. 1 2 8 -1 3 0 . 54. “B elief in m arebito” is h o w m o st folklorists in Japan have referred to the events an d sym bols su rro u n d in g th e visiting stranger. A lth o u g h I use th e term below, follow ing o th e r scholars w riting o n Japanese p o p u lar religion, it seems u n fo rtu n a te to lum p this im p o rta n t constellation o f ritual and highly im aginary sym bolic action u n d e r th e rub ric o f “ belief.” T h e use o f this w ord greatly limits o u r appreciation o f h ow these rituals w orked in Japanese society. Ifp e o p le did n o t “believe” in these visitors, does th a t m ean th e ir ritual appearance could n o t have som e so rt o f sym bolic affect o n th e com m unity? It seem s to m e th a t these n o m e n ­ clature b o rn in th e P ro te sta n t R efo rm atio n an d ab sorbed by th e Japanese folklore m o vem ent confuses th e d ata an d veils th e significance o f these custom s as ritual and sym bolic expressions o f a m ore sub tle ideology o f the cosm os. I t reduces o u r u n d erstan d in g o f w hy ritual actions are experienced as pow erful and transform a­ tive by people. 55. O rig u ch i’s discussion o f m areb ito is scattered th ro u g h o u t his large col­ lected w orks. H is w o rk o n this subject has recently been called into q u estio n by scholars in th e field o f Japanese religions, for th ey claim th a t he based his definitive discussions o f m areb ito o n scant evidence an d flights o f fancy w ith th e data. H is tend en cy to base assum ptions a b o u t th e “ essence o f th e Japanese m in d ” on a n ­ cient sources, an approach to Japanese folklore studies he ch am pioned, u n d e n i­ ably places him in th e nativist tra d itio n o f th e kokugaku scholars, and consequently the nationalist agen d a o f his w ork can n o t go u nexam ined. N evertheless, m any scholars have also recog n ized th a t w hile O rig u c h i’s evidence was at tim es scant and his m e th o d highly po etic at best, his insights w ere rem arkable, and his discus­ sion o f m arebito reflects a shared m eaning an d stru ctu re in a large n u m b e r o f disparate sources in Japanese religious history. W ith this in m in d , we will briefly review th e basic p a tte rn o f m areb ito belief, as set fo rth by O riguchi, H o ri, and oth ers from th e Japanese folklore studies m ovem ent. 56. O riguchi S h in o b u , O riguchi Shinobu zenshu, ed. O riguchi H akase Kinen Kodai K enkyujo (Tokyo: n .p .), 2 :3 3 -3 7 . 57. Yoshiko Y am am oto, The N am ahage: A fe s tiv a l in the N ortheast o f Ja p a n (Philadelphia: In stitu te for th e Study o f H u m a n Issues, 1978). 58. T h e m ap o n p. 17 in Y am am oto, N am ahage1 show s th e d istrib u tio n o f these “N ew Year’s visitors” practices th ro u g h o u t Japan. 59. T his term also includes a n u m b e r o f perform ances th a t d o n o t use puppets. A n exhaustive study o f this in terestin g class o f perform ances is Park Jo n Yul, K a dozuke no kdzd (Tokyo: K o b u n d o , 1989). 60. See C hie N akane, Japanese Society (H a rm o n d sw o rth : P enguin Books, 1973). See also E m iko O hnuki-T ierney, Illness a n d C ulture. 6 1 . M ichele M arra, Representations o f Power, p. 68. 62. F or this reading o f the term , I am follow ing th e suggestion o f M orita Yoshinori. See his Chusei senm in zd geind no kenkyii (Tokyo: Yusankaku, 1981), p. 117. 6 3 . T h e nam e is w ritte n w ith th o se four characters b u t is also seen w ritten w ith th e characters “ a th o u sa n d g o o d th in g s (kotobuki) te n th o u sa n d years.” 64. We will see vestiges o f this p erform ance form in the discussion o f th e Shiki Sanbaso rite in ch apter 5, w hich b orrow s m u ch o f th e language. O ne character even asks th e o th e r to “ dance to Senzu M an zai.”

65. In som e parts o f Japan, these perform ers w ere simply called m anzai. 66. Q u o te d in M orita, Chiisei senm in zogeino no kenkyii, p. 125. This passage is also m entioned by Hayashiya T atsu b u ro in his ChHsei getnd shi no kenkyii, pp. 3 2 0 -3 2 1 . 67. T his practice is widely described. C om parative studies from o th e r parts o f the w orld indicate th a t strangers often carry a m arker to indicate th at they com e in peace o r to identify them as com ing from outside the village they are entering. By m arking them selves as such, they m inim ize th e fear th a t they are trying to cause tro u b le, since they are openly adm ittin g th a t they are outsiders. 68. M isum i H a ru o , N ihon m in zo ku g ein o g a iro n (Tokyo: T okyodo Shuppan, 1974), p. 151. 69. W ritten w ith the characters “hail (th e precipitation) ru n n in g .” 70. W. G. A ston, trans. Nihongi: Chronicles o f Japan fro m the Earliest Times to A .D . 6 9 7 (R utland, Vt.: Charles E. T u ttle, 1972), p. 411. See also references on pp. 4 1 4 , 4 2 1 . A ston points o u t th a t the o th e r nam e for ararebashiri perform ances was m anzairaku. From this series o f references, it is possible to discern th at toka perform ances o f som e so rt were popular at c o u rt from this early period and were un d ersto o d to be ritual perform ances for the longevity o f the sovereign. 71. M orita, Chusei senm in to zo geinB no kenkyii, p. 123. 72. It is n o t possible to determ ine w hether o r n o t the Senzu M anzai is di­ rectly derived from the practice o f reciting toka, b u t the com m onalities betw een these tw o perform ance traditions are striking en ough to give the th eo ry som e w eight. 73. M orita, ChUsei senm in to zd geind no kenkyii, p. 125. M orita also points o u t th a t the practice o f N am ahage in th e Akita region o f Japan is also an example o f this belief. 74. M isum i, N ihon m inzoku g e in o g a iro n , pp. 1 7 0 -1 8 0 . 7 5 . Erika G erlinde de P o o rte r, tran s., M otoyosh i1S Sarugaku D angi: A Descrip­ tion a n d Assessment with A n n o ta te d Translation (Leiden, 1983), p. 156. In her note o n this section, the translator defines m atsu-bayashi as follows: “Special so rt o f music w ith dancing w hich was perform ed at N ew Year and was very popular in the first h alf o f the 15 th century. It was presented by singing m onks o r ordinary people. They visited the residences o f noblem en and perform ed this m usic, for which they were th en rew arded” (p. 2 7 7 ). D e P o o rte r’s ann o tated translation o f Sarugaku D a n g i contains excellent source m aterial in the notes. 76. My discussion o f these aspects o f kadozuke is inform ed in p art by O riguchi S hin o b u ’s analysis, cited in Jacob Raz, A udience an d Actors: A Study o f Their Interaction in the Japanese Theatre (Leiden: E. J. Brill, 1984), pp. 3 6 -4 1 . 77. See N inom iya, “ Inquiry C oncerning E ta,” p. 118. 78. G erd Baum ann, “ Ritual Im plicates O th e r s ’: R ereading D urkheim in a Plural Society,” in U nderstanding R itu a ls, ed. D aniel de C oppet (L ondon: R outledge, 1992), p. 98. Baum ann argues o u t the problem s w ith this series o f assum p­ tions by looking at rituals in w hat can clearly be considered pluralistic societies. The m ulti-ethnic L o n d o n sub u rb o f Southall is his chief example. 7 9 . E dm und Leach, C ulture a n d C om m unication: The Logic By Which Symbols A re Connected (C am bridge: C am bridge U niversity Press, 1976), p. 45, cited in B aum ann, “Ritual Im plicates O th e r s ,’ ” p. 97.

80. T h e jo ru ri tra d itio n , th o u g h u n d e r th e th reat o f to tal dem ise, was n o t yet com pletely d efunct. 81. T his was said to me by an old m an living in th e village o f F ukura o n Awaji in 1 9 8 8 , as he rem em b ered th e ch ild h o o d visits o f th e pup p eteers from the early p art o f this century. 82. H ajim e N akam ura, Ways o f T h in k in g o f Eastern Peoples (H o n o lu lu : U n i­ versity o f H aw aii Press, 1 9 7 1 ), p. 587. 83. R o b e rt J. S m ith, forw ard to Y am am oto, N am ahage, pp. 1—2. C hapter 3: A C rippled D eity, a P riest, and a P u p p et 1. J. Z. S m ith, “ Influence o f Sym bols,” p . 131. 2. T his case is significant because it has b een arg u ed th a t later k u g utsu groups developed o u t o f the p u p p e try g ro u p s affiliated w ith th e U sa shrine. 3. AU translations o f an d em phases in this tex t are m ine. I translate th e entire tex t an d discuss this rite in detail in “V iolence, R itual R een actm en t, and Ideology: T h e H o j6 -e (R ite for Release o f S en tien t B eings) o f the U sa H achim an Shrine in Jap an ,” H istory o f R eligions 3, no. 4 (M ay 199 4 ): 325—357. 4. This tw ist o n th e sto ry clearly reveals the role o f H achim an belief in the spread o f belief in goryo (m alevolent spirits o f th e deceased). A discussion o f th e G ion festival as a goryo-e can be fo u n d in N eil M cM ullin, “ O n Placating the G ods and Pacifying th e P opulace: T h e Case o f th e G ion Goryo C u lt,” H istory o f Religions 2 7 , no. 3 (F eb ru ary 1988): 2 4 6 -2 6 9 . See also Ich iro H o ri, Folk R e li­ g io n in Japan: C o n tin u ity a n d C hange (C hicago: U niversity o f C hicago Press, 1 9 6 8 ), 115. 5. C o rre sp o n d in g to 7 2 4 C. E. T h e n o te is in th e tex t and was probably added to m ake th e various accounts o f th e rite agree. 6. T his is also a co m m o n th em e in early C hinese m artial treatises and tales. T he recen t film To Live by th e C hinese film m aker Z h an g Yimou casts this issue o f pu ppeteers in b attle in a m o d ern lig h t w hen tw o shadow puppeteers are taken priso n er an d m anage to survive by p erfo rm in g for th e R ed A rm y d u rin g the fight against th e K u om intang. 7. See G ina L. Barnes, Protohistoric Tam ato: Archaeology o f the First Japanese State (A nn A rbor: U niversity o f M ichigan Press, 19 8 8 ), pp. 2 6 9 —2 77. 8. Yamagami Iz u m o , M iko no rekishi, ch ap ter I. 9. See N ak an o H atayoshi, H a c h im a n shinko-shi no kenkyii (Tokyo: Yoshikawa K obu n k an , 1 9 7 6 ), 1 :9 2 -9 3 . 10. W hile o n o n e level this p roblem seem s arcane, th e issue o f th e “ uniqueness o f Jap an ” is a central ideological p o in t in m uch Japanese scholarship relating to ritual perform ance. 11. C ited in M isum i H a ru o , Sasuraihito n o g e in d shi (Tokyo: N H K Books, 19 74 ), p. 35. 12. See D u n n , E arly Japanese P uppet D ra m a , pp. 6 4 -6 5 . 13. T su n o d a, N ingyd g eki no seiritsu n i ka n suru kenkyii, pp. 1 9 8 -2 0 4 . We assum e th a t T su n o d a ’s “gypsies” are E astern E u ropean G reek puppeteers. 14. D o n ald K eene, B u n ra k u : The A r t o f the Japanese Puppet Theatre (Tokyo: K odansha In te rn a tio n a l, 1 9 7 3 ), p. 19,

15. Ib id ., p. 20. 16. T sunoda, N ingyd g e k i no seiritsu n i kan suru kenkyu, p. 352. 17. T h e standard version o f this text used by scholars is found in Gunsho ruijii (C ollection o f texts), vol. 9 (1 9 2 8 ), pp. 3 2 4 -3 2 5 . I have used this text in my translation, th o u g h m any reliable versions are available scattered th ro u g h o u t Japa­ nese secondary sources o n perform ance and puppetry. For a discussion and partial translation o f the text in English, see D onald Keene, B unraku, p. 20. In G erm an, a translation appears in H agen Blau’s Sarugaku u n d Shushi: Beitrage z u r A usbildung dram atischer Elemente im weltlichen u n d religiosen Volkstheatre der H eia n -Z eit under besonderer Berucksichtigung Seiner Sozialen G rundlagen (W iesbaden: O tto H arrassow itz, 1966), p. 235. Kawajiri T aijipresents aversion o f the text in m odern Japanese in his Nihon Ningyo Geki H a tta tsu Shi, Ko (Tokyo: Bansei Shobo, 1986), pp. 100—101. F o rfu rth e rd isc u ssio n in Japanese, see Yamaji K ozo “ K ugutsu,” in Chusei no minshU to Geino, ed. Yamaji Κ δζδ, M ori Fum iko, and Kyoto Buraku Shi KenkyOjo (Kyoto: A unsha, 1986), 5 4 -5 9 . Hayashiya T atsusaburo discusses th e text in his Chiiseigeind shi no kenkyii, 8 th ed. (Tokyo: Iwanam i S hoten, 1987), pp. 3 2 3 -3 2 6 . T hese are b u t a few examples o f discus­ sions o f this text. To facilitate reading th e text in translation, I have added stanzas, divisions following th e topic divisions used by m ost Japanese scholars w hen dis­ cussing th e text. 18. H ere, O e n o Masafusa uses the pro p er nam e for a group o f Chinese per­ form ers very popular in H an C hina, the yulong m anyan zhixi (literally “tran s­ m utations o f fish and drag o n s” ). This direct reference in the text indicates O e M asafusa’s familiarity w ith Chinese perform ance. These perform ers are discussed by Wu H u n g in “A Sanpan C h ario t O rn am en t and the Xiangrui Design in W est­ ern A rt,” Archives o f A sia n A r t 37 ( 1984): 3 8 -5 9 . See discussion below. 19. T here is a character missing from th e text. T he w ord “ dazzle” is suggested by Tsunoda. 20. This w ould indicate th a t they do n o t even raise silkworms, since m ulberry leaves are food for these insects. 2 1 . T h e tex t uses th e characters “ hyaku shin” (h u n d red g o d s). F o r a com plete discussion o f the issues raised by this reference in this text, see my article “ O f Plagues and Puppets: O n th e Significance o f the N am e H yakudayu in Japanese R eligions,” in Transactions o f the A siatic Society o f Japan, 4 th ser., 8 (1993): 1 0 7 -1 3 1 . 22. T h e term in the text is fu ku su ke, a ritual object representing a person, decidedly phallic in shape. 23. T h e w om en were singers and dancers. A num ber o f the names m entioned here appear elsewhere in texts from the period. It is difficult to know just how th e au th o r in ten d ed these nam es to be read, and m ost scholars tolerate a variety o f guesses. 24. T he next tw o lines o f the text list types o f songs sung by the w om en. Imayd and Furukawayo refer to tw o styles o f sh o rt poem s o f an en tertain in g nature p u t to music. 25. ta-u ta 26. ka m i-u ta

27. toka T h e reader will recall the discussion o f this perform ance form in chap ter 2. 28. tsuji-uta 29. fu z o k u 30. hdshi 31. Y oshinobu ln o u ra and T oshio Kawatakc, The T ra d itio n a l Theater o f Japan (N ew York: W eatherhill5 1 9 8 1 ), p. 173. 32. Hayashiya, ChUseigetnd shi no kenkyH, p. 323; em phasis m ine. 33. W u H u n g “ Sanpan C h ario t O rn a m e n t,” p. 53. 34. T su n o d a, N ingyogeki no s e irits u n ik a n s u m kenkyu, pp. 3 3 2 -3 3 5 . Also see K eene, B u n ra k u , p. 20. 35. Yamaji “ K u g u tsu ,” p. 55. 36. K eene, B u n ra k u , p. 20. 37. Yamaji, “K u g u tsu ,” p. 55. 38. Ib id ., pp. 54-59. 39. See Felicia Bock, Classical L ea rn in g a n d Taoist Practices in Early Japan w ith a Translation o f Books X V I a n d X X o f the Engi-shiki (Tem pe: C en te r for Asian Studies, A rizona State University, 1 9 8 5 ), p. 45. 40. Ib id ., p. 83 n. 35. T h e w ood and bark o f the peach tree are also m edicinal. 4 1 . T he e th n o b o ta n ist C hristine F ra n q u e m o n t suggested to me th a t “peachw o o d ” could have been applied to a tree in Japan th a t did n o t necessarily co rre ­ sp o n d to th e tree d esignated by th e same characters in C hina. T his, she n o ted , is a com m on p attern in th e history o f early Japanese botany. 4 2 . C harles H . L o ng , Significations: Signs, Symbols a n d Im ages in the Interpre­ tation o f R eligion (Philadelphia: Fortress Press, 1986), p. I. 43 . C ited in T su n o d a, N ingyog e k i no seiritsu n i k a n suru kenkyu, p. 46 8 . 44. Sadafusa, K a m m o n gyoki, vol. I (1 9 4 4 ), p. 12. 45 . T he m ost exhaustive study o f draw ings o f puppeteers in source books is S hinoda J u n ’ichi and N ingyo Butai-shi Kenkyukai, eds. N ingyd jo ru ri butai-shi (Tokyo: Yagi S h o ten , 1992). T h e w ork includes a detailed discussion o f every pictographic reference to p uppeteers in Japan th a t exists. T h e concern o f the book is to trace th e developm ent o f th e jo ru ri stage, b u t it includes interesting discus­ sions o f th e various perform ances. Pictures w ere also used to great effect in the reco n stru ctio n o f the early history o f old jo ru ri by D u n n , Early Japanese Puppet D ram a. H is book includes m any draw ings o f th eaters and stages b u t only one p ictu re o f a kugutsu-m aw ashi (th e same one th at I include). 46 . As rep rin ted in S hinoda and N ingyo Butai-shi K enkyukai, N ingyo jo ru ri butai-shi, p. 191. 47. Ib id ., pp. 1 9 1 -1 9 2 . 48 . T h e in terestin g th in g a b o u t this perform ance co n tex t is th a t it is identical to o n e th a t I have seen o n Awaji in recent years (the w ork o f an elderly gentlem an w ho is reviving E bisu perform ance), dow n to th e detail o f children gathered in the street. I discuss his perform ance in chap ter 6. 49 . T h e box is considered to be th e first “ stage” o f th e p u p p e t th eate r in Japan. 50. Ib id ., p. 192. 51. Ib id ., p. 194.

52. Ib id ., p. 196. 53. Ibid. 54. Ibid. 55. C ited in M orita, Chiisei senm in to zd geino no kenkyu, p. 1 9 1 . 56. T he diary itself spans th e period from Bunm ei (1 4 6 9 -8 7 ) th ro u g h the e n d o f E do. 57. See Kotaka Kyo, com p. C hiiseigeindshi nenpyo(Tokyo: M eicho Shuppen, 1987), pp. 3 5 0 -4 4 0 (sporadic references in assorted listings). 58. Ib id ., p. 438. 59. F or an exhausting, if n o t exhaustive, study o f the textual references to changes in this shrine, see Richard P onsonby-F ane, The Vicissitudes o f Shinto (Kyoto: T h e Ponsonby M em orial Society, 1963), pp. 3 4 -8 0 . T he reference to M ontoku jitsu ro ku is on p. 68. P onsonby-F ane’s detailed and garbled discussion o f this shrine confuses the m atter in th a t he insists on referring to it as H iroda, although by th e Kamakura period this was no longer the nam e used to refer to this shrine. 60. This hypothesis was argued m ost vehem ently by Ponsonby-Fane to sup­ p o rt his ow n theory o f spirit aspects. H e may n o t have been w rong, b u t his evi­ dence is far from convincing. 61. T he H iro d a shrine appears to have been divided from th e earliest times in to tw o parts, th e main shrine (called H iro d a) and the shrine on the southern beach (H am a-m inam i). Perhaps this is the m eaning o f the reference to directions in O e no M asafusa’s snippet. In 1132, th ere appears to have been a struggle to separate the tw o shrines in to au to n o m o u s units. Ponsonby-Fane has suggested th at the H iro d a shrine was originally th e place for the w orship o f the rough spirit o f A m aterasu, and th at “an extraordinary m etam orphosis . . . befell H iro ta ,” and this single deity site was transform ed into a site w here five deities were w or­ shipped, and the identity o f A m aterasu O m ikam i was entirely lost. H e notes th a t this change to o k place in Jisho I (1 1 7 7 ). By the m id-K am akura period, it was again united; and w hen it was subsequently divided, it becam e tw o shrine units, the H iro d a shrine (referred to as th e n o rth e rn hokorct, or w orship site) and the Ebisu shrine. B oth were referred to as Nishinom iya after this tim e, leading to a great deal o f confusion th a t no one (including shrine priests at N ishinom iya) has been able to so rt o u t. 62. F or a list o f theories ranging from th e plausible to the ridiculous explaining the m eaning o f the nam e “w estern shrine,” see Yoshii, Ebisu shinkd to sono fu d d , pp. 3 8 0 -3 8 7 . C oncerning the Ise theory suggested by Ponsonby-Fane, Yoshii notes th at no one in Ise believes it. 63. Yoshii Tar5, “N ishinom iya no k u g u tsu ,” in M inzoku to rekishi I , no. I (Taisho 8): 29. 64. I discuss th e various associations o f this nam e in my article “O f Plagues and P u p p ets.” 65. M orita, Chiisei senm in to zo geino, pp. 1 9 6 -1 9 7 . See also Yoshii, “N ishi­ nom iya n o k u g u tsu ,” p. 30. 66. A hand-draw n plan o f th e shrine from 1978 can be found in Yoshii Sadatoshi, Ebisu shinkd to sono fu d o (Tokyo: K okusho Rikkokai, 1989), pp. 3 2 8 -3 2 9 .

67. Ja n e t G oodw in has w ritten a b o u t th e role o f itin e ra n t sacred specialists in th e v o lu n tary d o n atio n s cam paigns o f m edieval Japanese B uddhist centers. She suggests th a t this strategy used by B ud d h ist centers probably “ inspired a rash o f im ita to rs.” F o r th o ro u g h discussions o f these m e th o d s o f securing funds, see her “Alms for Kasagi T em p le,” in The J o u rn a l o f A sia n Studies, 4 6 , no. 4 (N ovem ber 1987): 8 2 7 -8 4 1 . 68 . Yoshii, Ebisu shinkd to sono fu d d , p. 10. 69. Ib id ., pp. 1 5 -3 8 , p resents tw en ty -fo u r E bisu shrine placards depicting E bisu. T hose o n pp. 2 1 , 2 2 , 2 6 , 2 7 , 2 8 , 32, and 33 show the standard im age o f E bisu as a cheerful o ld m an. 70. N am ihira E m iko, “ Suishintai o E bisu to shite m atsuru shinko: S ono imi to kaishaku,” M in zo ku g a ku K enkyu 4 2 , no. 4 (M arch 1978): 3 3 4 -3 5 5 . 71 . K ato G enchi, “A S tudy o f th e D ev elo p m ent o f Religious Ideas am o n g the Japanese People as Illu strated by Japanese Phallicism ,” Transactions o f the A sia tic Society o f Ja p a n , 2 d ser., supp. to vol. I (D ecem b er 1924): 13, 53—54. 72. O g u ra M an ab u, “ N o to n o kuni hyochakugam i no k o ,” K oku g a ku in zasshi 55, no. 3 (1 9 5 4 ): 3 0 -4 3 . See also Sakurada K atsu nori’s in -d ep th study, “ N o to no hyochakugam i n o kenkyu,” N ihon m in zo ku g a ku I , no. 4 (1 954): 91—94. 73 . Sakurada K atsunori, “T h e E bisu-gam i in Fishing V illages,” in Studies in Japanese Folklore, ed. E dw ard D o rso n (B lo o m in g to n , Indiana U niversity Press, 1 9 6 3 ) p . 124. 74. Sakurada, “T h e E bisu-gam i in Fishing V illages,” p. 125. 75. See Yoshii, Ebisu Shinko to sono fiid o , pp. 1 8 9 -1 9 1 , for a list o f examples. 76. Ib id ., p. 3. 77. N am ihira, “Suishintai o E bisu to shite m atsu ru sh in k o .” 78. R o b e rt J. S m ith, Ancestor Worship in C ontem porary Japan (Stanford: S tan­ ford U niversity Press, 1 9 7 4 ), p. 45. 79. U n til th e tw en tieth century, th e m ajor event o f the Ebisu shrine was a festival in th e n in th m o n th on th e tw enty-sixth day, w hen th e o-mikoshi o f the N ishinom iya shrine was carried from th e shrine precincts along the shore line to th e area called W adaham a. 80. Yoshii, Ebisu shinko to sono fiid o , p. 2. 81. In early m edieval Japan, th e idea th a t conversations in dream s could be taken literally as in stru ctio n s to th e dream er was n o t unco m m o n . T his idea o f “ conversations w hile d re a m in g ” ( m uchu mondo) has been in tro d u c ed by W illiam L aFleur in his The K a rm a o f Words: B uddhism a n d the L iterary A r ts in M edieval Ja p a n (Berkeley: U niversity o f California Press, 1 9 8 3 ), pp. 4 -8 . See also G eorge J. T an ab e, Myoe the Dreamkeeper: Fantasy a n d Knowledge in Early K a m a k u ra Buddhism (C am bridge: C ouncil o n E ast Asian Studies, H arvard University, 19 92 ). W hile b o th o f these studies argue for a B uddhist in terp retatio n o f d ream ­ ing, th e idea o f d ream ing as a m o d e o f direct revelation was co m m on in m edieval Japan. 82. See for exam ple, th e textual references in R ichard Ponsonby-F ane, Visiting Fam ous Shrines in Jap a n (Tokyo: P o n sonby-F ane M em orial Society, 19 6 4 ), 6 :3 4 1 -3 4 2 . 83. N ihon Rekishi D a ijite n , vol. 2 , s.v. “ E b isu.” 84. N am ihira, “Suishin o E bisu to shite m atsuru sh in k o ” p. 340.

85. Kitagawa, Religion in Japanese History, p. 14. For an extensive study o f this four-aspect p h en o m en o n , see the tangled discussion by Ponsonby-Fane in his Vicissitudes o f Shinto, pp. 4 1 -5 6 . It has been argued, based on references to the divine aspects in early texts, th a t the latter tw o are actually just aspects o f nigi-m i-tam a. 86. For example the “ro u g h ” spirit o f the sun goddess Amaterasu is enshrined in the Ara M atsuri no Miya (literally, shrine for w orshipping the rough aspect) at Ise. References to this location are m ade in prayers num ber 18, 19, and 22 o f the Norito. See Philippi, Norito, p. 84, s. v. “A ra-m atsuri-no-m iya.” 87. Takeuchi T., in Nihon shakai m inzoku jite n (Social and folklore dictionary o f Japan) (Tokyo: H ihon M inzoku Kenkyukai, 1960), s.v. “ E bisu.” 88. C ited by Philippi in K ojiki, p. 399 n. 4. 89. Ibid. This theory o f H iruko as th e “sun child” has been m ost strongly m aintained by the French scholar Jean H erb ert. See his D ieux a n d sectes populaires du Japon (Paris: E ditions Albin M ichel, 1967), p. 112, and Les dieux nationaux du Japon (Paris: Editions AIbin M ichel, I 9 6 9 ), pp. 54. 90 . See the creation m yth o f the M ande o f Africa, recounted in Charles H . Long, A lpha: Myths o f Creation (N ew York: G eorge Braziller, 1963), p. 134. T hat the failed creation is abandoned seems to be a com m on them e in cosm o­ gonic myths. 91. C ited in Ponsonby-Fane, Vicissitudes o f Shinto, p. 70. 92. C ited in N am ihira, ICegare no kdzd, pp. 3 4 1 -3 4 2 . 93. Ibid., p. 156. N am ihira’s discussion o f the m alevolent aspects o f Ebisu is unconvincing (pp. 1 5 6 -1 6 0 ). 94. Yoshii, Ebisu Shinko to Sono Fudo, p. 137. 95. K ojiki 1:29, 3. 96. Ibid. 1:32, I. References to this realm , which is the land o f Japan, usually use simply “T he C entral Reed Plain,” b u t in prayers various aspects o f this long name are often used. 97. Ibid. 1:32; 1:33. N o te th at Philippi discusses the implications o f these failed missions in his notes, p. 120f. 98. A nother name for Koto shiro nushi no kami. 99. K ojiki 1 :3 5 ,9 -1 3 . 100. Ibid. 1:36, 10. 101. Ibid. 1:37, 3. 102. Ibid., p. 134f. 103. Izum o is the location w here deities w ho were kicked o u t o f the Imperial Line, such as Susano, are w orshipped. It is understo od th at all the deities in Japan retire here for one m onth o f the year. 104. K ojiki, p. 135f. 105. A ston, N ihongi, p. 225. 106. This point is m ade by Philippi in Kojiki, p. 13 If. 107. K ojiki 1:2,7, 4. 108. Ibid. p. 412 n. 17. 109. Clearly the problem o f K otoshiro nushi’s reprisal was o f concern in myths at th at tim e. A ccording to N ihongi, aversion o f the K otoshiro nushi m yth has this

deity tran sfo rm ed in to an eig h t-fa th o m -lo n g bear sea m o n ster, a m enacing charac­ ter. See A ston, Nihoneji1 p. 61. 110. T hese prayers have been translated, by D o n ald Philippi in Norito: A Translation o f A n c ie n t Japanese R itu a l Prayers (P rinceton: P rinceton U niversity Press, 1990). AU w o rd in g o f titles and translations are from Philippi’s edition. 1 1 1 . O kuni nushi and his kin. 112. Philippi, N orito, pp. 6 9 -7 0 . 113. T h e assim ilation in this case is a m eetin g an d b lending o f B u ddhist and S hinto system s (shim butsu shugd), b u t it is im p o rta n t to keep in m ind th a t o th e r religious sources c o n trib u te d to th e h isto ry o f Japanese religions. 114. T h e case o f H ach im an is a n o th e r exam ple o f a deity w hose nam e does n o t appear in the chronicles being m oved in to m ainstream religious w orship, al­ th o u g h in this case it is largely th ro u g h identification w ith B uddhist figures. 115. It is difficult to identify all th e diseases th a t ravaged Japan th ro u g h o u t history, in large p a rt because local nam es for th e same diseases varied, b u t mainly because historical descriptions o f diseases frequently did n o t provide en o u g h in ­ fo rm ation to identify th em . Sm allpox, measles, dysentery, cholera, and venereal diseases are, how ever, readily identifiable from records o f epidem ics in Japan. A l­ th o u g h som e scholars argue th a t b u b o n ic plague existed b u t was n o t adequately described in th e records to be identified, A nn B ow m an Janetta wisely points o u t th a t “such arg u m en ts are hardly convincing. T h e sym ptom s o f bu b o n ic epidem ic are am o n g th e m ost graphic o f diseases,” so it is unlikely th a t anything as sy m p to ­ m atic as b u b o n ic plague w o u ld have go n e u n d o c u m e n te d in Japanese history ( E p­ idemics a n d M o rta lity in Early M odern Ja p a n [P rinceton: P rin ceto n U niversity Press], p. 191). F u rth e r su p p o rtin g th e view th a t Japan did n o t suffer from b u ­ bonic plague epidem ics, she w rites, “ In A ustria, a cordon sanitaire th a t was created in 1728 to keep th e b u bo n ic epidem ic from e n te rin g from T urkey was successful, b u t th e q u aran tin e req uired an en o rm o u s effo rt o n th e p a rt o f th e A ustrian gov­ ern m en t. U nlike A ustria, Japan h ad th e advantage o f being su rro u n d e d on all sides by a natural cordon sa n ita ire th a t was apparently able to keep th e bubonic plague o u t o f the c o u n try until th e en d o f th e n in e te e n th c e n tu ry ” (p. 2 0 0 ). 116. See H a rtm u t O . R o te rm u n d , Hdsogami: ou la petite verole aisement: m a teria u x p o u r PStude des epidemies dans Ie Japon des X l N ile , X I X e siecles (Paris: M aisonneuve e t L arose, 1 9 9 1 ) for a discussion o f m agical practices related to sm allpox d em ons in m edieval an d early m o d ern Japan. H is study m entions Ebisu only in passing o n p. 148, and he argues th a t m any o f th e N ew Year’s rituals o f felicity an d longevity sho u ld be u n d e rsto o d w ithin the larger c o n tex t o f epidem ic co n tro l at a ritual level. 117. N akajim a Y oichiro, Byoki nihon s h iTokyo: Y uzankaku, 1 9 8 8 ), p. 76, and Fujikaw a Yu, N ihon shippei shi (Tokyo: H eib o n sh a, 19 6 9 ), cited in Jan n etta, E p i­ demics a n d M o rta lity, p. 68 . Jan etta p o in ts o u t th a t several o f the intervals listed by Fujikaw a w ere q u ite g reat, in d icatin g th a t perhaps sm aller epidem ic in betw een these o ccu rred b u t w ere n o t recorded. So, th e n u m b e r o f sm allpox epidem ics in Japanese h isto ry is probably som ew here betw een fifty five and one h u n d red . 118. W illiam M cN eill, Plagues a n d Peoples (G arden City, N.Y.: D oubleday, 1 9 7 6 ), pp. 1 3 9 -1 4 1 .

119. Technically, a ddsojin is a spirit which protects travelers. Such spirits also have decidedly sexual overtones and may have been phallic deities. D epictions o f ddsojin are often phalli. 120. I have relied o n T su n o d a’s transcription o f this text in Ningyb Jjeki no seiritsu n i kan s u m kenkyii, p. 399. T akano Tatsuyuki has suggested th at th e miss­ ing character in the q u oted passage is “w o o d ,” which w ould give a reading o f “ the nu m b er o f these they carve o u t o f w ood is in th e h u ndred and th o u san d s.” Konishi Jin ’ichi suggests, however, th a t th e missing character was probably a com m a, m aking th e text read “ People carve these by th e hundreds and thousands” (ibid.). 121. T h is r e f e r s to th e L e e c h C h ild . 122. Q u o te d in Yoshii, “N ishinom iya n o kugutsu, p. 28 123. Ibid. 124. “ Kotaku o shinseki shukurin hoso sh u g oshingo.” T he calligrapher re­ ferred to in rhis piece was H osoi K otaku, a fam ous neo-C onfucian scholar, famed calligrapher, and m an o f letters w ho lived from 1658 to 1735. For this translation, I have used the transcription in T sunoda, Ningyo geki no seiritsu ni kan s u m kenkyii, pp. 4 1 7 -4 1 8 . 125. O ne w ould hope th a t a sentence like this w ould serve as a rem inder to contem porary historians and ethnographers never to assume th at the obvious may be om itted from descriptions or discussions. Frequently the obvious th at ties to g eth er m any uncertain threads is the m ost vulnerable to om ission in the historical record. 126. Yoshii, “ Nishinom iya no k u g u tsu ,” pt. I , p. 29. 127. Smallpox was largely a disease afflicting children. Janetta has docum ented and analyzed the records o f infant and childhood deaths based on tem ple records ( kakocbd) d uring the Tokugaw a period. See Jannetta, Epidemics and M ortality, pp. 6 1 -1 0 7 . 128. Bock, Engi-Shiki, p. 20. 129. Yoshii, “Nishinom iya n o k u g u tsu ,” p. 30. 130. T he play A m id a 3S R iven Breast has been translated in to English by C. J. D u n n in Early Japanese Puppet D ra m a , pp. 1 1 1 -1 3 4 . 131. C ited in Shinoda and N ingyo Butai-shi Kenkyukai, Ningyo jo ru ri butaishi, p. 195. 132. Yoshii, “N ishinom iya no k u g u tsu ,” pt. I , p. 29. This record o f the event at the H yakudayu shrine is dated S h otoku 4, the sixteenth day o f the eighth m onth. Chapter 4: A D ead P riest, an A n gry D eity, a Fisherm an, and a P uppet 1. W akam ori T aro, ed., A w ajishim a no m inzoku (Tokyo: Yoshikawa K obunkan, 1974), p. 3. 2. M iyam oto Tsuneichi, Setonaikai no kenkyii (Tokyo: Kadokawa Shoten, 1965), p. 104. 3. U m ehara Takeshi, A m a to tenno, 2 vols. (Tokyo: Asahi Shinbunsha, 1991). 4. N akano, H achim an shinko-shi no kenkyii. See in particular his discussion o f the confederation o f various tribes into the U sa clan.

5. See N iim i Kanji, A w a ji no ningyo shibai (Tokyo: Kadokawa S h o ten , 19 7 4 ), p. 17, and W akam ori, A w a ji no m in zo k u , 1 1 -1 3 . A lth o u g h this hypothesis was cred ited to W akam ori, th e bulk o f th e research was co n d u cte d by H an d a Yasuo o f O ita University. P rofessor H a n d a d ied before th e research was published, and so it was p u b lish ed by W akam ori. W akam ori stated in his research rep o rts th a t he w ished credit for this hypothesis to g o in p a rt to H anda. 6. T om bs o f am azoku w ere discovered o n the tiny island o f O kinoshim a (ab o u t five h u n d re d m eters o f f th e so u th w estern coast o f Awaji and actually a t­ tach ed to Awaji d u rin g low tid e), indicating th a t th e island was settled by these m aritim e people. T h ere is a curious sto ry c o n n ected w ith this island, w hich ap ­ pears to have som e tru th to it. It is believed th a t th e am azoku bu ried on this island did n o t w an t th e ir tom bs to be b o th e re d an d th a t from a very an cien t tim e this island was co nsidered ta b o o . O ral tra d itio n m aintained th a t any fisherm an w ho w en t to o close to th e island w ould crash o n th e rocks o r dro w n , o r som e o th e r terrible calam ity w ou ld befall him o r his family. In the 1950s, ig n o rin g this local legen d , a g ro u p o f archaeologists from Tokyo Kyoiku D aigaku excavated the is­ land, and w ithin several m o n th s a n u m b e r o f th e m m et tragic deaths. A year and a h alf later, according to m any inform ants all o f th em had died. A t th e tim e I w rote my docto ral dissertatio n , n o fu rth e r research has been do n e on th e tiny island o f O kinoshim a. In th e su m m er o f 199 0 , th e new o w ner o f th e island decided th a t it had great p o ten tial as a to u rist attractio n , and so he arran g ed to make it accessible. H e hired som e w orkers to clear brush and p lan t som e wildflowers. Sunset cruises to see th e tiny island w here the am azoku w ere b u ried w ere offered to the public. T h at sum m er, I crossed over to th e island to lo o k at th e tom bs w ith a few o th e r scholars, b u t we m ade sure th a t we w ere safe and to o k along elaborate offerings to the spirits o f th e am azoku in th e form o f salt, dried fish, rice and sake. W hen we arrived on th e island, we discovered th a t th e w orkers w ho had been there to clear brush had taken care o f this before us, an d a sm all appeasem ent sto n e had been erected for th e am azoku. N evertheless, we m ade o u r offerings in earnest before looking at the tom bs. Early in th e fall o f 1988 and again in th e su m m er o f 1 9 9 0 , I tried to follow up o n th e U sa hypothesis by tracin g references to the deity H yakudayh th ro u g h ­ o u t th e In la n d Sea. M y in te n t was to d eterm in e if the constellation o f am azoku to m b s, Sanjo districts, ap peasem ent p erform ances, w om en o f pleasure, k ugutsu, and epidem ics spirits attach ed to this nam e m ig h t make it possible to suggest th a t th ere was a co m m o n trad itio n o f ritual p u p p etry spreading w ith H achim an belief o f w hich b o th N ishinom iya an d Awaji w ere representative centers. T his research was b o th in terestin g an d enjoyable, b u t in the end I aband o n ed it. A l­ th o u g h still u n p ro v ed , th e hypothesis a b it fu rth e r developed th an w hen H an d a left it. See my article “ O f Plagues and P u p p e ts” also Sw anberg-Law , “ Puppets o f th e R o a d ,” pp. 9 7 —114, 1 2 1 -1 4 8 . I have com e to th e conclusion th a t we m ust satisfy ourselves w ith an in ad eq u ate h isto ry o f ancient origins in the case o f this trad itio n . 7. In h e r Shinto m in z o k u β ή η δ n o genryH Suzuka C hiyono attem p ts to c o n ­ stru ct a co n tin u o u s tra d itio n from th e am azoku th ro u g h H achim an w orship to various p u p p e try trad itio n s (T okyo, 1988).

8. This interpretation was m entioned in th e early-nineteenth-century text A iva ji no k u n i meisho zue. 9. The reader is referred to chap ter 2 and my discussion o f the significance o f Sanjo districts. 10. O th e r local versions also m aintain th a t this figure is a dosojin, a p ro te c to r o f travelers. In o th e r locations H yakudayu was also th o u g h t to be a dosojin. I have suggested elsewhere th a t the female figure may be Tam ayori H im e, or some o th e r sham an. 11. M ihara-gun has several H achim an shrines, b u t the m ajor three are the Koda H achim an shrine, the Kashu H achim an shrine, and the Fukura H achim an shrine. T here are tw o main lineages in th e H achim an tradition: U sa, with head­ quarters in O ita prefecture’s U sa, and Iw ashim izu, w ith headquarters in Kyoto. T he Koda and Fukura shrines are in th e Usa H achim an lineage, and the Kashu shrine is in the Iwashim izu H achim an lineage. Later, the Tsurugaoka, with h ead ­ quarters in Kamakura, becam e an im p o rtan t center o f H achim an belief th ro u g h its connection w ith th e m ilitary rule in Japan during the Kamakura and up th ro u g h the Tokugaw a period. 12. Niim i, A w a ji no ningyd shibai, p. 18. 13. See a discussion o f th e phen o m en o n o f w akam iya in Yanagita Kunio, “H ito w o Kami ni M atsuru F fishu,” T anagita K u n io zenshii (Tokyo: Sogeisha, 1968), 7 :2 5 6 -2 9 4 . 14. T he G enroku-era d o cu m en t from Awaji entitled A w a ji tsHki (Records o f a Passage th ro u g h Awaji) does n o t m ention any o f these shrines, although descrip­ tions o f shrines and tem ples is o ne o f the m ajor features o f the text. I exam ined the original text at the Awaji Rekishi Shiryokan in S um oto in the sum m er o f 1988. 1 5 . T here are m any offerings o f statuary and structures presented around 1805 in b o th K oda and O m ido by th e Sanjo-m ura puppeteers, indicating th a t they perceived the relationship betw een these tw o sites. 16. Niim i, A w a ji no ningyd shibai, pp. 18—19. 17. AU three o f these shrines were im p o rtan t in th at oracles from them were com bined to au gm ent th e au thority o f the imperial family th ro u g h the religious constructions ofY oshida (Yui-itsu) Shinto. 18. T he reader will recall th a t in chapter 2 we reviewed the significance o f Sanjo districts and th e m eaning o f attaching the nam e Sanjo to a perform er’s name. 19. A ston, N ihongi, pp. 1 3 -1 7 . Som etim es Awaji is counted am ong the islands and O nogorojim a is the placenta. (Readers are encouraged to dust o ff their Latin for the coital scene in A ston’s translation.) 2 0 . A b o u t seven kilom eters o ff th e southeastern coast o f Awajishima, there lies a small island m easuring no m ore than three or four kilom eters in length, called N ushim a. This island also claims to have the site which is the original O nogorojima. O n N ushim a, one climbs up th e steep m ountain path and arrives at a small worship hall in which a painted picture depicting Izanam i and Izanagi creating the w orld is hung. T he people o f N ushim a, w ho n u m bered around three h u ndred in 1988, have freq u en t co n tact w ith th e people o f Awajishima. W hen questioned as to how there could be tw o O nogorojim as, an elder on the island o f N ushim a told me, “W ho knows? W ho cares? We have it and they have it. I t’s a pretty place on

N ushim a, th o u g h , q u ite sacred, I th in k .” It is likely th a t o th e r places called O n o goro jim a p ro b ab ly exist th ro u g h o u t th e In lan d Sea, su p p o rtin g th e suggestion m ade to m e by Joseph Kitagawa th a t O n o g o ro jim a probably d enotes a type o f sacred space an d n o t an actual place. As this m yth g o t redacted in to th e K ojiki and N ibo n g i narratives, its m ean in g was transform ed. 21. W akam ori, A w a jisb im a no m in zo ku , pp. 2 -3 . 22. de P o o rte r, MotoyoshVs Saru g a ku Dangi., p. 159. 2 3 . G unji M asakiyo, “ K ugutsu n o Ki,” in Geind ronshu, ed. H o n d a Yasuji (Tokyo: K inseisha, 1 9 7 6 ), p. 6 1 9 . G u n ji’s title is a p u n o n th e title o f th e tex t by O e no M asafusa discussed in th e previous chapter. 24 . G oshiki C h o Kenkytikai, Goshiki chd shi (S um oto: G oshiki C ho Rekishi K enkyukai, 1 9 8 6 ), p. 4 8 8 . 25 . T his discovery was originally an n o u n ced in H a b a ta k i (published by the Iw ate Ken B unka Shinko Jigyo D an [Iw ate C u ltural P ro m o tio n G ro u p ]), no. 6 (2 0 O c to b e r 198 7 ): 4. F o r a com plete discussion o f the discovery, see Kadoya M itsuaki an d Y am am oto R eiko, “M orioka han no ayatsuri shi suzue shirobe shiryo ni tsu ite ,” Iw a te kenritsu hakubu tsu ka n kenkyii bdkoku, no. 6 (A ugust 1988): 1 -4 6 . 2 6. T h e m eaning o f th e suffix “ tayii” in th e nam es H yakudayu and Kikudayu has shifted th ro u g h o u t Japanese history. O riginally it was a sign o f a c o u rt rank. By th e late m edieval p erio d , it d e n o te d th a t o n e was a principal m em b er o f a p e rfo rm ­ ing g ro u p (o r th a t on e had a shrine rank). 2 7. Som e later texts say th a t H yakudayu “ fell in love w ith and was b e tro th e d to th e d a u g h te r,” reflecting perhaps a co n cern w ith th e sexual m ores o f later tim es. 28. T h e te x t uses th e w o rd rin ji, w hich m eans “im perial o rd e r” and refers to d o cu m en ts such as th e o n e m e n tio n e d above, d ated 1570. It seems th a t as they traveled aro u n d Japan, p u p p eteers had to carry these d o cu m en ts statin g their nam e (often g ran ted th em by an im perial p erso n ) and th eir business. 29 . T h e te x t was reissued by F ukuura K o b u n d o in 1972. 30. M eisho zue frequently w ere g eared to regions w hich w ere adjacent to , o r p art of, m ajor pilgrim age ro u tes, since a pilgrim age was a socially acceptable way to take a nice trip. Awaji is n ear th e Shikoku pilgrim age ro u te , and o n e could take a ferry from w hat is n ow T okushim a to th e so u th ern tip o f Awaji. T h ere were also m iniature pilgrim ages o n Awaji, called m a m e-junrei, m odeled on the larger ro u tes such as S hikoku and Saikoku. See Kodansha Encyclopedia o f J a p a n , s.v. “ m eisho z u e .” 3 1 . M isum i H a ru o n o tes th a t this m o tif o f th e supernatural stran g er m arrying a w om an in th e village is a co m m o n th e m e in p erform ance traditions in Japan. H e reco u n ts th e follow ing storyline: O n e n ig h t a stran g er appears in tow n and is allow ed a n ig h t’s lodg in g . T h a t n ig h t, he has sex w ith a w om an. In th e m o rn in g , he goes to th e beach, gets in to a b o at, an d disappears on th e sea. T h e girl becom es p reg n an t an d later gives b irth to a boy, w ho is n am ed Saburo. H e show s excep­ tional stre n g th a n d g ro w th , an d is m ade chieftain o f th e village. T his storyline has at least tw o m otifs w hich have b earing o n o u r discussion o f th e fo u n d er o f Awaji p uppetry. First, a stra n g e r appears from n o w here and exhibits o v ert sexual behav­ io r (i.e., gets a w om an p reg n an t). S econd, he disappears on th e sea, indicating the m aritim e cosm ology o f pow erful beings th a t com e and go across th e sea to the

o th e r w orld. T hird, th e offspring has the nam e Saburo, often associated w ith Ebisu. T hese m otifs recur in th e text translated below, the Ddkumbo Denki. M isum i H a ru o , “Sasuraibito n o gikyoku,” in Geind shi no m inzoku teki kenkyii (Tokyo: T okyodo Shuppan, 19 7 6 ), pp. 2 1 8 -2 1 9 . 32. I have looked for records o f these litigations and have n o t been able to find them . 33. T h e text uses th e w ord shake, w hich denotes a hereditary priest at a shrine. T he w ord can also be used to refer to a kannushi, b u t since the text refers to M ori Kanedayu as a shake and to M ori T ango as a kannushi, we have to suggest th a t perhaps the w ord shake was used in a m ore general sense. Niimi interpolates shake as shinshoku, which m eans a shrine priest o f low er rank. See N iim i, A w a ji no ningyo shibai, 14. 34. T he text says, “Settsu Amagasaki no Shonenji to iu tera ni tayori,” indicat­ ing th a t he perhaps had friends or family th ere w hom he could “ rely u p o n ”— the nuance suggested by “ ni tayori.” 35. W andering perform ers often sang the Tales o f Heike, and standard narra­ tive styles developed for presenting these stories. These ballad recitations becam e a broad source for joruri recitations, w hich later becam e p u p p et plays. 36. A Japanese reader w ould perhaps u n d erstand th a t a m em ber o f the im pe­ rial family w ould n o t be allowed o u t o f the im perial palace. To explain how the prince was able to see the perform ance, it was therefore necessary to indicate the circumstances. 37. T he text uses the characters “dojo d o g e ,” which literally m ean “those w ho go up (in to th e presence o f the em peror) and those w ho stay d o w n .” 38. Cited in N iim i, A w a ji no ningyo shibai, p. 16. 39. T h e text uses the expression kido a ira ku , which translates roughly as “ re­ joicing, anger, sorrow, and pleasure,” a co m p o u nd th at refers to the range o f em otions in joruri recitation and perform ing arts in general. 40. Ib id ., p. 16. 41. T he family m entioned in these texts, the H ikida family, continues to be a pow erful family on Awaji even today, and the p uppetry group referred to in these texts became the largest and m ost successful o f the joruri troupes on the island. 42. See Edw ard Shils, Tradition (C hicago: University o f C hicago Press, 1981). 43. In Shinoda, N in^yd jo ru ri butai-shi, p. 297. 44. See my discussion o f th e transform ation o f m yths in the Usa H 5 jo -e case in “V iolence, R itual R eenactm ent, an d Ideology.” 45. F or a transcription o f the ICanbun text, see Yamada Shijin, iiDokumbd D enki: Awaji ningyo shibai no shison d en sh d ,” in Doshisha kokubunjyaku, no. 13 (M arch 1978): 97. T he rep o rt o f ICadoya and Y am am oto, “ M orioka han no ayatsuri shi suzue shirobe shiryo ni tsu ite ,” also presents a transcription o f the text. 46. T he text is also referred to as th e D dkum bd y u ra i and the A w a ji hisho (Awaji secret letter). 47. See M iyata N o b o ru , Jisha engi (Tokyo: Iw anam i S hoten, 1975). 48. These scrolls have recently been relocated to the M ihara M unicipal Archives.

49. See Nagata Kokichi, Ikite iru ninjyyo shibai (Tokyo: Kinseisha, 1983), p. 195. O f these eight docum ents, I have personally examined all but that in Nagano Prefecture. It is apparent that these scrolls are all transcribed versions o f an original docum ent, since they are nearly identical. One version I examined at the Awaji Ningyd Joruri Kan in Fukura is clearly a later copy o f the text. It contains a num ber o f mistakes in the characters and leaves o u t a few im portant characters in placename com pounds. These are merely transcription mistakes, no t textual variants. O f the eight, two at the Awaji theater and those at the hom e o f Toyota Hisae appear to be quite old. The M orioka text appears to have been recopied within the last one hundred years or so, m ost certainly from the docum ent taken to M orioka in the middle o f the seventeenth century by Suzue Shirobe. 50. The translation is my own. T he division o f the text into two parts is a device intended to highlight the appropriation o f the Nihonyyi narrative as the opening o f the docum ent. 51. At this point in the text, H yakudayu’s name has the suffix “D o k u m b o .” 52. Literally, the “w orld-opening island,” the place from which the Japanese archipelago was created. 53. Niimi Kanji’s discussion o f the Awaji text makes the assum ption that the Awaji reading o f the name as Wakokuzaki rather than Wadazaki (or Wada hama) is an error. See his A w aji no ningyd shibai, p. 18. C h a p te r 5: P u p p ets o f th e R o ad , P u p p ets o f th e F ield 1. The term ayatsuri was also used to refer to puppeteers during the nine­ teenth century, but this term was no t tied to ritual puppetry per se, whereas the term Dokumbd-mawashiparticularly m eant a person who perform ed Ebisu o r Sanbaso rites. 2. N um erous scholars in Japan have discussed the origins and developm ent o f the Sanbaso rite outside o f the N oh tradition. I suggest that by looking at other rites which resembled the Sanbaso perform ance, it is easier to trace its history. Q uite a num ber o f references to Senzu m anzai hdshi (probably hum an beings and not puppeteers) perform ing with Daikoku dancers (Daikoku being a deity coupled w ith Ebisu) from the fifteenth and sixteenth centuries suggest th at these tw o types o f performances were paired. T he SanbasS and Ebisu pair that we see in the Awaji tradition, I suggest, grows ou t o f this earlier pairing o f Senzu manzai and Daikoku. 3. Neary, Political Protest, p. 12. 4. Niimi, A w a ji no ninjyyo shibai, pp. 169-170. 5. I was unable to interview people o ff Awaji, as it was n o t possible to find villages where one could be sure ritual puppeteers had perform ed. D uring the years I lived in the Kansai area, I encountered several elderly people who recalled Awaji puppeteers com ing at the New Year. Performances in Kobe, Osaka, and the area around Kyoto were com m on until the 1930s according to people on Awaji. Shikoku, particularly the Awa region, has large num bers o f ningyS-mawashi. 6. B oth were considered to be o f lower status than puppeteers working in the theaters m anipulating joruri puppets.

7. It is possible th at this distinction betw een locations o f perform ance is based on the T em po reform s (1 8 3 0 -4 4 ), designed to exacerbate the distinction b e ­ tw een outcasts and th e peasantry so as to increase the control o f the shogunate over th e ir peasants’ tax base, u p o n which the shogunate was largely d ep en d e n t for capital. O n e p o rtio n o f these reform s forbade an outcast from stepping beyond the entryw ay to a peasant’s hom e. Kobayashi Shigeru, “ Kinsei ni okeru buraku kaiho to so ,” Rekishi koron, no. 6 (June 1977): 8 7 -9 4 . 8 . T h o u g h com m on th ro u g h o u t th e Japanese perform ing arts, it originated in the Sarugaku tradition and was originally presented by Shushi perform ers. This ritual perform ance was p resented th ro u g h o u t Japan by itinerant perform ers from as early as th e late eleventh century up th ro u g h th e postw ar period, w ith, needless to say, considerable variation and developm ent in its ritual context and perfor­ m ance style. T h e fundam ental stru ctu re o f the ritual, from the puppets and ritual chant used to the various symbols displayed d u ring the rite, shows rem arkable continuity th ro u g h o u t Japanese history. 9. N ote th e parallel betw een this conception o f the characters in the rite and the perception o f the perform ers as itin eran t o thers, discussed below. 10. In these o th e r perform ing arts, th ere is a strict distinction betw een the dance o f O kina ( O kin a-m a i) and th e dance o fS anbaso (Sanhaso-mai). In the case o f ritual puppetry, however, the Sanbaso perform ance has absorbed the dance o f O kina and th e term s are used interchangeably at tim es, leading to no small am o u n t o f confusion for scholars n o t familiar w ith puppetry rituals b u t aware o f the ritual repertoire in o th e r perform ing art form s in Japan. In the tradition o f Sanbaso perform ance, SanbasS is often closely identified w ith the Leech Child. Som e variations o f th e oral trad itio n even equate the tw o, and thus th e actions o f this character in the perform ance, which are graceful and ritually p o ten t, show the pow er o f a failed creation (abandoned o n the sea by his parents), a fine example o f “negative capability.” 11. This strategy was suggested to me by Karen Brazell. 12. This typology was explained to me by U m azum e M asaru. 13. N agata, Nihon no ningyo Shiba-i, p. 89. 14. This practice also sheds light o n th e role o f H yakudayu as an epidem ic spirit at th e N ishinom iya shrine, as seen in chapter 3. 15. Neary, Political Protest, p. 22. 16. N iim i, A w a ji no ningyo shibai, pp. 1 7 9 -1 8 0 . 17. In versions o f the Sanbasb rite in Kagura, K abuki, and N oh, the dance o f Sanbaso is largely seen as a hum orous skit. T his is probably a later in terpretation o f th e earlier sham anic dance, perhaps to make fun o f the frenzied gestures o f shamans in trance. A n u m b er o f o th e r Sanbaso dances around Japan, how ever, are quite ribald. (For exam ple, he is present at his ow n conception and crawls up betw een his parents as they are m aking love.) 18. Som etim es masks are p u t on tw o o r even three o f the puppets, and they all g et possessed, alth o u g h this seems to be a variation o f the perform ance d e ­ p en d en t largely on th e n u m b er o f perform ers available and the desire for real dram atic force. 19. This is my interp retatio n o f th e significance o f the h at’s design.

20. M aurice B loch, “Sym bols, S o n g , D ance and F eatures o f A rticulation: Is R eligion an E xtrem e F o rm o fT ra d itio n a l A u th o rity?” European J o u rn a l ofSociol­ ogy 15 (1 9 7 4 ): 5 5 -8 7 . 21 . T h e im plication h ere is th a t v om iting, unlike defecation, is beyond o n e ’s c o n tro l, an d hence he sh o u ld n o t be seen as guilty o f an offense. 22 . P hilippi, K o jik J pp. 7 9 - 8 0 ; em phasis m ine. 2 3 . Ib id , pp. 8 0 -8 1 . 2 4 . L aurel Rasplica R o d d , trans. Kokinshu: A Collection o f Poems A n c ie n t a n d M odern (P rin ceto n : P rin ceto n U niversity Press, 1 9 8 4 ), p. 35. 25 . K onishi Jin ’ichi, A H istory o f Japanese L itera tu re, vol. 2 , The Early M iddle A ges, trans. A ileen G a tte n , ed. Earl M iner (P rin ceton: P rin ceto n U niversity Press, 1 9 8 6 ), p. 113. K onishi sees th e m edieval Japanese perception o f th e spirit o f w ords as influenced by a n o th e r strain o f m agical language in Japanese religions, nam ely th e co n cep t o f d h a ra n t from esoteric B uddhism . 26 . G ary L. E bersole, K itu a l Poetry a n d the Politics o f D eath in E arly Ja p a n (P rin ceto n : P rin ceto n U niversity Press, 1 9 8 9 ), p. 19. In ch ap ter I , “ Ritual P o etry in th e C o u rt,” E bersole discusses th e ritual p o ten cy o f song and poetry. See espe­ cially pp. 1 7 -2 3 . 27 . Frank H o ffh a s tran slated a n u m b e r o f ritual chants used in related O kina rites. H e correctly situates these p o etic texts w ithin th eir perform ance c o n te x t to rend er th e cryptic lines m eaningful. H o ff, “T h e ‘E v o catio n ’ and ‘B lessing’ o f Okina: A P erform ance V ersion o f Ritual S h am anism ,” A lcheringa/Ethnopoetics, n .s., 3, no . I (1 9 7 7 ): 4 8 -6 0 . 2 8 . T h e tex t o f th e Sanbaso rite is th e N o h tex t o f th e D ance o f O kina. In m ost p u p p et versions o f Sanbaso, large parts o f th e p erform ance have been ren d ered incom prehensible th ro u g h w h at can only be regarded as a m isu nderstanding o f the classical lines in th e te x t by p erform ers, w h o th e n h an d ed d o w n the variations to su b se q u e n t generations. I have fo u n d th e m eanings o f using w hat could be considered a garbled tex t to be in terestin g , b u t for th e purposes o f this discussion, I have follow ed th e advice o f th e N o h scholar K aren Brazell and have used the N o h yokyoku to “re c o n stru c t” th e text. 29. I discuss th e significance a n d possible m eaning o f these u tterances below. Possible origins an d m eanings o f these syllables have been suggested. T h e Russian linguist o f Japanese and U ral Altaic languages, Sasha Vovin, has suggested th a t the w ords com e from ancient K orean an d may be references to the m oon. 30. T h ese im ages reinforce th e visual im agery flashing in to view o n th e p u p ­ p e t’s costum es. R eferences to th e crane an d tu rtle are tw ofold. T h e costum e o f Sanbaso has d etailed draw ings o f b o th cranes and im aginary turtles. T h e colors o f his costum e— red , w hite, black, and g o ld — are colors used to d ep ict the senba z u r u (th o u san d -y ear crane), a w h ite bird w ith a g o ld -co lo red beak, a red crest, and black u n d e r its wings. T h e tu rtle , in ad d itio n to th e picture o n Sanbaso’s costum e, is rep resen ted in th e co stu m e o f Senzai by th e six-sided m otif. T his shape is also stitch ed o n to th e b o tto m o f S anbaso’s hat. 31. T h e I in e is f r o m R ydjin hisho. 32. H e re , th e m eaningless syllables are tran sfo rm ed in to on o m ato p o eia. “ D o do d o ” suggests th e th u n d e rin g so u n d o f a w aterfall.

33. T h e Japanese for “three so ngs” is sankyoku, and can refer to three styles o f m usic in N o h , namely th e ryusen, which im itates the sounds o f flowing water, the m kuboku, w hich is a w oodpecker, and th e ydshinso. Each type o f refers to a differ­ en t use o f perform ed sound to im itate the natural w orld. Im itative magic in the ritual (described by the general term nazoraeru) implies in this case the ability o f language, w hen perform ed in th e p ro p er ritual context, to create the realities to which it refers. 34. T h e “ old m en” are to O kina, Senzai, and Sanbasd, w ho have been sum ­ m oned by the music, chanting, and felicitous imagery. 3 5 . This is literally the “dance o f te n thousand years,” used widely in felicitous occasions in Japanese perform ing arts. Recall th e discussion o f the toka dance and Senzu m anzai in chapter 2. 36. It is interesting th at in the p uppet version o fth is perform ance, as opposed to the N oh version, masks are used to d en o te possession. I t is possible th at origi­ nally p u ttin g masks on puppets was a way th a t puppeteers im itated the N o h th e a ­ ter by doing ay a t s u n Noh, discussed briefly in chapter 3. 37. T he term ka m i asobi (deity play or playing w ith the sacred) refers to the aspect o f a m atsuri which is in ten d ed to entertain, and thus appease, the deity w ho has been invoked. 38. G enjoraku, literally “goin g to th e capital m usic,” refers also to a type o f Gagaku dance, in which a snake is su b d u ed by a person w earing a dem on mask. 39. T he Japanese line appears to leave th e subject intentionally am biguous. T he line could also be read, “T og eth er we enjoy the voice m ade by the pine trees and the w in d .” In this final line o f the text, the am biguity o f the relational aspect o f th e language allows all th e disparate threads o f th e text—th e dancers, th e im ­ agery, and the audience— to be woven into one final m om ent. 40. O ne interpretation o f these syllables, apparent in notes on the N o h ver­ sions o f this piece, suggest th a t these syllables are intended to be the hum an voice im itating th e musical in strum ents w hich are used to lure deities, namely the tsuzum i and fue (a small handheld d ru m and flute). Yokyoku zenshii, vol. 12 o f Nihon onjjaku zenshii (Tokyo: N ihon O ngaku Z enshu Kikokai, 1948), p. 92 n. T he entire N oh version o f this chant, “ O k in a,” is transcribed in this collection, altho u g h it differs from th e p u p p et version o n som e points. T he N o h version presents O kina as the shite (principle role), Senzai as the tsure (subordinate role), and Sanbaso as th e K yogen actor (a h um orous acting position, vulgar in contrast to the refined N o h sensibilities). 41. A kalpa is defined by Sir M onier M onier-W illiams’ Sanskrit-English dic­ tionary as “a fabulous period o f tim e (a day o f Brahm a or one thousand Yugas, a period o f four thou san d , three h u n d red and tw enty millions o f years o f m ortals, m easuring th e duration o f the w orld; a m o n th o f Brahm a is supposed to contain thirty such kalpas; according to the M ahAbharata, twelve m onths o f Brahm a co n ­ stitute his year, and one h u n d re d such years his lifetime; fifty years o f B rahm a’s are supposed to have elapsed, and we are now in th e Svetavaraha-kalpa o f the fifty first; at the end o f a kalpa the w orld is annihilated. . . . with B uddhists, the kalpas are n o t o f equal d u ra tio n .” W hether one accepts the o rth o d o x Sanskrit definition o f a kalpa or the reference to tim e or defines the term as the length o f tim e needed to w ear dow n a rock w ith a robe o f feathers, th e reference is to a very long tim e.

Since it is an incantation for longevity and fertility, it is optimistic indeed. Sir M onier Monier-Williams, Sanskrit English Dictionary (Oxford: Oxford Univer­ sity Press, 1979), s.v. “ kalpa.” (In a Pali narrative explaining a cognate term , kappa, the robe o f feathers is a silk cloth.) 42. Across the N aruto Straits from Awaji there lived an Ebisu-kaki who con­ tinued to make rounds with his puppet until the early 1950s. A num ber o f ph o to ­ graphs in folklore books and studies o f the Inland Sea show this puppeteer carrying his smiling puppet, which was about half his height. In many o f the p h o ­ tographs, he is standing holding his puppet at the entrance to a house, dressed casually and with a bag on his back, as if going on a stroll. See for example the dear photograph in Kawajiri, Nihon ningyogeki hattatsu shi—koy p. 90. 43. Gunji, “ ICugutsu no Ki,” p. 619. 44. This practice still continues, and all o-fuda, protective talismans (o-m am ori ), and other religious articles o f this sort are taken to shrines during the New Year period and burnt in a receptacle for the disposal o f sacred objects (dondoyaki). It is understood that they are “exhausted” from the previous year. The cycle o f purity and efficacy must be renewed for the com ing year. 45. Sec the transcription o f the Ebisu-mawashi’s chant in Tosa no m inzoku geino. Niimi also transcribes a popular one from Awaji in A w aji no ningy5 Shibai1 p p . 177-179. 46. In theater and ritual studies, there is a great deal o f discussion about the issue o f transform ing rituals into theater, I discuss this issue in greater depth in C hapter 6. To a certain extent the performances by Ebisu- and Sanbaso-mawashi in this chapter already reflect the early stages o f this process o f transformation. This is not wholly a tw entieth-century phenom enon, however. D uring the nine­ teenth century, when two- , three- , and four-day puppetry festivals (called ningyd matsuri) on Awaji became popular, there was great dem and to have these itinerant arts made available. Both the Sanbaso and Ebisu pieces then underw ent a great transform ation and were elaborated into single pieces with several performers and a set script. In the case o f Ebisu-mai, rather than the traditional hitori-zukai (sin­ gle puppeteer) m ethod o f puppet m anipulation, Ebisu is m anipulated by three puppeteers. This performance absorbed many motifs and scenes from popular itinerant performances while also coming to resemble more a joruri performance with chanters, multiple manipulators, and elaborate staging. A striking feature of the Ebisu rite was the addition o f a scene at the beginning indicating the kadozuke origins o f the rite. Ebisu would appear at a hom e asking for sake, and the house­ holder would serve the deity with great style (perhaps some staged wishful think­ ing on the part o f puppeteers, who were no t always so well treated by hosts in the kadozuke context). This more formalized performance o f Ebisu-mai could be presented at these festivals because there was more time to prepare an elaborate performance. The text and performance description o f Ebisu-mai presented here reflects this transform ation, in which we can see a num ber o f individual scenes fused into one hum orous performance. In the rite, since Ebisu catches the first fish o f the season, the proper channels for a successful fishing year are acknowledged and a productive and safe year at sea is requested. We will use this stage context as the basis for our discussion because it is the m ost com m on, the m ost representa­ tive o f the deity’s character, and also the most lively. The rite, one o f great merri-

m ent and intoxication, clearly reveals th e unity o f perform ers and audience. T he hum an audience is at once th e host o f the rite, th e recipient o f its efficacy, and a participant in its festivities. 47. Sakurai T okutard, M in zo ku g irei no kenkyii (Tokyo: Yoshikawa K obunkan, 1987), pp. 1 5 3 -1 5 4 . 48. M ura Torasuke, A w a ji ndgake jo ru ri shibai, pp. 1 0 4 -1 1 1 . 49. This is th e sam e collective th a t agreed to a “ reen actm en t” ofE bisu-m ai on th e island o f Susaki in Fukura h arb o r in 1990. See the discussion in chapter 6. 50. F rom p h o to in M une, A w a ji nogake jo ru ri shibai, p. 106. 51. Ibid. 52. Ib id ., p. 107. 53. This is a recessed area in a hom e or office used to display pieces o f artw ork, flower arrangem ents, etc. See ibid., p. 105. 54. A t th e Ebisu festivals I atten d ed d u rin g my studies on Awaji, w hen p resen t­ ing oneself before the shrine, one drinks a small cup o f sake w ith the deity. While this is also done in shrines dedicated to o th e r kam i, it is a very im p o rtan t part o f the worship at an Ebisu festival, as this deity likes to drink. 55. Ib id ., pp. 1 0 8 -1 0 9 . 56. A com m on feature o f Ebisu festivals was the large am o u n t o f alcohol co n ­ sum ed. A village could usually afford to host only one o f these festivals a year at best. T oday o n Awaji th ere has been a real effort to reduce the consum ption o f alcohol at these festivals, because local cam paigns to address alcoholism have had far-reaching effects. B ut num erous people to ld me about how dru n k people used to g et a t these festivals. 57. An im p o rtan t p art o fE b isu -m a i was th e distribution o f o-fuda depicting Ebisu to people in fishing villages. O -fuda were u n d ersto o d to em body the pow er and sacred au thority o f th e central shrines (usually N ishinom iya) wherever they were hung. T hey were frequently p u t in the small kam idana in the cabin o f a boat and invoked th e presence o f th e deity. An o-fuda was som etim es un d ersto o d to have magical abilities, as in the case o f H yakudayu o-fuda described in chapter 3. T h e central shrine for E bisu w orship in Japan, as we saw in chapter I , was th e N ishinom iya Ebisu Daim yojin Shrine. Because puppeteers presenting the Ebisu-m ai th ro u g h o u t Japan were technically sacred specialists representing this shrine, the o-fuda they d istributed w ere those bearing th e nam e N ishinom iya. For this reason, Ebisu-kaki presenting th e Ebisu rite had to have prior approval o f the Nishinom iya shrine. As we have n o te d , the relationship betw een these puppeteers and the cerem onial centers was often a strained one, because the p u p ­ peteers becam e popular and in tro d u ced their ow n popular interpretations o f Ebisu w orship in to the perform ance. N evertheless, the arrival o f the Ebisu-kaki at New Year w ith th e new and fresh o-fuda for the w orship hall o f the local N ishi­ nom iya shrine served to revitalize the shrine and restore contact w ith the cerem o­ nial center, no m atter how far away. W hen the p u p peteer b ro u g h t w ith him the sacred pow er and authority o f the cerem onial center and bestow ed it on the small w orship sites scattered around the Inland Sea, he was perform ing a service for people to o busy to make annual pilgrim ages to N ishinom iya to procure ritual im plem ents such as o-fuda.

58. T his tran slatio n is based u p o n th e tex t used by the Awaji p u p p e t th ea te r and supplied to m e by th e th e a te r director. A n alternate version used by th e Tosa E bisu-m aw ashi (across th e straits o n S hikoku) can be found in Takagi Keio, ed., Tosa nojjeino: Kochi ken no m in zo ku g ein o (K ochi-shi, 1 9 8 6 ), pp. 2 8 7 , 2 88. 59. T h e Japanese is in u i no ho yori (from th e direction o f in u i). I n u i c o rre ­ sponds to n o rth w e st and is a d irectio n in th e T aoist cosm ology. 60. Literally, “ th e b est g o d o flu c k in th e th ree co u n tries.” T h e reference is to the th ree realm s from S h in to and p o p u lar Japanese cosm ology: th e realm o f hu m an beings (U tsush iy o ), th e realm o f th e dead (yo m o -tsu -k u n i), and the realm o f heaven (Takam a ga hara). 61. O -m ikki (sake for th e g o d s), used for special ritual purposes. 62. A sa k a zu k i, a small flat cup, is used for serving sake on cerem onial and polite occasions. 63. T h e nam e o f th e era in w hich he is said to be bo rn is fictitious, w ritten w ith the characters m eaning “luck” and “v irtu e .” To be born on the th ird day o f the first year o f an era, ju st before sunrise, is as felicitous a tim e o f birth as one can im agine. A child b o rn easily is also th o u g h t to be extra lucky. 64. Ubuyu. In Japan, th ere is th e cu sto m o f giving a n ew b o rn baby a b ath in ritually p rep ared w ater to rem ove th e p o llu tio n o f ch ildbirth. 65. E bisu is always d ep icted w earing a kazaori eboski, a type o f black h a t care­ fully folded to flop over o n one side. 66. Som e people rem em b er th e ir fam ily’s g iving sake cerem o n io u sly to the ddkum bd-m aw ashi at N ew Year. 67. I have been to ld th a t in som e versions o f this perform ance, once th e nam es o f all actual fish have been given, E bisu starts to m ake up nam es to get th e chance to drin k m ore. S om etim es, this rite is p erfo rm ed for occasions o th e r th an good fishing. In such a case, th e event for w hich it is perfo rm ed is in serted here. F or exam ple, if th e piece is presen ted at a local school g athering, the wish may be for “g o o d stu d ies.” A t a h o m e fo r th e elderly, it m ay be for “g o o d h e a lth .” A t a sum m er rural festival, for “ a g o o d h arv est,” etc. A gain, if th e event is o th e r th a n a fishing rite, appropriate an d frequently very clever wishes are in serted th ro u g h o u t this section, m u ch to th e am u sem en t o f th e audience as th e perform ers are able to second guess p eo p le’s real in ten tio n s for particip ating in th e rite. Perform ers can also in te rje c t co m m en ts a b o u t individuals in the audience, usually w ith a slight elem en t o f teasing. 68. T hese are place-nam es near Awaji, w hich can be changed to indicate any place from w hich fisherm en set out. 69. A t this p o in t in th e p erform ance, th e p u p p e te er may perform fam ous dances and skits from o th e r Japanese trad itio n s, show ing o ff his prow ess and know ledge. This extends th e p erform ance and creates an e n terta in m e n t rite w ithin th e rite. 70. In th e original these last lines are aO ki wa tairyo, oka wa m a n za ku , shdbai hanjo, tsukisenu jyodai koso m edctakere 71. E da m a m e are soy beans in th e p o d , prep ared by p ar boiling th em in salt water. T h ey are eaten cold as a snack. 72. N iim i, A w a ji no nm gyo Shibai, p. 165.

73. An excellent p h otographic record o f the nogake butai jo ru ri was m ade by M une Torasuke and published u n d er th e title A w a ji ndgake joruri shibai (Tokyo: Sogeisha, 1986). 74. See H o ri, Folk R eligion in Japan, p. 2 1 . W hile I do n o t discuss them in any d ep th , there were also perform ances held all over this section o f Japan on special stages in H achim an and Ebisu shrines, called nogakegoya. These stages differ from their tem porary coun terp arts in th a t they rem ain up all year ro u n d and te n d to be used for a n u m b er o f events. K udo Takashi (A w a to A w a ji no ningyo shibai [T o­ kushima: n .p ., 1 9 7 8 ], pp. 9 5 -1 2 2 ) lists in detail th e location and style o f all the perm anent nogakegoya in Tokushim a Prefecture. T he region o f Japan th at in ­ cludes the low er half o f Awaji and th e adjacent province on Shikoku has the third highest n u m b er o f “shrine stages.” 75. Raz, A udience a n d Actors, p. 28. 76. See for example Izum i F usako’s discussion o f Sanbaso in her K ashira no keitd (Miyazaki: n .p ., 1984), pp. 3 6 4 -3 7 1 . 77. This is called in Japanese kannigiw ai. See S onoda M inoru, “T he Religious Situation in Japan in R elation to S h in to ,” A c ta A siatica 51 (1987): 1 -2 1 , esp. 8- 12.

78. This sort o f perform ance was last presented at the N ational T heater by the Awaji tro u p e in 1960. This nana-bake was described to me by U m azum e-san in the sum m er o f 1988. 79. In 1988 I interview ed m any elderly people concerning w hat they rem em ­ bered ab o u t th e annual perform ances. W ith o u t fail, everyone m entioned the dogugaeshi as one o f the m ore m em orable features o f the perform ances they had seen d uring the heyday o f Awaji p uppetry w hen o u td o o r perform ances were th e norm . 80. See M u n e , A w a ji ndgake jo ru ri shibai. 81. F o r a b rief general reference in E nglish, see LaFleur, L iquid Life, pp. 1 4 4 -1 4 5 . 82. T he reverse was also tru e. A com m on practice u p o n th e death o f a p u p p e­ teer was to have his puppets m anipulated by o th e r puppeteers to present his funer­ ary rite. O n Septem ber 1 ,1 9 8 8 , the funeral o f a famous puppeteer in th e Bunraku theater in Japan was carried o u t in p art by his puppets, m anipulated by o ther m em bers o f th e troupe. T he female puppets were dressed in w hite, as w ere the u n h o o d ed puppeteers m anipulating them . 83. M y friend U em ura Yasuko gave m e som e suggestions and cases for this classification. C h ap ter 6: P u p p e ts a n d W h irlp o o ls

1 . This organization was foun d ed in 1950 by th e M inistry o f E ducation, o sten ­ sibly to p ro m o te the recognition and preservation o f local perform ance traditions th ro u g h o u t Japan. This new form at, w ith g overnm ent m oneys, was a reform ula­ tion o f th e Kyodo Buyo to M inyo no Kai (Local D ance and Folk M usic C onven­ tio n), w hich held annual perform ances from 1925 th ro u g h 1936. 2 . See, for example, the late 1980s film M a c A rth u r’s Children (a creative trans­ lation o f “Setonaikai no Yakyii K urabu” ), set on Awaji d u ring the O ccupation. A

brief puppetry scene in the film shows performers from the theater presenting Shiki SanbasS. 3. The nativist origins o f Yanagita’s enterprise are carefully set forth in Harootunian, Things Seen and Unseen, pp. 4 1 8 -4 3 6 . An excellent collection o f essays critiquing the agendas ofYanagita’s enterprise is J. Victor Koschman, Oiwa Keibo, and Yamashita Shinji, eds., International Perspectives on Yanagita Kunio and Jap­ anese Folklore Studies (Ithaca: Cornell University Press, 1985). I also discuss this issue in Swanberg-Law “ Puppets o f the R oad,” chapter I. 4. In his work Zuroku Nihon no m inzokugeind (Tokyo: Kinseisha, 1960), p u b ­ lished in 1960, H onda established five categories into which folk perform ance traditions can be classified: kagura (performances within Shinto shrines or for Shinto kam i); dengaku (performances connected with the cultivation o f rice); fu ryii (performances used to divert disasters and pestilence); shukufukugei (dances and performances presented on felicitous occasions and to bless events); zn d g a iraimyaku (performances o f foreign, largely Chinese, origin). 5. I would argue that the use o f the term shinkd to demarcate Japanese religios­ ity already shows the influence ofW estern scholarship and conceptions o f religios­ ity, where orthodoxic religion is viewed as normative. 6. A thorough discussion o f this legislation and its ramifications over the last three decades in Japan is Barbara E. T hornbury’s excellent article “The Cultural Properties Protection Law and Japan’s Perform ing A rts,” Asian Folklore Studies 53, no. 2 (1994): 211-2 2 7 . The chronology o f this discussion is drawn from her study. 7. I discuss this ritual in “Violence, Ritual Reenactm ent, and Ideology.” 8. Many o f the photographs o f prewar and early postwar performances in this book were taken by Fudo Saiichi. 9. Fred Davis, Yearning fo r Yesterday: A Sociology o f Nostalgia (New York: Free Press, 1979), pp. 1—2. 10. Ibid., p. 2f. 11. After surveying university students in the mid 1970s, Davis reports that while only half o f the students associated the word nostalgia with homesick­ ness, many more associated it with “warm, old times, childhood, yearning.” Ibid., p. 4f. 12. Ibid., p. 6. 13. Stephen Nussbaum , introduction to Yanagita, “The Evolution o f Japanese Festivals: From M atsurito S a ir e if in Koschmann, Oiwa, and Yamashita, Interna­ tional Perspectives on Yanagita K unio, pp. 168-69. 14. Raymond Williams, The Country and the City (New York: Oxford Univer­ sity Press, 1973). 15. Fredric Jameson, “Postm odernism , or the Cultural Logic o f Late Capital­ ism,” New Left Review, no. 146 (S eptem ber/O ctober 1984): 53—92. 16. Ibid., p. 85. 17. Ibid., p. 66. 18. Ibid., p. 71. 19. Davis, Yearning fo r Yesterday, p. 31. 20. Jameson, “Postm odernism ,” p. 72. 21. Ibid., p .7 3 .

22. H aro o tu n ian argues th at this feature o f the contem porary folklore m ovem ent has its roots in the Tokugaw a nativist m ovem ent (kokugaku). By the early nineteen th century, the kokugaku agenda had shifted from reclaim ing a philosophically “au th en tic” Japanese spirit to the form ulation o f a theory o f peas­ antry. Such a theory o f peasants and peasant labor addressed a grow ing need for a co herent social discourse to m aintain the social o rder in the face o f the grow ing unrest w ith th e rigid class system. This th eo ry o f peasantry, w rites H aro o tu n ian , was m ost clearly form ulated by H irata A tsutane, w ho equated the aohitogusa (lit­ erally “ blue grass p eople,” his term for “th e folk” w ho spring up everywhere and always grow back and bounce back from even th e w orst o f conditions) w ith the landless peasants and the A ncient Way (kodoron) w ith agricultural labor. T he o u t­ com e in this direction in kokugaku was th a t nativist discourse posited th a t the w orld o f th e peasant becam e th e w orld o f the real, and yet the public and external discourse o f w hat the peasant was and w hat peasants did was set up n o t by peasants b u t by scholars such as H irata A tsutane. See H aro o tu n ia n , Things Seen a n d U n­ seen, p p . 2 3 -2 5 . 23 . This thirty -th ree site pilgrim age rou te in w estern Japan (the literal m ean­ ing o f th e ro u te ’s nam e) is dedicated to the bodhisattva K annon (Avalokitesvara in Sanskrit) and is said to have been established in the n in th century. Pilgrim age is an o th er good exam ple o fw h a t has com e to be considered an “exotic practice.” Ian Reader suggests the recent popularity o f pilgrim age reflects a nostalgic return to the past. See Ian Reader, “From Asceticism to the Package T our— T he Pilgrim ’s Progress in Japan,” Religion 17 (April 1987): 140. 24. Interestingly, the first tim e I saw this scene perform ed was at the third pu pp etry sum m it in the fall o f 1993. A lthough it was presented by the students from the local high school, it is a shattering piece o f theater. M any o f th e theater enthusiasts w ith me th a t day w ere m oved to tears and speechlessness by the scene. 25. T his conversation occurred o n July 22 , 1 9 8 8 , in th e th eater in Fukura. 26. R enato R osaldo, C ulture a n d Truth: The R em a kin g o f Social Analysis (Boston: Beacon Press, 1989), p. 70. 27. Ib id ., p. 69; emphasis mine. 28. M arilyn Ivy, “ D iscourses o f th e Vanishing in C ontem porary Japan” (P h.D . diss., C ornell University, 198 8 ), pp. 3 9 -4 4 . 29. I borrow th e phrase “a past o f things p resen t,” a reversal o f th e A ugustinian d ictum , from Joseph Kitagawa’s article by th at nam e, and I shall make the same p o in t a b o u t the uses o f history in Japan. See Kitagawa “ ‘A Past o f T hings Pres­ e n t,” ’ pp. 4 3 -5 8 , esp. 5 3 -5 8 . 30. Richard Schechner, Between Theater a n d Anthropology (Philadelphia: U n i­ versity o f Pennsylvania Press, 1985), p. 75. 31. Ib id ., pp. 7 7 -7 8 . 32. See my article “T he P u p p etas B ody S u b stitu te,” in Religious Reflections on the H u m a n Body, ed. J. M. Law (B loom ington: Indiana U niversity Press, 1995). 33. I atten d ed this ritual perform ance w ith ab out thirty fisherm en and my re ­ search assistant, Jiro N akam ura, to film th e event. N akam ura was at the tim e a stu d en t in the C ornell C ollege Scholar Program .

34. See my discussion o f this center in chapter 4. 35. Again, my research assistant Jiro Nakamura was videotaping the perfor­ mance. 36. Several years after this day, the perform er who m anipulated the puppets told me th at th ro u g h o u t the entire rite and the repetition, tiny ants had crawled up inside his haori hakam a and were driving him crazy. H e said that because this event was being filmed, he did n o t want to jum p up and scream, but had it “only been a real perform ance” (honto no sbibai dake dattara), he would have probably done so. 37. Barbara T hornbury m entions the “sum m it” m odel being used in preserva­ tion o f folk perform ing arts. See her article “From Festival S ettingto C enter Stage: Preserving Japan’s Folk Perform ing A rts,” A sinn Theater Journal 10, no. 2 (fall 1993): 163-178. She briefly m entions the problem s inherent in presenting rituals as stage entertainm ent. 38. I discuss below how my presence as a foreign Ph.D . was a scripted part of these trips. 3 9 . For example, in the third sum m it, a new puppetry group from Kobe called “Ningyo O ur K obe” participated. T hough strictly am ateur, they are working on reviving joruri in Kobe at a popular level. 40. See Thornbury, “Cultural Properties Protection Law,” for a discussion o f this view. 41. Sasahara Ryoji, “Kimyo na butai, bimyo na butai: M inzoku geino taikai to m inzoku geino kenkyusha,” M inzoku geind kenkyii, no. 12 (1990): 13-17. 42. Keir Elam, The Semiotics o f Theatre and D ram a (London: M ethuen, 1980), p. 7. 43. Jiri Vetrusky, “ M an and O bject in the T heater,” in A Prague School Reader on Esthetics, Literary Structure and Style, ed. Paul Garvin (W ashington: G eorge­ town University Press, 1964), p. 84. 44. See the beautiful black-and-white photographs o f farmers m eeting to practice joruri recitation in hom es after leaving their m uddy farming shoes at the doorstep in M une Torasuke’s collection o f photographs, A w a ji nogake Joruri shibai, esp. pp. 28, 29. This book, a collection o f photographs taken in the early 1960s, shows activities o f Awaji puppeteers as they built their tem porary stages in the countryside. The collection itself was part o f a nostalgic process o f “capturing the vanishing.” 4 5 . A related point, perhaps overemphasized in Japan these days, concerns the use o f whalebone in the inner workings o f puppet heads. As this is no longer allowed, puppet-head makers now use plastic, b u t they claim that one cannot make a puppet head properly w ithout whalebone. This has become an issue in the environm ental m ovem ent in Japan, as many traditional arts that use whale m ateri­ als have had to substitute other materials. A return to the traditional is not always in line with a return to a m ore “eco-friendly” life. 46. I have also noticed this tendency to make the perform ance less a reenact­ m ent and m ore o f a relevant perform ance am ong the high school groups perform ­ ing it on Awaji. At the third ningyo sum m it in O ctober 1993, the Mihara High School Puppetry Club presented the Ebisu perform ance. The tayu, a young girl o f

only fourteen, was q uite pow erful, and th e script included many references to teenage life and studying for exams, which drew many laughs from the audience. 47. U m azum e M asaru, “Awaji ningyo shibai,” in A w a ji ningyd kashira, ed. H ySgo Kenritsu Rekishi H akubutsukan (H im eji: H yogo Kenritsu H akubutsukan, 1985), pp. 2 1 -2 2 . My first research trip, after my initial introduction to the area in 1977, was in 1984, w hen I w ent to Fukura to set up the field research that began in 1987. 48. His reference, I assum e, was to the problem o f ritual puppeteers as purifiers and therefore as outcasts. 49. In Japan, unlike the U nited States, m uch scholarship on issues o f regional interest gets d one by high school teachers. Academia is n o t lim ited to universities, and frequently there is m ore exchange betw een local scholars and national univer­ sities than one sees in the U nited States. 50. U nion Internationale des M arionettes, an international consortium o f puppeteers and perform ers w ith representatives in several countries. T he director o f U N IM A in Japan is Takeda Shunosuke, w ho directs puppet th eater in Kanagawa.

Bibliography Ja p a n e se L a n g u a g e S o u rc e s A m ano Fum io. “ O kina S arugaku no seiritsu: jS gyodo shflshSe to no k a n re n .” B u n g a ku 51, no. 7 (1 9 8 3 ): 1 6 6 -1 7 7 . A m ino Yoshihiko. M uen, kugai, raku: nihon chUsei no jiy ii to heiwa. H eib o n sh a Sensho 58. Tokyo: H eib o n sh a, 1978. A m ino Yoshihiko et al. Chusei no tsu m i to batsu. Tokyo: T okyo D aigaku Shuppankai, 1983. F u d o Saiichi. A w a ji ningyB no yurai. M ihara-gun: Awaji N ingyo H o z o n Kai, 1937. Fujikaw a Yu. N ihon Shippei Shi. Tokyo: H e ib o n sh a, 1969. Fujiw ara T oshio. “Shin S arugakki.” In K o d a i seiji shakai skiso, ed ite d by Yam agishi T okuhei e t al., pp. 1 3 3 -1 5 2 . Tokyo: Iw anam i S h o ten 1979. G oshiki C h o Kenkyu ICai. Goshiki chd shi. S u m oto: G oshiki C h o Rekishi ICenkyukai, 1986. H am agishi Koichi. “ ICishi tanabe n o ebisu m a tsu ri.” M a tsu ri, no. 4 6 (spring 1986): 5 2 -6 1 . H arad a T om ohiko. “ D an zaem o n Y uishogaki k aidai.” In N ihon shomin seikatsu shiryo shusei, ed ited by H arad a T o m o h ik o e t al., pp. 4 2 7 -4 2 8 . Tokyo: Sanichi Shobo: 1971. H ayashiya T atsu sab uro . Chusei b u n ka no kicho. Tokyo: T okyS D aigaku Shuppankai -----------Chiiseig ein o shi no kenkyu: ICodai kara no keisho to sBzo. Tokyo: Iw anam i S h o ten , 1960. H o n d a Yasuji. M in zo ku geinB saihdroku. Tokyo: N ih o n H o sS S huppan Kyokai, 1971. -----------“ N S n o hassei: m ikom ai kara nS e .” B u n g a ku (M arch 1977): 2 6 9 -2 8 1 . H o ri Ichiro. M in k a n shinkS. Tokyo: Iw anam i S h o ten , 1951. -----------M in k a n shinkB shi no shomondai. Tokyo: M iraisha, 1971. -----------N ihon no sh a m a n izu m u . Tokyo: ICodansha, 1971. -----------W ggakuni m in k a n shinkd shi no kenkyu. 2 vols. Tokyo: Sogensha, 1 9 5 3 1955. H y S g o ken shi h en sh u sen m o n annai kai. HyBgo ken shi. Tokyo: ICawakita Insatsu Kabushiki ICaisha, 1974. Ishio Yoshihisa. M inshu undo kara m ita chiisei no h inin. Tokyo: S an’ichi S hobo, 1981. Iwasaki T oshio. HonpB shBshi no kenkyu. Tokyo: Iwasaki H akushi G ak u i-ro n b u n Shuppan ICSenkai, 1963. Iw ate Ken M in zo k u n o ICai. Oshira-sama: oshira-sama shim pojium u kiroku. M orioka: Iw ate ICenritsu H a k u b u tsu k an S h u p pan, 1987. Izum i Fusako. K ashira no keito. M iyazaki: n .p ., 1984.

Kadoya M itsuaki and Y am am oto R eiko. A w a ji ningyo to Iw ate no geino shudan. M orioka: S higunaru Sha, 1991. ---------- “M orioka han no ayatsuri shi suzue shirobe shiryo ni tsu ite.” Iw ate kenritsu hakubutsukan kenkyii hdkoku, no. 6 (A ugust 1988). Kawabata Yoshimizu. “ O ita-ken ni o keru kug u tsu keito— geinb m u ra.” Usa-shi, O ita, 1987. Kawajiri Taiji. N ihon ningyogeki hattatsu shi—ko. Tokyo: Bansei Shobo, 1986. Kim ura H idcshiro, ed. A w a ji no k u n i meisho zu e. Vol. 3. Sum oto: Fukuura B unzo Awaji, 1895. Kita Keiji. Hyogo ken m in zo ku geino shi. Tokyo: Kinseisha, 1977. Kita Teikichi. “ D osojin to kaite sai n o kami to yom u k o to no k 6 .” M inzoku to rekishi I , no. 3 (19 2 1 ): 2 9 7 -3 0 3 . ------------“Sanjo hoshi ko: Kodai shakai soshiki no kenkyu.” M inzoku to rekishi I, no. 3 (1921): 1 2 3 -1 2 6 . Kitahara Taisaku. “Sanjo hoshi ko: kodai shakai soshiki no kenkyu.” M inzoku to rekishi 3, no. 4 (19 2 1): 1 -1 0 . ---------- Senm in no kdei— waga kutsujoku to hankd no hansei. Tokyo: C hikum a S hobo, 1974. Kitam ura Tetsuro. “N in g y o .” N ihon no bijutsu, no. 11 (M arch 1967): 1-30. Kobayashi Shigeru. “Kinsei ni okeru buraku kaiho to so .” Rekishi koron no. 6 (June 1977): 8 7 -9 4 . Kotaka Kyo, com p. C huseigeindshi nenpyd. Tokyo: M eicho S huppan, 1987. Kudo Takashi. A w a to A w a ji no ningyo shibai. Tokushim a, 1978. Nihon geino no shigen teki kenkyii. Tokyo: San’ichi S hobo, 1981. K uroda Toshio. “Chusei n o m ibunsei to hisen k an n en .” B uraku m ondai kenkyu 33 (1984): 2 3 -5 7 . Kyoto Buraku Shi Kenkyu Jo. Chusei no minshii to geino. Kyoto: A unsha, 1986. M ihara Rekishi Kenkyukai. M ih a ra g u n shi. M ihara-gun: n .p ., 1985. M isum i H aru o . Geind shi no m inzoku teki kenkyu. Tokyo: T okyodo Shuppan, 1975. ---------- N ihon m inzoku gein o g a iro n . Tokyo: T okyodo Shuppan, 1974. ---------- Sasuraibito no geino shi. Tokyo: N H K Books, 1974. M iyam oto Tsuneichi. Setonaikai no kenkyu. Tokyo: Kadokawa S hoten, 1965. Miyata N o b o ru . Jisha engi. Tokyo: Iw anam i S hoten, 1975. ---------- O nna no reiryoku to ie no kam i. Tokyo: Jinbunshoin, 1983. M orita Yoshinori. Chiisei senm in to zo geino shi. Tokyo: Yusankaku, 1 9 8 1 . M oriya Takeshi. Geino togenzo. Kyoto: T ankosha, 1985. M une Torasuke. A w a ji nogake jd ru ri shibai. Tokyo: Sogeisha, 1986. N agahara Keiji. Chiisei seiritsuki no shakai to shisd. Tokyo: Yoshikawa K obunkan, 1977. ---------- N ihon no chusei shakai. Tokyo: Iw anam i S hoten, 1973. N agata Kokichi. Ikite iru ningyo shibai. Tokyo: Kinseisha, 1983. ---------- N ihon no ningyo shibai. Tokyo: Kinseisha, 1974. ---------- “N ingyo shibai n o kigen shiryo.” In Geind ronshU, edited by H o n d a Yasuji. Tokyo: Kinseisha, 1976. Nakajima Yoichiro. Bydki nihon shi. Tokyo: Y uzankaku, 1988.

Nakano Hatayoshi. H achim an shinko-shi no kenkyii. Tokyo: Yoshikawa Kobunkan, 1976. N am ihira Emiko. Kegare no kozd. Tokyo: Seidosha, 1992. ------------“Suishintai o Ebisu to shite m atsuru shinko: Sono imi to kaishaku.” M inzoku kenkyii 42, no. 4 (M arch 1978): 335-368. Niimi Kanji. A w a ji no ningyo shibai. Tokyo: Kadokawa Shoten, 1974. Niimi Tsuneyuki. “Um igam i to Ebisu shinko.” Matsuri, no. 46 (spring 1986): 2 7 -4 8 . ----------- “Awaji no ningyo joruri shiryo.” Dento engeki, no. 13: 48-58. Obayashi Taro. Nihon shinwa no kigen. Tokyo: Kadokawa Shoten, 1961. O gura M anabu. “N oto no kuni hyochakugami k o .” K okugakuin zasshi 15, no. 3 (1954): 3 0-43. Oka Masao. Nihon minzokugaku no rekishi to kadai. Tokyo: n.p., 1958. Okam ura Akio. A w a ji no bunka. Sum oto: Awaji Shin’yo Bijutsu Kan, 1960. Okiura Kazuteru. Nihon minshii bunka nogenryu: hisabetsu buraku no m inzoku to geino. Tokyo: Kaiho Shuppansha, 1984. O riguchi Shinobu. “ O kina.” N ihongeinoshi rokko, no. 2 (1944): 14-28. ______ “ Okina no hassei.” Origuchi Shinobu zenshu. Vol. 2. Tokyo: n.p., 1948. -___ .. “ Okina no kenkyii.” Geino 4, no. 6 (sum m er 1972): 310-366. ______ “Tokoyo oyobi m arebito.” M inzoku 4, no. I (1948): 1—62. Park Jon Yul. Kadozuke no kozd. Tokyo: Kobundo, 1989. Sakurada Katsunori. “N oto no hyochakugami no kenkyii.” Nihon minzokugaku I , no. 4 (1 9 5 4 ) :9 1 - 9 4 . Sakurai Tokutaro. “ Goryo shinko no hassei to tenkai.” In Goryd shinko, edited by Sakurai Tokutaro and Shibata M inoru. Tokyo: Yuzankaku, 1984. Sakurai Tokutaro and Shibata M inoru, eds. Goryd shinko. Tokyo: Yuzankaku, 1984. Sasahara Ryoji. “ Ki my δ na butai, bimyo na butai: M inzoku geino taikai to m inzoku geino kenkyusha” M inzoku geino kenkyii, no. 12 (1990): 13-17. Shinoda Ju n ’ichi and Ningyd Butai-shi Kenkyukai, eds. Ningyo jd ruri butai-shi. Tokyo: Yagi Shoten, 1992. Suzuka Chiyono. Shintd m inzoku geino no genruii. Tokyo, 1988. Takagi Keio, ed. Tosa no geino: Kdehi-ken no m inzoku geino. Kochi-shi: Kdchi-shi Bunka Shink5 Jigyodan, 1986. Takeda Seiichi. “ Nishinomiya kugutsu shi to Awaji ningyo shibai.” Shiritsu Nishinomiya kdkdgakko kiyo: ahydron“ no. 6 (1972): 1—11. Takeda Shin’ichi and Tokuda Toshiharu, eds. A w a ji junrei. Tokyo: M icho Shuppan, 1981. Takeuchi, T. “E bisu.” In Nihon shakai m inzoku jite n (Social and folklore diction­ ary o f Japan). Tokyo: N ihon M inzoku Kenkyukai, 1960. Tsunoda Ichiro. Ningyd geki no seiritsu n i kan suru kenkyii. Osaka: Kuroya Shuppan, 1964. Tsurum i Kazuko, ed. Yanagita K unio shu. Kindai nihon shiso taikei 14. Tokyo: Chikuma Shobo, 1975. Um azum e Masaru. A w a ji ningyo shibai. Fukura: Awaji ningyo za, 1975.

U m azum e M asaru. “Avvaji ningyo shibai.” In A w a ji ningyo kashira, edited by H y ogo K enritsu Rekishi H akubutsukan. Him eji: H yogo Kenritsu H akubutsukan, 1985. U m ehara Takeshi. A m a to tenno. 2 vols. Tokyo: Asahi Shinbunsha, 1991. W akamori T aro, ed. A w ajishim a no m inzoku. Tokyo: Yoshikawa K obunkan1 1974. W atatani Kiyoshi. B u n ra ku to aw aji ningyo za. Tokyo: N ara S h o b o 1 1956. Yamada Shijin. iiDdkumbo D enki: Awaji ningyo shibai shison d en sh o .” Ddshisha kokubungaku, no. 13 (M arch 1978): 9 1 -1 0 1 . Yamagami Izu m o . M iko no rekishi. Tokyo: Y uzankaku1 1984. ---------- Nanase harai n o g e n ry u . K odai b u n k a 2 2 , no. 5 (A ugust 1970): 1 2 4 -1 5 1 . Yamaji Κ δζδ. “K u g u tsu.” In ChUsei no m inshu to geino, edited by Yamaji Κ όζδ, M ori F um iko1 and Kyoto Buraku Shi Kenkyujo. Kyoto: A unsha1 1986. Yamaji K ozo1 M ori Fum iko1 and Kyoto Buraku Shi Kenkyukai1 eds. Chiisei no minshH to geino. Kyoto: A unsha1 1986. Yamaori T e tsu r6 . K a m i kara O kina e. Tokyo: Seiddsha1 1984. Yam am oto Shunosuke. Sado no ningyo shibai. N iigata-ken, S ado-gun, Shinmachi: Sado Kyodd Kenkyukai1 1976. Yanagita Kunio. “H ito o Kami ni M atsuru F u sh u .” In T anagita K unio zenshu. Vol. 7. Tokyo: Sogeisha1 1968. Yokoi Kiyoshi. Chusei m inshu to seikatsu bunka. Tokyo: Tokyo D aigaku Shuppan Kai1 1975. Yoshii Sadatoshi. Ebisu shinkd to sonofudo. Tokyo: K okusho Rikkokai, 1989. Yoshii Tar6. “ N ishinom iya no k u g u tsu .” M inzoku to rekishi I , no. I (Taisho 8): 2 8 -3 1 . W estern L a n g u a g e S ou rces

Akazawa1 Takeru. “ M aritim e A daptation o f Prehistoric H u n ter-G atherers and their T ransition to A griculture in Japan.” In The A fflu e n t Foragers: Pacific Coasts East a n d West, edited by S. Koyama and D. H . Thom as. Senri E th n o lo g ­ ical S tu d ie s ,n o .9 , p p .2 1 3 -2 6 0 . Osaka: N ational M useum o f E thnology, 1981. A m ino, Yoshihiko. “Som e Problem s C oncerning the H isto ry o f P opular Life in M edieval Japan.” A c ta A siatica 4 4 (1983): 7 7 -9 7 . Araki1 James T. The B allad D ra m a o f M edieval Japan. Berkeley: U niversity o f California Press, 1973. A ston, W. G ., trans. Nihongi: Chronicles o f Japan fro m the Earliest Times to A . D . 6 9 7 . R utland, Vt.: Charles E. T uttle, 1972. B aird1 BiL The A r t o f the Puppet Theater. N ew York: M acm illan, 1965. Barnes, Gina L. Protohistoric Tamato: Archaeology o f the First Japanese State. U n i­ versity o f M ichigan C en ter for Japanese Studies, M ichigan Papers in Japanese Studies, no. 17. A nn A rbor: U niversity o f M ichigan Press, 1988. Barnes, Gina L 1 and K. H u tte re r1 eds. Windows on the Javanese Past: Studies in Archaeology a n d Prehistory. Ann A rbor: U niversity o f M ichigan Press, 1986. Becker, A lton. “Text B uilding, Epistem ology and A esthetics in Japanese Shadow T h eatre.” In The Im a g in a tio n o f Reality: Essays in Southeast A sia n Coherence

Systems, ed ited by A. L. Becker and Aram A. Y engoyan, pp. 2 2 1 -2 4 2 . N o r­ w o o d , N . J.: Albex Publishing, 1979. B ender, Ross. “T h e H ach im an C u lt and th e D oky6 In c id e n t.” M on u m en ta N ipponica 2 4 , no. 2 (su m m er 1979); 1 2 5 -1 5 4 . ----------. “ M etam o rph o sis o f a D eity: T h e Im age o f H achim an in Y u m i T a w a ta Y M o n u m en ta N ipponica 3 3 , no. 2 (sum m er 1978); 1 6 5 -1 7 8 . B ethe, M onica, and K aren Brazeil. D ance in the N o Theater. Ithaca: C ornell U niversity Press, 1982. B cttelh cim , B runo. Freud a n d M a n ’s Soul. N ew York: K nopf, 1983. Blacker, C arm en. The C a ta lp a Bow: A Study o f Sham anic Practices in Japan. L o n d o n : G eorge A llen an d U n w in , 1975. Blau, H agen. Saru g aku u n d Shushi: Beitrage z u r A u sb ild u n g dram atischer Elem ente im Weltlichen u n d religiosen Volkstheatre der H e ia n -Z eit under besonderer Berucksicbtigung Seiner Sozialen G rundlagen. W iesbaden: O tto H arrasso w itz, 1966. Bock, Felicia. Classical L ea rn in g an d Taoist Practices in Early Japan w ith a Trans­ lation o f Books X V I a n d X X o f the Engi-Shiki. C e n te r for Asian Studies, O cca­ sional Papers, no. 17. T em pe: A rizona State University, 1985. B ou rd ieu , Pierre. O u tlin e o f a Theory o f Practice. T ranslated by Richard Nice. C am bridge: C am b rid g e U niversity Press, 1977. Bowers, F au b io n . Japanese Theater. N ew York: H erm itag e H o u se, 1952. C orbey, R aym ond, an d Joep L eersen, eds. A lterity, Identity, Im age: Selves a n d Others in Society a n d Scholarship. A tlanta an d A m sterdam : R o d o p i, 1991. C ornell, Jo h n B. “ F rom C aste P atro n to E n tre p ren e u r and Political Id eologue: T ransfo rm atio n in N in e te e n th and T w e n tie th C e n tu ry O u tcaste Leadership E lites.” In M odern Japanese Leadership: T ra d ition a n d Change, ed ited by Ber­ n ard S. Silberm an and H . D . H a ro o tu n ia n , pp. 5 1 -8 2 . Tuscon: U niversity o f A rizona Press, 1966. C raig, E. G o rd o n . “T h e A cto r an d th e L Jberm arionette.” In Total Theater: A C ritica l Anthology, ed ited by E. T. Kirby. N ew York: D u tto n , 1969. Davis, Fred. T e a m in g f o r Yesterday: A Sociology o f N ostalgia. N ew York: Free Press, 19 79. de C o p p e t, D aniel, ed. U nderstanding R itu a ls. L o ndon: R o u tled g e, 1992. de P o o rte r, Erika G erlinde, trans. MotoyoshPs Sarugaku D angi: A Description a n d Assessment w ith A n n o ta te d Translation. L eiden, 1983. DeV os, G eorge, an d H . W agatsum a, eds. J a p a n ’s Invisible Race: Caste in C u lture a n d Personality. Berkeley: U niversity o f C alifornia Press, 1966. D o rso n , E dw ard, ed. Studies in Japanese Folklore. B loom ington: Indiana U niver­ sity Press, 1963. D o ug las, Mary. P u rity a n d D anger: A n A nalysis o f the Concepts o f Pollution a n d Taboo. L o n d o n : R o u tled g e and Kegan Paul, 1966. D u m o n t, Louis. H om o H ierarchicus: A n Essay on the Caste System. C hicago: U n i­ versity o f C hicago Press, 1966. D u n n , C. J. The Early Japanese P uppet D ra m a . L o ndon: L uzac, 1966. D ykstra, Yoshiko K urata. “ Jizo, th e M o st M erciful: Tales from th e Jizo Bosatsu R eig en k i.” M o n u m en ta N ipponica 33 , no. 2 (su m m er 1978): 1 7 9 -2 0 1 .

Ebersolc5 Gary L. R itu a l Poetry a n d the Politics o f D eath in Early Japan. Prince­ ton: P rinceton University Press, 1989. Egam i, N am io. “ L ight on A ncient Japanese C ultural O rigins from H istorical Ar­ chaeology and L eg en d .” In Japanese Culture: Its Development a n d Characteris­ tics, edited by R o b ert J. Sm ith and Richard K. Beardsley, pp. 1 1 -1 6 . Viking F u n d Publications in A nthropology, W enner-G ren F ou n d atio n for A n th ro p o ­ logical Research. Chicago: Aldine, 1962. Eilberg-Schw artz, H ow ard. The Savape in Judaism : A n Anthropolopy o f Israelite Relipion a n d A n cie n t Judaism . B loom ington: Indiana University Press, 1990. Elam , Keir. The Semiotics o f Theatre a n d D ram a. L ondon: M ethuen, 1980. Eliade, Mircea. Shamanism: Archaic Techniques o f Ecstasy. Translated by W illard R. Trask. Bollingen Series, no. 76. Princeton: Princeton University Press, 1964. Elisseeff, Vadime. Archaeolopia M undi: Japan. Translated by James H o g arth . Geneva: N agel Publishers, 1973. Feher, M ichel, R am ona N addaff, and N adia Tazi. Frapments fo r a History o f the H u m a n Body. P art I. N ew York: Z o n e, 1989. F oucault, M ichel. Madness a n d C ivilization: A History o f Insanity in the A pe o f Reason. N ew York: P an th eo n Books, 1965. Fustel de Coulanges. The A n c ie n t City: A Studyon the Relipion, Laws, a n d In stitu ­ tions o f Greece a n d Rome. Translated by W illard Small. G arden City, N . Y.: D oubleday, 1956. G arvin, Paul, ed. A Prapue School R eader on Esthetics, Literary Structure a n d Style. W ashington: G eorge W ashington University Press, 1964. G eertz, Clifford. “M aking Experiences, A u th o rin g Selves.” In The Anthropolopy o f Experience, edited by V ictor T urn er and Edw ard B runer, pp. 341—380. C ham ­ pagne: University o f Illinois Press, 1986. G oodw in, Janet R. “Alms for Kasagi T em ple.” Jo u rnal o f A sian Studies 46, no. 4 (N ovem ber 1987): 8 2 7 -8 4 1 . G rappard, Alan. “L otus in the M o u n tain , M o u n tain in the Lotus: R okupo Kaizan N im m on D aibosatsu H o n g i.” M onum enta N ipponicaAA, no. I (spring 1986): 2 1 -5 0 . H aro o tu n ian , H . D . Thinps Seen a n d Unseen: Discourse a n d Ideolopy in Tokupawa N ativism . Chicago: University o f Chicago Press, 1988. H arris, William Wayne. Population, Disease a n d L and in Early Japan: 645-900. E astA sian Papers. C am bridge: H arvard U niversity Press, 1985. H arrison, Bernard. Form a n d Content. N ew York: H arp er and Row, 1973. H arrison, John A., trans. “ N ew L ight on Early and M edieval Japanese H isto ­ riography.” U niversity o f Florida M onographs, Social Sciences, no. 4 (fall 1959). H avens, N orm an, trans. M atsuri: Festival a n d R ite in Japanese Life. In stitute for Japanese C ulture and Classics, C ontem porary Papers on Japanese Religion, no. I. Tokyo: K okugakuin University, 1988. H erb ert, Jean. D ieux et sectes populaires d u Japon. Paris: Albin M ichel, 1967. ---------- Les dieux n a tio n a u x d u Japon. Paris: AJbin M ichel, 1969. H off, Frank. Sonp, Dance, Storytellinp: Aspects o f the Perform inp A rts in Japan. Cornell University East Asia Papers, no. 15. Ithaca: C ornell University Press, 1978.

H oltom , Daniel C. “The M eaning o f Kami.” M onum enta Nipponica 3 (1940): 2 - 2 7 ,3 2 -5 3 . ---------- The N ational Faith o f Japan: A Study in Modern Shinto. 1938; reprint, New York: Paragon Books, 1965. H ori, Ichiro. Folk Religion in Japan: C ontinuity and Change. Edited by Joseph M. Kitagawa and Allen Miller. Chicago: University o f Chicago Press, 1968. ---------- “ Mysterious Visitors from the Harvest to the New Year.” In Studies in Japanese Folklore, edited by Edward D orson, pp. 7 6 -1 0 3 . B loom ington: Indi­ ana State University Press, 1963. Inoura, Yoshinobu, and Toshio Kawatake. The Traditional Theatre o f Japan. Tokyo: The Japan Foundation, 1981. Ishida, Eiichiro. “N ature and the Problem o f Japanese Cultural Origins.” In Jap­ anese Culture: Its Development and Characteristics, eds. R obert J. Smith and Richard K. Beardsley. Viking Fund Publications in Anthropology, WennerGren F oundation for Anthropological Research. Chicago: Aldine Publishers, 1962. Ivy, Marilyn. Discourses o f the Vanishing in Contemporary Japan. Ph.D . diss. Cornell University, 1988. ---------- Discourses o f the Vanishing: Modernity, Phantasm, Japan. Chicago: University o f Chicago Press, 1995. Jameson, Fredric. “Post-m odernism , or the C ultural Logic o f Late Capitalism.” New Left Review, no. 146 (S ep t./O ct. 1984): 53-92. Janetta, Ann Bowman. Epidemics and M ortality in Early Modern Japan. Prince­ ton: Princeton University Press, 19. Kamata, Hisako. “D aughters o f the Gods: Shaman Priestesses in Japan and O ki­ nawa. ” In Folk Cultures o f Japan a n d East Asia. M onum enta Nipponica M ono­ graphs, no. 25. Tokyo: Sophia University Press, 1966. Kapferer, Bruce. “ Performance and the Structure o f M eaning and Experience.” In The Anthropology o f Experience, edited by Victor T urner and Edward Bruner. Champagne: University o f Illinois Press, 1986. Kato, G enchi. “A Study o f the D evelopm ent o f Religious Ideas am ong the Japa­ nese People as Illustrated by Japanese Phallicism.” Transactions o f the Asiatic Society o f Japan, 2d ser., supp. to vol. I (D ecem ber 1924): 5-70. Kawatake, Shigetoshi. Development o f the Japanese Theatre A r t. Tokyo: Kokusai Bunka Shinkosai, 1935. Keene, D onald. Bunraku: The A r t o f the Japanese Puppet Theatre. Tokyo: Kodansha International, 1973. Kidder, J. E., Jr. Early Japanese A r t: The Great Tombs and Treasures. Princeton: Von N ostrand, 1964. ---------- Japan Before Buddhism. London: Praeger and Sons, 1966. Kitagawa, J. M. On Understanding Japanese Religion. Princeton: Princeton U n i­ versity Press, 1987. ------------Religion in Japanese History. New York: Columbia University Press, 1966. Koschmann, V ictor J., Oiwa Keibo, and Yamashita Shinji, eds. International Per­ spectives on Tanagita K unio a n d Japanese Folklore Studies. Cornell University East Asia Papers, no. 37. Ithaca: Cornell University Press, 1985.

Kristeva, Julia. Powers o f Horror: A n Essay on Abjection. Translated by Leon S. R oudiez. N ew York: C olum bia University Press, 1982. LaFleur, William. “ H u n g ry G hosts and H u n g ry P eople.” In Fragments fo r a H is­ tory o f the H u m a n Body, p art I , edited by M ichel Feher, pp. 2 7 0 -3 0 1 . New York: Z one, 1989. ---------- The K a rm a o f Words: Buddhism a n d the Literary A rts in M edieval Japan. Berkeley: U niversity o f California Press, 1983. ----------. L iquid Life: Abortion a nd Buddhism in Contemporary Japan. Princeton: P rinceton U niversity Press, 1992. Law, Jane M arie. “ O f Plagues and Puppets: O n the Significance o f the N am e H yakudayu in Japanese R eligions.” Transactions o f the A siatic Society o f Japan, 4 th ser., 8 (1993): 1 0 8 -1 3 2 . ----------. “T h e P uppet as Body S ubstitute: Ningyd in the Japanese Shiki Sanbaso Perform ance.” In Religious Reflections on the H u m a n Body, edited by Jane M arie Law. B loom ington: Indiana University Press, 1995. ---------- “ Religious A uthority and Ritual Puppetry: T he Case o f Ookumbd D enki.” M onum enta Nipponica 27 , no. I (spring 1992): 7 8 -9 7 . ---------- “V iolence, Ritual R eenactm ent, and Ideology: T he H ojo-e (Rite for R e­ lease o f Living Beings) o f th e U sa H achim an Shrine in Japan.” History o f R e li­ gions 33, no. 4 (M ay 1994): 3 2 5 -3 5 7 . ----------, ed. Religious Reflections on the H u m a n Body. B loom ington: Indiana U n i­ versity Press. 1995. Ledyard, G ari. “ G alloping A long with the H orseriders: L ooking for the Founders o f Japan.” Journal o f Japanese Studies I , no. 2 (1975): 2 1 7 -2 5 4 . L ong, Charles H . Significations: Signs, Symbols a n d lm a g esin the Interpretation o f Religion. Philadelphia: Fortress Press, 1986. M arra, M ichel. The Aesthetics o f Discontent: Politics a n d Reclusion in M edieval Japanese Literature. H on o lu lu : University o f Hawaii Press, 1991. ---------- Representations o f Power: The Literary Politics o f M edieval Japan. H o n ­ olulu: U niversity o f Hawaii Press, 1994. M aruyam a, M asao. Studies in the Intellectual H istory o f Tokugawa Japan. T rans­ lated by Mikiso H ane. Princeton: P rinceton U niversity Press, 1989. Mayer, Fanny H agin. “T he Yanagita K unio A pproach to Japanese Folklore Studies.” The Transactions o f the A siatic Society o f Japan, 3d ser., 13 (D ecem ber 1976): 1 2 9 -1 4 3 . M cN eill, William. Plagues a n d Peoples. G arden City, N.Y.: D oubleday, 1976. M encher, Joan P., ed. Social Anthropology o f Peasantry. Bombay: Somaiya, 1983. Miki, Fum io. H aniw a. Tokyo: W eatherhill, S hibundo, 1974. Miller, Richard J. Ja p a n ’s First Bureaucracy: A Study o f Eighth C entury Govern­ ment. China-Japan Program . Ithaca: C ornell U niversity Press, 1978. Miwa, Kimitada. “T ow ard a Rediscovery o f Localism: Can the Yanagita School o f Folklore Studies O vercom e Japan’s M odern Ills?” Japan Q uarterly 2 3 , no I (January-M arch 1976): 3 8 -4 9 . M orse, Edw ard A. “Yanagita K unio and the M odern Japanese C onsciousness.” In In tern a tio n a l Perspectives on Yanagita K u n io an d Japanese Folklore Studies, edited by J. V ictor K oschm ann, Oiwa Keibo, and Yamashita Shinji, pp. 1 1 -2 8 . C ornell E astA sia Papers, no. 37. Ithaca: C ornell University Press, 1985.

N agahara, Kciji. “T h e M edieval O rigins o f E ta - H in in .” Jo u rn a l o f Japanese S tu d ­ ies 5, no. 2 (sum m er 1979): 3 8 5 -4 0 4 . N akam ura, H ajim e. Ways o f T h in k in g o f Eastern Peoples: In d ia ■—C h in a —Japan. H o n o lu lu : U niversity o f H aw aii Press, 1978. N akane, C hie. Japanese Society. H a rm o n d s w o rth : P enguin Books, 1973. N arim itsu , M atsudaira. “T h e C o n c e p t o f Tam ashii in Jap an .” In Studies in Ja p a ­ nese Folklore, e d ite d by E dw ard D o rso n . B loom ington: In d ian a U niversity Press, 1963. N eary, Ian. P olitical Protest a n d Social C ontrol in P re-W ar Japan: The O rigins o f B u ra k u Liberation. A tltan tic H ig h lan d s, N .J.: H u m anities Press In te rn atio n a l, 1989. N iculescu, M argareta, ed. The Puppet Theatre o f the M odern World. T ransated by Ew ald O sers an d E lisabeth Strick. B oston: Plays, In c ., 1967. N inom iya, Shigeaki. “An In q u iry C o n cern in g th e O rig in , D ev elo p m en t, and P resen t S itu atio n o f th e E ta in R elation to th e H isto ry o f Social Classes in Jap an .” Transactions o f the A sia tic Society o f Japan, 2d ser., 10 (D ecem b er 1933): 4 7 -1 5 2 . O hnuki-T ierney, E m iko. Illness a n d C u ltu r e in C ontem porary Japan. C am bridge: C am bridge U niversity Press, 1984. -----------M onkey as M irror: Symbolic Transform ations in Japanese H istory a n d R itu a l. P rinceton: P rin ceto n U niversity Press, 1987. Parker, R o b ert. M iasm a: P ollution a n d P urification in E arly Greek Religion. O xford: O x fo rd U niversity Press, 1983. Philippi, D o n ald L. trans. K ojiki. Tokyo: U niversity o f Tokyo Press, 1968. ---------- K ojiki: Translated w ith a n In tro d u ctio n a n d Notes. P rinceton: P rinceton U niversity Press, 1969. -----------Norito: A T ranslation o f A n c ie n t Japanese R itu a l Prayers. P rinceton: P rin ceto n U niversity Press, 1990. P o nso n b y -F an e, R ichard. The Vicissitudes o f Shinto. Kyoto: P onsonby-F ane M e­ m orial Society, 1963. ---------- . V isiting Fam ous Shrines in Japan. Tokyo: T h e P onsonby-F ane M em orial Society, 1964. R az, Jacob. A u d ien ce a n d Actors: A Study o f Their Interaction in the Japanese Theatre. Leiden: E. J. Brill, 1984. -----------“P o p u lar E n te rta in m e n t an d Politics: T h e G reat D engaku o f 1 0 9 6 .” M o n u m e n ta N ipponica 4 0 , no. 3 (a u tu m n 1 985): 2 8 3 -2 9 8 . Redfield, R o b ert. The L ittle C om m unity: Viewpoints fo r the Study o f a H u m a n Whole. Ithaca: C o rn ell U niversity Press, 1955. ----------. Tepoztlan: A M exican Village. C hicago: U niversity o f C hicago Press, 1930. R einiger, L o tte. Shadow Theatre a n d Shadow Films. L o n d o n : B. T. B atsfbrd, 1970. R eischauer, Jean , an d R o b e rt Karl. E arly Japanese H istory (c. 40 B .C . — a . d . 1167). P rinceton: P rin ceto n U niversity Press, 1937. R ickm an, H .P ., ed. an d trans. Dilthey: Selected W ritings. C am bridge: C am bridge U niversity Press, 1976. R icoeur, Paul. H erm eneutics a n d the H u m a n Sciences. E d ited and translated by

John B. T hom pson. E ditions de la M aison des Sciences de PH om m e. C am ­ bridge: C am bridge University Press, 1981. R oterm u n d , H a rtm u t O . Hosog am i: ou In petite verole aisement: m ateriaux pour Petude des epidemies dans Ie Japon d esX V IIIe, X I X e siecles. Paris: M ai sonneuve et Larose, 1991. Sahara, M akoto. “T h e Yayoi C u ltu re.” In R ecent Archaeological Findings in Japan, pp. 3 7 -5 4 . Tokyo: C en ter for East Asian C ultural Studies, 1987. Said, Edw ard. Orientalism . N ew York: P antheon Books, 1978. Sansom , Sir G eorge. A History o f Japan. Stanford: Stanford U niversity Press, 1958. Schaum ann, W erner. “ Leistet die M ythe vom V erlorenen Angelhaken (UmisachiTamasachi no Shinw a) E inen Beitrage zur K alrung des Problem s der H erkunft der Japanischen K ultur?” In Beitrage z u r Japanischen Ethnogenese: 100 Jahre nach H einrich von Siehold, pp. 129—147. B onner Zeitschrift fur Japanologie, vol. 2. B onn, 1980. Schechner, Richard. Between Theater a n d Anthropology. Philadelphia: U niversity o f Pennsylvania Press, 1985. Scott, A. C. The Puppet Theater o f Japan. R utland, Vt.: Charles E. T uttle, 1963. Segawa, Kiyoko. “M enstrual T aboos Im posed upon W om en.” In Studies in Ja p a ­ nese Folklore, edited by Edw ard D orson, pp. 2 3 9 -2 5 0 . B loom ington: Indiana University Press, 1963. Sim m en, Rene. The World o f Puppets. N ew York: T hom as Y. Crow ell, 1972. Sm ith, Jonathan Z. M ap is N o t Territory: Studies in the H istory o f Religions. Leiden: E. J. Brill, 1978. Sm ith, R o b ert J. Ancestor Worship in Contemporary Japan. Stanford: Stanford University Press, 1974. Sm ith, R o b ert J., and Richard K. Beardsley, eds. Japanese Culture: Its Develop­ m ent a n d Characteristics. V iking F und Publications in A nthropology, W ennerGren F o undation for A nthropological Research. Chicago: A ldine, 1964. Sonoda, M inoru. “T he Religious Situation in Japan in Relation to S h in to .” A cta A s ia tic a S l (1987): 1 -2 1 . Swanberg-Law, Jane M arie. “Puppets o f the R oad: Ritual Perform ance in Japa­ nese Folk R eligion.” P h.D . diss., University o f Chicago, 1990. Takayanagi, S h u n ’ichi. “Yanagita K unio Survey Review.” M onum enta Nipponica 2 4 , no. 3 (autum n 1974): 3 2 9 -3 3 5 . Tam biah, Stanley. “A Perform ative A pproach to R itual.” In Proceedings o f the British Academ y, vol. 65. L ondon: O xford U niversity Press, 1979. Tanaka, M igaku. “T h e Early H istorical P eriods.” R ecent Archaeological Discover­ ies in Japan. Tokyo: C enter for East Asian C ultural Studies, 1987. T heunissen, M ichael. The Other: Studies in the Social Ontology o f Husserl, H e i­ degger, Sartre a n d Buber. Translated by C. M acann. C am bridge: M IT Press, 1984. Todorov, Tzvetan. Literature an d Its Theorists: A Personal View o f Twentieth C en­ tury Criticism. T ranslated by C atherine Porter. Ithaca: C ornell University Press, 1988. Tsuda, Sokichi. “T h e Idea o f Kami in A ncient Japanese Classics.” T young Pao. Vol. 7, bks. 4 -5 . Leiden: 1966.

T su d e, H iro sh i. “T h e K ofun P e rio d .” In R ecent Archaeological Discoveries in Ja p a n . Tokyo: C e n te r for E ast Asian C ultural Studies, 1987. T su n o d a, R yusaku, and L. C a rrin g to n G o o d rich . Japan in the Chinese D ynastic Histories. P erkinsA siatic M o n o g rap h s, n o . 2. S o u th Pasadena, Calif.: P. D. and lo n e Perkins, 1951. T su n o d a , R yusaku, W. T h e o d o re de Bary, and D o nald K eene. Sources o f Japanese T radition. N ew York: C olu m b ia U niversity Press, 1958. T u rn e r, V ictor W. The R itu a l Process; Structure a n d A n ti-S tru c tu re. Ithaca: C ornell U niversity Press, 1969. T u rn er, V icto r W ., an d E dw ard M . B ru n er, eds. The Anthropology o f Experience. C ham pagne: U niversity o f Illinois Press, 1986. W akita, H a ru k o , and Susan B. H anley. “ D im ensions in D ev elo p m en t o f Cities in F ifteen th and S ixteenth C en tu ry Jap an .” In Ja p a n Before Tokugawa, ed ite d by Jo h n W. H all, N agahara Keiji, and Yam am ura K 5zo, pp. 2 9 5 -3 2 6 . P rinceton: P rin ceto n U niversity Press, 1981. W alton, M ichael J, C ra ig on Theatre. L o n d o n : M e th u en , 1983. W illiams, R aym ond. The C ou n try a n d the City. N ew York: O xford U niversity Press, 1973. Wu H u n g . “A Sanpan Shan C h a rio t O rn a m e n t and th e X iangrui D esign in W est­ ern A rt.” Archives o f A sia n A r t 3 7 (1 9 8 4 ): 3 8 -5 9 . Y am am oto, Yoshiko. The N am ahage: A Festival in the Northeast o f Japan. P hila­ delphia: In stitu te for th e Study o f H u m a n Issues, 1978. Yamashita, Shinji. “ Ritual an d U n conscious T rad itio n : A N o te o n Yanagita K u n io ’s A b o u t O u r Ancestors.” In In te rn a tio n a l Perspectives on T a n agita K u n io a n d Japanese Folklore Studies, e d ite d by J. V ictor K oschm ann, O iw a K eibo, and Yamashita Shinji, pp. 5 5 -6 4 . C ornell E ast Asia Series, no. 37. Ithaca: C ornell U niversity Press, 1985. Yoneyam a, T oshinao. “Y anagita and H is W orks.” In In te rn a tio n a l Perspectives on T a n a g ita K u n io a n d Japanese Folklore Studies, edited by J. V ictor K oschm ann, O iw a K eibo, and Yam ashita Shinji, pp. 2 9 -5 2 . East Asia Series, no. 37. Ithaca: C ornell U niversity Press, 1985. Yoshida, T eigo. “T h e St ran g er as G od: T h e Place o f th e O u tsid e rin Japanese Folk R elig io n ,” Ethnology 2 0 , no . 2 (1 9 6 4 ): 8 7 -9 9 . Y oung, Jo h n . The Location o fT a m a ta i: A Case Study in Japanese Historiography, 720-1945. Jo hns H o p k in s U niversity S tu d ie sin H istorical an d Political Science, series 75, no. 2. B altim ore: Jo h n s H o p k in s U niversity Press, 1958.

Index

agemaki, 187 Agency for Cultural Affairs, 2 0 6 Ai Suzue, 149 Ainu, 4 3 akafujo {red pollution). See menstruation All Japan Puppetry Summit, 11, 2 4 1 , 2 4 4 . See also zenkoku ningyo shibai samitto All Japan Regional Performing Arts Convention, 2 0 4 , 2 1 0 , 2 4 2 ^ 3 alterity, 12, 53, 2 4 6 ; and identity, 54; and oppression, 54 Amagasaki, 135, 153, 154 amagatsu (type o f effigy), 35, 36, 38 amajjoi (ra\n making rite), 175 Amaterasu, 111, 123, 1 2 5 , 181 amazoku (ama), 62, 138, 1 4 0 , 1 4 1 , 2 5 6 Ame-no-Uzumc-no-Mikoto, 4 3 Amida no Munewari (early puppet play), 134 amulets, 132 appeasement, ritual, 18, 32, 4 4 , 92ff., 130, 133,136,140,150 ara-ebisu, 121 ara-mi-tama (rough spirit aspect o f deities), 111, 121 arai Kannon (Avalokitesvara), 39 architecture, Japanese, 78 Aristotle, 231 ashibyoshi, 177 audience status, 2 4 0 avant garde theatre, 2 6 aversion, 6 Awa, 2 2 1 Awaji, 3, 6, 8 6 , 87, 156; amazoku hypothesis, 138—41; early tradition, 10; historical archives, 9, 2 5 6 ; myth tradition, 115; preservation of, 2 0 6 ; puppetry tradition, 3, 6, 110, 134, 1 3 6 , 137, 138, 1 4 6 ; uniqueness of, 138; Usa hypothesis, 141. See also Awaji puppetry Awaji Awa Ningyo Za Kyiiki Zassan, 154 Awaji Mihara-gun Sanjo mura, 153, 159 Awaji-gusa^ 151, 163 Awaji puppetry: Ebisu rite, 134; festival (ningyd matsuri), 2 0 1 ; historical beginnings, 6; bozonkai (preservation society),

2 1 3 ; as movement from shrine work, 148; narrative origins, 137ff.; tradition formation in, 4, 7, 147; tradition revival in, 4 Awaji Ningyo Joruri Kan, 5, 7 , 8, 13, 157, 167, 1 9 9 , 204ff., 2 3 0 , 2 3 7 , 2 4 4 , 248ff., 259 Awaji no Kuni Meisho Zue, 151 Awaji Ningyo Shibai No Xurai, 2 1 3 Awaji Sanjo-mura, 156 Baird, Bil, 2 1 , 22 Ballet Folklorico, in Mexico, 2 3 1 Band5 Sentaro, 2 6 5 Banto Pilgrimage route, 38 Banzairaku, 68 barefoot, as social rank, 1 0 6 Bateson, Gregory, 2 3 1 bathing. See cleansing Baumann, Gerd, 85 Bharatana.tya.rn (classical Indian dance), 231 Bloch, Maurice, 181 Bock, Felicia, 102, 133 body, as symbol for society, 56, 6 6 Bogatyrev, Petr, 2 4 6 boundary situations, 71 Budapest Puppet Theatre, 2, 7 Buddhism, and dreams, 119. See also Chinese Buddhism Bunraku, 4, 17, 18, 2 7 , 4 4 , 1 7 0 , 2 0 9 , 2 2 0 , 228,243,251 burakumin, 67, 68, 2 2 7 Byodoin, 7 3 chants, sacred (tonaegoto), 7 3 , 180ff. Chikamatsu Monzaemon, 23 childbirth, 6 1 , 65 child protection, through effigies, 131 China: artists, 103; Buddhism, 9 4 ; ethnographic poetry, 9 9 ; putative origins of puppet theatre, 94ff.; shadow puppet theatre, 2 7 , 2 8 ; transformation artists, 98 chinkonsai. See spirit pacification Chiri Mashiho, 4 4 Chiribukuro (Bag o f Dust), 105, 106

314 Chuyuki, 7 3 , 100 cleansing, 62. See ai.ro purification colonialism, and nostalgia, 226 containment, ritual of, 1 1 2 - 2 0 cosmology, 59, 71 The Country and the City, 216 Craig, E. Gordon, 19, 2 9 , 30 Cultural Properties Protection Law, 2 1 0 , 213 Curtiss, Edward, 2 3 0 daiyaku, 202 Danzaemon, 162 death: reversing, 2 8 ; rites of, 6 0 - 6 2 , 65 Davis, Fred, 2 1 5 , 2 1 6 , 2 1 8 , 2 2 6 defecation, 63 defilement, 6 6 deities. See kami deko-mnivashi, 51. See also kugutsu dekozamma, 202 Dengaku, 2 4 2 dentogeittd biimu, 241 Denver, 1 0 4 Dewa Province, 37 dhalnng, in Java, 23 disaster victims, spirits of, 117 Diet, Japanese, 2 1 0 discrimination, 6 6 , 7 3 , 8 4 , 169 divination, 7 diviner, 73; Taoist, 34, 73 dogu-gaeshi, 1 2 - 1 3 , 1 9 9 , 2 0 0 , 2 0 6 Dokun. See Dokumbo Dokumbo, 9, 131, 143, 144, 146, 154, 160, 177, 2 3 7 dokumbo-mawttshi. See ritual puppetry Dokumbo-Denki (text), 12, 13, 1 3 7 , 1 4 9 6 3 ; calligraphy, 157; extant copies, 157 domination, social, through imagery, 230ff. iidsojin, 130 Douglas, Mary, 55, 66 drifted deity tradition, 1 1 6 - 2 0 , 133. See also hydchakugami Dumont, Louis, 6 7 Dunn, C. }., 106 Durkheim, 85 Ebersole, Gary, 182 Ebisu: deity, 69, 94, 110, 1 1 1 - 3 6 , 143, 144, 160, 189ff., 232ff.; containment o f dangerous potential, 112ff., 165; and fish, 113, 168, 188ff.; foreignness, 115;

INDEX Kotoshiro-nushi no kami, 124; and Leech Child, 6 4 , 1 1 1 , 1 1 5 , 1 2 2 - 2 4 , 1 5 8 ; meanings of, 120; modern ritual performance, 2 5 5 ; non-deity, 124; patron of puppets, 9, 9 0 ; physical depiction, 113; and pollution, 112, 1 2 4 , 165; the puppet, 3, 4, 4 4 , 4 7 ; reenactment and revival, 2 3 0 ; repressed qualities, 1 2 8 ; rites, 9 4 ; vs. Sanbaso performance, 1 6 5 , 168; sea deity, 140; shrines, 1 4 3 , 145; visiting deity, 1 2 1 ; worship, 1 5 6 Ebisu-kaki, 12, 8 6 , 1 0 9 , 1 1 1 , 1 3 4 , 1 3 7 , 151, 157, 189ff., 2 3 2 Ebisu-mai, 134, 135, 188ff., 2 3 2 , 253ff.; style differences, 2 5 5 Ebisu-mawashi, 51, 109, 188ff. See also kugutsu Edo, 162 Edo period, 151, 154, 2 1 7 effigies, 17, 18, 33, 34, 1 1 8 , 130; as body substitute, 34, 9 4 ; clay, 33; origin o f in puppetry, 133; to ward off disease, 132ff. ekibyogami (epidemic spirits), 1 2 9 , 133, 134 Elam, Keir, 2 4 6 Eliade, Mircea, 5 3 - 5 4 , 71 ema (shrine placards), 38 engi (shrine origin narratives), 1 1 5 , 118, 157, 159 Engi-shiki, 1 0 2 , 1 2 7 , 133 Enlightenment, European, 55 ennichi, 191 epidemic, Japanese, 1 2 9 . See also ekibyogami excreta, divine, 6 4 exorcism, 74 exoticism, 2 0 8 , 2 2 1 , 2 6 3 externalized spirit, 34; as divine, 39 fare, as inescapable, 2 5 , 2 6 felicity, 189 fertility ritual, 4 2 - 4 4 fetuses, aborted or miscarried, 37, 38 finitude, 2 6 fire, 6 2 floating buddhas. See hyocheiku^ami folk art: and culture, 13, 204ff.; performing, 5, 7, 204ff. Folk Performing Arts Convention, 2 4 2 , 243 folk religion, 13, 14

foreign o rig in th eo ry , 9 5 , 9 9 foreigner, role of, 2 5 6 F o u c au lt, 55; N aren sch iffen , 71 fra g m e n ta tio n o f h isto ry , 9 0 F u d 5 S atoshi, 2 1 3 , 2 5 3 , 2 5 6 Fujikaw a Yu, 129 Fujiw ara n o M u n e ta d a , 100 fu k k a ts u (revival), 2 0 6 fu k k d , 2 0 6 f u k u j i n , 12 F u k u ra , village of, 5 ,1 3 , 7 6 , 1 4 2 , 1 5 7 ,1 6 7 , 2 0 1 , 2 3 0 , 2 4 7 , 2 5 5 ; fishing collective, 1 9 2 ,2 3 2 funeral practice, 117 f u r i f u r i Sanbaso, 167. See also Sam abaso g a b u (trick p u p p e t h e a d ), 2 3 , 2 4 G agaku, 186 g a ik o n (ex tern alized sp irit), 34 g a ira ig o (w ords o f foreig n o rig in ), 2 1 4 gate c h a n te rs (shdmon hoshi), 73 g ateb a ll, 9 , 1 4 6 , 2 4 0 gatew ays, 7 8 , 7 9 , 108 gays a n d lesbians, 54 g en ital flashing, 4 3 , 4 4 G e n jo ra k u , 185 g e n k a n (as ritu al space), 1 0 8 , 109 G enroku k i A w a ji A y a tsu n -sh ib a i no Chijo K o k i., 155 g eo g rap h y (sacred): m arg in a liz atio n , 7 0 ; o f m e an in g , 111; o f o th e rn e ss, 7; p e rfo r­ m ance, 70; sym bolic, 10, 5 2 , 71 G io n M a tsu ri, 1 2 9 , 130 G o -H a n a z o n o , 106 gbhei (m arkers o f ritu al space), 183 goryd-shin (m alev o len t sp irits), 117 g o s h in ta i{spirit vessel), 1 2 ,4 8 , l l l f f . , 1 1 6 , 191. See also sh in ta i G o y o ze i, E m p ero r, 154 g ra n d narrativ e, failure o f, 9 5 ff., 9 8 , 1 2 2 , 155 gravedig gers, 70 g u ard ia n sh ip , 33 G u n ji M asakiyo, 189 H a c h im a n , 8 , 9 , 9 1 ff., 1 4 0 , 1 4 3 , 1 7 5 , 1 9 2 , 198; as b o d h isattv a, 9 2 , 9 3 ; w a ka m iya , 143 H ac h isu g a family, 1 6 1 -6 2 h aiden (w o rship hall), 1 4 3 , 2 3 3 hako-m aw ashi, 4 5 , 5 1 , 108. See also ku g u tsu

H a k u sa i, 2 3 H am am a tsu U ta k u n i, 131 h a n a g u sa (flow er v e n d o r), 199 H a n d a Y a su o , 141 H an iw a figurines, 3 3 , 35 harae-m ono (ritu al p u rifier), 73 hare, 6 0 , 6 1 . See also p urification H a ro o tu n ia n , H . D ., 15 H ase K a n n o n , 3 8 h a ya g a w a ri, 199ff. H ayashiya T a tsu sa b u ro , 7 2 , 9 8 H a y a to , 9 0 - 9 3 , 211 h ea lin g rites, 3 2 , 74 H e a rt S u tra, 76 h e g em o n y , 1 2 5 , 2 1 4 ff. H eian p e rio d , 12, 3 5 , 6 8 , 7 1 , 7 9 , 8 0 , 8 1 , 8 9 , 9 0 , 9 4 ff„ 1 0 0 , 1 0 2 , 1 0 6 , 182 H eik e w arriors, 1 4 7 H ik id a Awaji T ayii, 155 H in d u tra d itio n s, 231 h in in (n o n -h u m a n ), 9 0 , 2 4 6 H iro d a (H iro ta ), 1 1 0 , 1 1 1 , 1 2 3 , 1 2 4 , 130 h iru ko , 1 1 1 , 1 2 2 , 1 3 1 , 1 3 2 , 1 5 8 . See also E b isu ; L eech C h ild h iru m e , 1 1 1 , 122 hisabetsu b u ra ku . See o th e rn e ss; o u tsid ers HisadayQ tro u p e , 167 historical scholarship: b lin d h isto ricism , 1 0 0 ; erro rs in , 9 9 , 1 0 0 ; p ro b le m s in , 9 5 - 9 9 ; u b iq u ity , 134 h ito g a ta , 3 2 , 3 3 , 3 4 , 3 5 , 9 4 H o fe r, Jo h a n n e s, 2 1 5 H 5jo-e, 9 2 , 9 3 , 211 H o k k a id o , 13, 39 boko. See a m a g a tsu h o m e , p u rifica tio n o f, 175 hom esick n ess, 21 5 H o n d a Yasuji, 2 0 9 honden, 2 3 3 h o n o -sa n b a so . See sanbaso honsha (m a jo r sh rin e), 145 H o n s h u (Iw a te p re fe c tu re ), 1 0 , 1 4 7 , 162 hopa ra ta (A in u te rm ), 4 4 H o ri Ic h iro , 1 4 , 15 hoshi (p riest), 8 0 , 81 hozonkai (p reserv atio n so cieties), 2 1 2 , 243 H y ak u d ay u , 9 , 1 1 1 , 1 3 0 , 1 3 1 , 1 3 2 , 1 3 4 , 13 6 , 1 4 3 , 1 4 4 , 1 4 5 , 1 5 1 -6 3 , 1 6 4 , 2 3 7 ; as in te n tio n a lly lo st sto ry , 2 2 7 hydchakugam i, 1 1 7 -1 9 hyohaku no g e l, 77

316 Ichi (Sanjo, Awaji), 137, 138, 149 Ichimura Rokunojo, 142 icon creation, in revivalism, 208ff. identity construction, 5 4 - 5 5 , 204ff., 215 Ikite iru ningyd shibai, 157 imagination, theatrical, 2 5 , 2 6 Imayo, 97 immersion, in cleansing, 62 Imperial Palace, 7 8 Inai Haruichi, 2 5 3 - 5 6 infanticide, 37 Inland Sea, 7 6 , 9 4 , 1 3 8 - 4 0 , 2 0 6 International Puppetry Festival (lida-Shi), 9, 171, 2 5 8 interview techniques, 168 Iroha Jiruisho, 120 Isc, 111, 146 Ise ke bisho tanjo no ki, 35 Israelites, 65 itinerancy. Sec ku^gutsu-maivashi; outsiders itinerant performance, 7, 11, 13, 16, 5 1 , 59, 64, 6 6 - 7 0 , 7 3 , 7 8 , 81, 8 4 , 100, 103, 105ff., 112ff., 145, 161, 2 4 1 , 2 4 6 ; Awaji and roots of, 147ff.; campaign to end, 8 6 , 227ff.; as divine distribution system, 83; incorporated into staged performance, 2 4 9 ; non-ecclesiastical sacred specialists, 149; purification role, 149; rise of, 1 3 7 , 138; shame concerning, 2 2 8 ; spirit pacification, 149; and stage performance revival, 2 0 8 Ivy, Marilyn, 2 2 9 Iwakusubuna, 130, 160 Iwashimizu Hachiman shrine, 113, 146 Izanagi and Izanami, 6 4 , 1 2 2 - 2 6 , 146, 158 Izumo shrine, 70 Jameson, Fredric, 2 1 7 - 1 9 Japan Sea, 75 Japan: errors o f omission, 134; grand narrative, failure o f 9 5 f f „ 9 8 , 9 9 ; history of, unified, 1 5 - 1 6 ; medieval, 5 8 - 5 9 , 6 1 - 6 2 , 6 9 , 7 1 , 7 2 , 80, 103ff., I l l , 137, 182, 187 Java, 2 3 Jews, 5 4 Jingu Kog5, 126 Jito, Empress, 81 jiuta, 180 jizo, ingestion of, 39

INDEX Jizo, Bosatsu (ksitigharba), 3 7 joruri (ballad), 12, 2 3 , 4 4 , 4 6 , 132, 134, 145, 1 5 0 , 157, 175, 176, 1 9 2 f f . , 2 2 0 , 2 2 3 , 2 4 3 ; vs. kadozuke, 2 1 9 ; tradition, 4, 7, 18, 2 3 , 4 2 , 111, 153 jiiyo mukei minzoku bunkazai, 2 0 6 kabuki, 7 0 , 170, 2 0 9 , 2 2 0 , 2 4 2 , 2 4 3 kadozuke ningyo, 1 6 4 , 2 6 2 kadozuke nogei, 77 kadozuke. performance tradition, 11, 5 2 , 7 7 , 7 8 , 8 2 - 8 8 , 109, 171, 1 8 8 , 1 9 1 f f . , 208ff., 2 4 1 , 2 4 5 , 2 5 6 , 2 6 3 ; vs. joruri, 2 1 9 , 2 2 5 ; political denial of ancient, 2 2 7 ; and radical negativity, 8 4 ; reenactment, 232ff.; reenactment activities, 2 3 0 ; revival (1990s), 2 2 9 ; and scholarly Zeitgeist, 87; seasonal nature, 8 5 ; stage performance, 2 4 8 ; vulnerability of oral tradition, 2 2 7 Kagura, 9, 171, 186, 1 8 7 , 2 4 2 kairaishi-ki, 9 0 , 9 6 , 9 8 - 1 0 0 , 103ff. kalpa (Vedic measurement o f time), 187 Kamakura, 38, 8 0 , 82, 9 0 , 100, 110 kami, 13, 15, 6 3 - 6 5 , 120, 142, 158, 159 kami asobi, 185 kami yobai, 189 kami-okuri. See Sanbaso kami-okuri mngyo, 172 kamidana, 4 9 , 116, 157, 192 Kanedayu, 156 kannushi, 6 8 , 1 5 3 , 192 kamkuri, 2 5 1 Kasai family, 131 kashira, 1 8 9 , 2 0 1 , 2 5 0 , 2 5 4 ; carving, 8 , 2 4 , 4 6 , 4 7 , 2 5 1 , 2 5 8 ; repair, 2 0 1 Kasuga, 1 4 6 Kato Genchi, 115 ka-tashiro. See amajjatsu kawara. kojiki, 70 kawara mono. See riverbeds kmvata hyakusho, 1 6 6 kawaya (riverbed people), 70 ke (profane), 6 0 Keene, Donald, 9 5 , 9 6 , 9 9 , 103 kegare, 34, 6 0 , 61. See also purification rites keigaisha, 143 Keisci Awa no Narato, 17, 2 4 , 2 2 I f f . Ki no Tsurayuki, 181 Kida Teikichi, 73

317

INDEX Kikudayii, 151, 152 kinosukc, 4 2 Kitagawa, Joseph, 14, 15 Kitamura Tetsuro, 35, 36 Kobe, 104, 110, 118 Koda Hachiman shrine, 1 4 4 , 145 Kojiki, 59, 6 2 , 1 2 1 - 2 4 , 127, 140, 146, 157, 181 kokeshi: fertility symbols, 37, as phallus, 37; unformed shape, 36, 37 kokumm, 2 6 2 koma inu (guardian dogs), 143 Konishi Jin'ichi, 182 IConjaku Monogatari, 31 Konokoro-gusa, 109, 110 Korea, 9 5 kotobuki (felicity), 171 kotodama shinko, 1 8 1 - 8 2 Kotoshiro-nushi no kami, 1 2 4 - 2 7 Koyo, and Kohyo shrines, 9 3 , 2 1 1 , 2 1 2 ksitigharba. See Mizuko Jizo kubi-kaki mngyo, 4 5 , 108 kusutsu,7, 12, 8 9 , 9 4 , 98ff., I l l , 1 4 0 ^ 1 ; altaic contact, 9 5 ; alternative meanings, 96ff.; and government, 99ff.; and Hyakudayu, 132; meaning of, 95ff.; as outcasts, 9 0 ; as ritual specialists, 12; untaxed status, 9 7 ICugutsu no Ki, 9 7 kugutsu-mawashi, 68, 9 6 , 1 0 0 , 109, 135 Kunisaki, 9 2 Kuroda Ningyo Za, 9, 157 kurofujo (black pollution). See death Kurosawa Akira, 78 Kurotoriko, 9 7 kuyo, 2 0 1 - 2 kyodo buyd (regional dances), 2 0 9 kyodo jjeino, 2 0 9 Kyogen, 4 1 , 199 Kyoho era, 1 3 0 , 135, 175, 176 Kyoto, 73, 155 Kyushu, 10, 1 3 , 3 9 , 9 1 , 9 2 , 9 3 , 116, 117, 242 Lacan, Jacques, 2 1 8 language, levels of, 180ff. Leach, Edmund, 85 Leech Child, 6 4 , 111, 133, 154, 158, 159, 164 linguistic imagery, 180ff. Long, Charles J., 1 0 4 Lotus Sutra, 39

mabiki. See infanticide Machida-ke kyuzo rakuchu rakugai, 106 madness, and reason, 55 magic, 35, 39, 84 Manyoshu, 31 manzairaku, 7 9 , 9 7 , 184, 185 ma-rebito, 7 6 , 121, 170; shinko, 7 4 , 81; worship, 118 marginalization, 5 6 - 5 7 , 6 6 , 67, 70, 7 1 , 74, 82, 84, 8 9 , 101, 2 2 9 , 2 5 9 ; and Ebisu, 113 Marionette, 3 0 , 31; material and spiritual relation, 31 Matabee furyuraku zu, 109 Matsu-bayashi, 11, 52, 79, 83; Gion festival as, 83; senzu manzai as, 83 McNeill, William, 129 Mead, Margaret, 231 meaning, collective, 6 medetai (felicity), 189 mediumship, incarnate, 33, 72 medley effect in ritual language, 186ff. Meigetsuki, 1 0 0 Meiji period, 86, 132, 163, 2 1 1 Meiji Restoration ( 1 8 6 8 ) , 10 Meisho Nishinomiya Annai Sha, 130 meisho zue, 151 memory, as lens to tradition, 10 menstruation, 6 1 , 73. See also pollution Mihara-gun, 138, 149, 250ff., 3 5 8 miko (female shaman), 7 3 , 1 4 1 . See also shamanism Minamoto no Yoritomo, 100 Ministry o f Education, 2 1 0 minkan shinko, 2 0 9 minzokugeind, 5, 2 0 4 , 2 0 8 , 2 1 0 , 2 1 3 Minzoku Geino Taikai, 2 4 2 minzokjigaku (folk studies), 2 0 8 , 2 1 0 miscarriage, 37, 38, 63 Misumi Haruo, 80, 82 mitamashiro. See goshintai Mitamura Engyo, 154 Miyamoto Tsuneichi, 1 4 0 Mizuko Jizo, 37, 38 Mombu Daijin, 2 1 2 Mongol, Ebisu as, 115 monomorai (begging), 87 Morioka (Iwate prefecture), 9 - 1 0 , 3 6 , 149, 157, 180 Morooka Nanrin, 132 motifs, visual, mythic, and poetic, 2 0 0 Motoon Norinaga, 126

318 mudmen dancers, 2 3 1 Mune Torasuke, 1 9 1 , 2 0 1 munt-age o sat, 171 murder, 57, 61, 6 3 , 65 Murogimi, 158 Muromachi period, 31, 35, 7 2 , 73, 78, 8 0 , 8 2 , 8 3 , 147 music, 186ff.; ritual, 4 9 , 7 7 ; songs, 97 Muslims, 5 4 Mutsu province, 37 Myojfoki, 80 myths: Chinese, 2 7 ; creation, 1 2 2 - 2 4 , 156, 159; function of, 122, 127; Indian, 31; Japanese, 4 3 , 5 9 , 6 2 - 6 5 ; about puppetry, 57 Nagahara Keiji, 72 Nagano, 9, 157, 170, 2 5 8 ; and Awaji puppetry, 2 5 8 nagare botoke, 117 Nagata Kokichu, 17, 18, 31, 35, 157 Nakamura Hajime, 87 Nakanishi Hideo, 2 4 2 , 2 5 6 , 2 5 8 , 2 5 9 Nakano Toraichi, 2 1 3 Nakayama Taro, 124 Namahage, The, 7 5 , 7 6 , 87 name shifting, tradition creation, 161 Namihira Emiko, 6 0 , 1 1 3 , 114, 117, 120, 124 nana-bake, 199 Nandancho, 147 Naniwa Kunie, 2 0 6 , 2 3 8 , 2 5 6 Naniwa puppet troupe, 131 naorai, 198 Nara period, 33, 34, 35, 8 1 , 9 0 , 9 3 , 102 Naruto, 2 5 1 Naruto Kaikan, 2 2 4 , 2 4 7 Naruto Straits, 2 0 6 , 2 2 4 Naruto Whirlpools, 2 0 6 National Museum o f Moscow, 2 0 4 National Theater o f Japan, 2 0 5 nativism, 2 1 0 , 2 2 0 , 2 6 3 nativist interpretation theory, 95ff., 2 5 0 natsukashii (natsukashimi), 214, 234, 235 nawabctri (territories in performance), 167 nembutsu dancers, 68 New Year, 9, 75, 76, 80, 85, 8 6 , 108, 109, 118, 124, 131, 145, 1 8 8 , 2 0 5 , 237, 247 nichijo seikatsu (everyday life), 13

INDEX Nichiren, 39 nigi-mi-ta.mil, 121 Nihongi text, 59, 6 2 , 8 1 , 1 2 1 - 2 6 , 140, 146,157 Niigata Prefecture, 116 Niimi Kanji, 1 4 3 , 166 ningyo. See puppets Ningyo Ganso Honke Uemura Gennojo Jireki, 1 5 4 ningyo hozonkai, 2 4 2 Ningyo JoruriKan, 8, 1 1 1 , 2 0 1 , 2 5 3 ningyo joruri tradition, 135, 150, 164, 2 5 8 ningyo matsuri, 12, 197ff., 206ff. ningyo no furusato, 241 ningyo shibai, 4, 7 ningyo tsukat, 23 ningyo-mawashi, 51, 1 5 3 , 2 0 1 . See kugutsu ningyo-shi, 4 6 Nishinomiya, 1 1 1 , 147, 159, 162, 180ff.; Awaji litigation with, 1 5 1 - 5 3 ; Center, 12, 111, 116; Daimyojin, 1 5 8 ; and drifted deity, 118, 1 5 9 ; Ebisu, 12, 71, 9 1 , 1 1 1 - 1 9 , 123, 134, 1 4 3 ; Ebisu Jinja, 1 1 1 ; Hyakudayu, 132; kugutsu, 4 5 , 111, 1 3 4 - 3 6 , 142; myth tradition, 1 5 6 - 6 3 ; non-religious ballads, 134; puppetry at, 131 ff., 1 5 0 , 1 5 2 , 1 5 6 ; Sanjo, 135; shrine, 70, 83, 106, 110, 111, 1 1 8 , 123, 130, 131, 148, 1 5 9 ; splintering o f tradition at, 135, 1 3 6 , 1 3 7 , 141, 156 nogake butai, 196ff. Noh, 9 , 4 0 , 4 1 , 108, 109, 1 3 4 , 170, 184, 185, 1 8 7 , 199, 2 0 0 , 2 0 9 , 248ff. nokanki, 198 nori-naoshi, 181 Noroma NingyS, 4 0 , 4 1 , 4 2 , 1 9 9 nostalgia, 204ff., 2 3 5 , 2 5 8 ; colonialism, 2 2 6 ; over decline of puppetry, 5, 2 0 8 ; deconstruction, 2 1 7 ; English historiography, 2 1 6 ; homesickness, 2 1 5 ; imperialist, 2 2 6 ; invoked by language, 181ff.; and medicalization, 2 1 5 ; nostarujii, 2 1 4 ; and political agenda, 2 1 4 , 2 1 6 ; reenactments as, 2 3 2 ; and revival of folk culture, 11, 13, 213ff., 2 4 4 , 2 6 1 , 2 6 2 ; signs as fetishes, 2 1 7 ; and tradition formation, 10; as veil for inhumanity, 226ff. nostalgic mode, 2 1 7 Noto Peninsula, 116 Nushima, 140 Nussbaum, Stephen, 2 1 6

IN D E X o-bon, 76 o-fuda, 112, 190ff. o-mikoshi, 118, 124 o-san ningyo, 35 o-shogatsu. See N ew Year O braztsov, Sergei, 30 O e M inosukc (puppet head-carver), 8, 251, 252 O e no M asafusa, 90, 9 6 , 9 8 , 9 9 , ZOO, 1 0 2 -6 , 130 O ga peninsula, 75 O gura M anabu, 116 O hnuki-T ierney, Em iko, 60 O ita prefecture, 91 O kina (the p u p p et), 3, 4 , 157, 176 O kina trad itio n , 9, 170, 176, 183, 248 O kina-w atari, 175, 176 O kuni-nushi no kami, 123, 125, 1 2 6 -2 8 O m ido, 151 O m ido H achim an D aibosatsu Shrine (Sanjo), 8, 71, 138, 142, 143, 145, 2 0 4 ,2 3 0 ,2 3 7 Ongyoku M ichi Shirube, 153 onmyoji (Taoist divm ers), 34, 6 8 , 73 O nogorojim a, 146, 147 ono m atopoeia, 186 O po-kuni-nushi, 127 O riguchi S hinobu, 74, 7 5 , 121, 140, 240 O saka, 72, H O , 228, 243 Osaka B unrakuza, 25 0 otherness: alterity, 71ff., 1 0 3 ,2 4 6 ; function of, 11, 74ff.; h abitat, 69; identification w ith, 88; and social m obility, 74; status of, 51, 52, 77, 84, 8 5, 161; transform a­ tion in, 81. See also itinerancy; m arginali­ zation, outsiders O tto , R udolph, 53, 74 outcasts. See otherness; outsiders outsiders, 5 1 -5 9 , 74ffi, 1 6 9 ,2 2 8 ; apprecia­ tion , 2 5 9 -6 1 ; changed for reenactm ent, 2 36; shift from pupp eteer to audience, 246; as traditional sym bolic elem ent for N ew Year, 76 Papua N ew G uinea, 231 parent and child shrines, 143 pariah. Se? outsider Parvati, 31 pastiche, 218 past o rientation, 207 patronage, 166 paulow nia w ood, 1 8 ,4 1

319 peachw ood puppets, 9 7 -1 0 2 perform ance territories, 166 Philippi, D onald, 126, 181 Pim iko, burial place, 141 Plagues a n d Peoples, 129 play, in fieldwork, 257 poetic definition, 105 pollution, 70; and blood, 73; districts of, 71, 73; levels of, 61; and taboo, 55, 5 9 -6 3 p ostm odernism , 15, 2 2 0; aesthetic m ode, 217; and tradition building, 243 Prague School, 246 preservation societies, 212 profu n d ity and mystery, 2 0 0 p u p p et nam es, 3 p u ppet theater, classical, 4 ; gypsy, 95 p u p p et trad itio n (ningyo shabai), 19, 21; art form s, 78; as images o f g o d , religion, 20 , 2 1 , 26 p u p p et trad itio n , C hinese, 95; com m unica­ tio n , 22 ; cultural com m onalities, 21; deconstructions, 21; defined, 2 0 , 21; history, 1 0 -1 1 ; itinerant n ature of, 4; and life stories, 6 p u p p et trad itio n , dram atic en tertain m en t, 106, 141; revival, 5, 88 p u p p et trad itio n , foreign origins of, 9 5 , 99 p u p p et trad itio n , as m ultivalent, 11, 86; Awaji prewar, 9 , 51 p u p p et trad itio n , narrative origins, 137ff.; perform ance, 2 1 , 74; representation in puppetry, 2 1 , 70 p u p p et troupes, 1 6 1 -6 3 puppeteers, 6 9 -7 1 , 7 3 , 9 4 , 104, 108, 111; appeal to children, 168; controlled by shrine, 148; as dead priests, 2 0 3 ; Ebisu w orship, 148; gypsies (k u k ie n re n g e ro / khelepaskero), 9 5 ; itinerancy, 166; leaving shrine service, 148; opposites, em b o d im en t of, 2 4 6 ; as outsiders, 51, 67 , 68; relation to puppets, 4 8 ; religious rites, 9 , 4 9 -5 1 ; ritual o f arrival, 85; role, 51, 5 7 -5 8 , 6 9 , 201ffi; specialization, 165, 166; status, 148, 161, 162; te rrito ­ ries, 164ff.; in unified trad itio n , 155. See also outsiders puppetry, ritual, 3, 5, 76, 150, 164, 167, 2 0 5 , 26 3. See also Awaji p u ppetry puppets, 11, 12, 32, 34, 4 7 , 6 8 , 89, 157; abode for divine beings, 3 0 -3 2 ; in b at­ tles, 9 Iff.; body substitutes, 133, 177, 20 2 ; cem etery (dekozam m a), 202;

puppets (cont.) channeling danger, 4 , 32; C hinese ety­ m ology (kuai-luai-tzu), 95; concrete m etaphorical self, 20 2 ; costum es, 178; and creativity, 27; driving o u t evii, 1 7 8; fear of, 84; and food, 197; and freedom o f expression, 23; as go-betw eens for hum an and divine, 4 8 , 52, 68; G reek (koukla), 95; as icons, 204ff.; as ideal ac­ tors, 28, 29; idenfication w ith d u rin g re ­ pressive eras, 27; im itative qualities, 203; in lieu o f hum an perform ers, 18, 19, 23, 25, 26, 30, 31; m anipulation of, 4 5 , 109; as m arginal, 84; as m etaphor, 2 7 , 48; opinions held of, 22 ; poetic definition, 105; origins of, 132, 133, 140; and pow er o f generalization, 30; as protective shield, 202; relative size of, 17, 165; as representational equals, 32, 34; and social co n tro l, 27, 4 8 , 203 ; and spiritual life, 2 2 , 23, 48; as spiritual vessels, 32, 3 4 ,4 7 , 4 8 , 69, 172, 202; static and m anipulated, 32, 45, 89, 94; as substitutes for small children, 3 2, 35, 132, 133; as theater, 22, 23, 29, 30, 40; Turkish (kukla), 95; wild nature of, 18 Pure Land B uddhism , 79 purification, 4, 7 ,9 , 18, 32, 33, 3 8 ,4 4 ,4 9 51, 59, 6 0 -6 3 , 65, 6 6 , 82, 164, 175 purification rites, 4, 7, 9 , 18, 3 2 , 60, 93 purity, hierarchical, 67 purity codes, 4, 7, 52, 56, 59, 6 4 -6 7 , 77 questionnaire as fieldwork m ethodology, 8

240ff., 24 7 ; stages of, 208ff.; and th eater vs. itinerancy, 2 0 8 ff. rites: developm ent, 4 , 5, 6 , 5 Off.; efficacy, 5; exchange o f goods, 86; history, 6; m ean­ ing, 6 , 31, 32, 182. See also kadozuke perform ance tradition ritual, 3, 6 7 , 68 ; chan t, 180ff.; language and efficacy, 180ff.; and m arking, 85, 24 0 ; perform ances, 12, 4 9 -5 0 , 7 2 , 21 0 ; substitutes, 33, 122. See also purification R itu a l Poetry a n d the Politics o f D eath in Early Japan, 182 ritual theory, 56; p e rfo rm e r/h o s t distinc­ tio n , 85; as shared system o f m eaning, 5, 4 4 , 56, 69 rituals, use of: appeasem ent, 18, 32, 44, 92ff., 130, 133, 158, 164, 21 1 ; ch an n el­ ing danger, 4 , 3 2 , 4 9 -5 0 ; children, 131; con tainm en t in, 165; dance in, 7 9 , 80, 24 9 ; vs. dram a, 164; driving o u t evil, 4 4 , 51; Fertility', 37, 4 0 -4 4 ; and hum an d e­ velopm ent, 60; insect pests, feelings of, 173; and itinerancy, 58; language, 182; music in, 4 9 , 7 7 , 171; sake drinking, 192ff.; social co n tro l, 66, 94; space, 2 3 7 , 240; substitutes, 35, 2 0 3 ; and symbolic action, 20 2 ; and theater, 2 2 , 2 0 9 , 244ff.; use o f effigies, 31 riverbeds, 66, 6 9 , 70; districts, 70; as po llu ted places, 70; and p opular im agination, 71 road, the, 10, 232; as sacred geography, 10 Rosaldo, R enato, 226 Rydjin Hishd, 104, 186

Kasbomon (Rasho G ate), 78 Redfield, R obert, 13 regionalism , 204ft'., 220ff. reincarnation, 37 religion, Japanese: authority, 147; C hinese elem ents in, 54; dctachability o f spirits, 34; as heterogeneous, 13; and history, 7, 9, 15, 16, 32; non-ecclesiastical practice, 5, 13, 74, 110, 138; pre-B uddhist, 60; p rotestantization, 88 religious authority, as m arginalized, 16 religious history, in general, 5 3 -5 4 Renaissance, 55 restored behavior, 231 revisionist revival, 207ff., 232ff. revival o f puppetry, 204ff.; and exoticism, 2 2 9 , 244f£; role o f foreign scholars, 2 0 8 ,

sacredness: as danger, 8 3 -8 4 ; d ouble-edged n atu re of, 121 Sado Island, 9 , 39, 116; sexual puppets, 39, 4 0 , 4 1 ,4 2 Saibara, 8 0 , 9 7 , 170, 183 Saikoku pilgrim age, 221 sakaki, 191ff. saki-m i-tam i, 121 salt, 62 sam urai, 162 Sanbaso, 9, 164ff., 2 3 7 , 245ff.; am agoi (rain rite), 175; chan t, 182; dance in, 170, 184, 2 4 9 ; decline, causes of, 205; Ebisu com pared to , 165; eyes m oving, 16 8 ; f u r i f u n Sanbaso, 167; h at, 178; hand Sanbaso, 171; kam i-okuri Sanbasd, 172; kotobuki Sanbasd, 171; lapse in prac-

INDEX

tice, 205; music in, 171; O kina, 179; p er­ form ance specialization, 165; the p u p p et, 3, 4, 157; revival (1990s), 229ff.; ritual, 9, 31, 86, 145, 146, 170; Senzai, 179; shiki Snnbaso, 171, 175ff., 248ff. Sanbaso rites, 183, 2 0 1 , 2 4 6 ; hono, 170, 174; O ktna-m ai, 170; O kin a -w a ta ri, 170; Shikt, 170, 174, 176ff. Sanbaso-mawashi. See kugutsu, 6 6 -6 9 , 86, 87, 170, 182 Sanjo districts (Awaji), 6, 7, 9, 50, 6 6 -7 5 , 137, 253; affiliation with cerem onial centers, 7 3 , 111; child b irth as pollu tio n , 73; crem ation preparation, 73; ind o o r, changes, 246; as m ode o f being, 74; m ura, 1 3 5 -3 6 ; as otherness, 74; and taxation, 73; as unsignifiable, 72 Sanjo D ok u m b o , 146 Sanjo hoshi, 74, 148 Sannom iya, 104 Sarugaku, 106 sasbikomi-mngyd, 45 saw ari (soliloquies), 199 Schechner, R ichard, 2 0 9 , 231 sea deities, 115, 116, 118, 119; strangeness o f object, 119; unco ntro lled arrival, 119 seamlessness: in Japanese scholarship, 15; in schizophrenia, 218 Senzai, the p u p p et, 3 , 4 , 170, 176, 183 Senshu m anzai. See Senzu m a n za i Senzu m a n za i, 11, 52, 79 , 81, 8 2 , 83; hoshi, 7 9 , 80, 81 Setsuyo O chiboshu, 131 Settsu N ishinom iya Shrine. See Nishinom iya, shrine sexual foreplay, 63 Shakespeare, 220 sham anism : female sham ans, 111, 140, 141; magic, 32, 33, 4 7 , 80; and m arking, 85; rejection of, 87. See also miho shamisen, 18, 2 3 , 111, 199, 258 shintai, 189, 234. See goshintai Shinto perform ance, 211 Shin to practice, 5 9 -6 0 , 92; and syncretism , 128 S hin to priests, 40 shinzo, 118 Shiva, 31 shrine guilds, 40 S hugendo practice, 39, 93 signification, 104ff. signifiers, 217ff.

321 Significations: Signs, Symbols a n d Images in the Interpretation o f Religions, 104 sm allpox, 129, 133 Sm ith, R o b ert J., 87 social discrim ination. See m arginalization; otherness society, Japanese: national identity, 88; reenactm ent, 232ff.; regionalism in, 13; revivalism in, 13 Sokoloff, A ndre, 26 sources, prim ary vs. secondary, 7 space, sacralized, 4 9 spirit pacification, 34, 43 spirituality, unconscious, 127 stage m ovem ent, 244ff. State C entral P u p p e tT h e a tre (M oscow ), 30 strangers. S rro u tsid ers sum o, 9 2 , 211 Su m o to H achim an shrine, 176 Susanoo no o -m ik o to , 123, 181 Suzuka C h iy o n o , 94 syncretism , 128 Szilagyi, D ezso, 27 ta no kam i, 83 t a i ( sea bream ), 189 Takeda S h in ’ichi, 258ff. Tales o f the H eike, 153 ta m a (spirit), 182 ta m a fu r i (spirit shaking), 177 tam ashii (soul), 32, 48 ta m a yo n ningyo, 172, 174 Tanki M a n roku, 131 ta ta ra reru , 124 ta ta ri-g a m i (curses), 117, 121 tate, 202 taxation and pollu tio n , 73 tayu, Japanese, 2 3 , 2 5 , 131, 176, 221 tek u g u tsu , 106 T en g u H isa, 4 6 tesarugaku, 106 th eater and ritual, 209 th eatre, British, 29 theodicy, 2 7 T heunissen, M ichael, 53 T h ird R eich, 54 T ogenuki Jizo, 39 T o h o k u d istrict, 2 1 , 3 6 , 3 7 , 73 toka (perform ance form ), 80 Tokugaw a period, 2 7 , 37, 6 9 , 7 1 , 7 3 , 79, 84, 89, 124, 125, 1 3 2 , 142, 147, 149, 153, 157, 161, 162, 2 2 1 , 2 2 8 ; as closed

322 Tokugawa period (cont.) society, 56; nativists, 54; repression of culture, 227ff. Tokushima, 162, 221, 225 tolerance, 2 6 tonaegoto, 180 torimono, 80, 82, 2 0 2 tourism, Japanese, 2 2 0 Toyota Hisae family, 157 tradition formation, 1 5 9 - 6 3 ; and postmodernism, 2 4 3 traditionalism, 2 0 9 Trance and Dance in Bali, 2 3 1 transformation, 115 transformation artists, 9 8 , 9 9 , 103 Tsuda Sokichi, 122 Tsunoda Ichiro, 9 5 , 9 9 turtles, 1 7 9 , 187 tutelary deity, 132, 143 Uber-Marionette, 2 9 Uemura Bunrakuken, 2 2 9 Uemura Gennojo, 1 4 2 , 143, 150, 154, 155, 157, 1 7 5 , 2 0 1 Uemura Kanetayu, 153 Uesugi-ke kyiizo rakuchii rakugai, 106ff. Uji, 73 Ujigami, 36, 132 Umazume M a s a r u , 4 6 , 1 7 1 , 1 7 6 , 1 9 9 , 2 0 0 , 203, 204, 219, 220, 2 3 6 , 2 3 8 , 2 4 0 , 2 4 1 , 2 4 2 , 2 4 3 , 2 4 4 , 2 5 6 , 262ff. umi-ire, 171 UN1MA, 2 5 8 Urakame, 251 urination, 63; as comic relief in performance, 4 1 , 4 2 Usa Hachiman, 7 1 , 9 1 f f . , 1 1 5 , 130, 141, 211 Usa hachimangu bdje-e mgi, 9 I f f . Usa H 6 j o - c , 2 H Usa lineage, 143 Usa Shrine, 9 1 f f „ 2 1 1

INDEX Vedas, 187 violence, 25—26 vomiting, 63 Wadahama, 1 1 8 , 160 Wakamatsu-shi, 1 1 6 , 1 1 7 Wakamori Taro, 1 3 9 - 4 2 , 147 Wakokuzaki (place name), 158, 1 5 9 , 160 Waseda Jinja, 1 7 1 , 174 Waseda Ningyo Za, 9 Waseda shrine, 174 waza-bito, 68 Williams, Raymond, 2 1 6 Williams, Tennessee, 2 4 3 Wu Hung, 9 8 Yamagami Izumo, 38 Yamaji Kozo, 9 9 Yamamoto Shunosuke, 4 0 , 4 1 , 87 Yamamoto Yoshiko, 75 Yamanashi prefecture, 157 Yamazaki Yoshinari, 131 Yanagita Kunio, 4 , 13, 15, 2 0 8 , 2 2 1 Yokoi Kiyoshi, 59, 6 2 , 6 6 Yokyoku (Noh texts), 170 Yomi (land o f dead), 126 yorishiro (spirit vessel), 6 9 . See also jioshintai; shintai YoshiiTaro, 111, 118 Yoshikawa-ke kyuzo rakuchii rakugai, 109 yujjen no sekai, 200 Tujoki, 130 yulong manyan zbiyi (Chinese performance form), 9 8 , 103 zashiki (as performance space), 4 9 , 7 8 , 2 5 5 Zeami, 39^10, 8 3 , 148 Zenkoku Minzoku Geino Kai, 2 1 0 Zenkoku Ningyo Shibai Samitto, 11, 2 4 2 , 244,253 Zeshi Rokujo Igo Sarujjaku Dangi, 83 Zoho Joruri Taikeizu, 154

ABOUT THE AUTHOR

Jane Marie Law is Assistant Professor of Japanese Religions at Cornell University. She is editor of Religious Reflections on the Human Body.