Akinari matsuno cause of death

TALES OF MOONLIGHT AND RAIN

Ueda Akinari

Columbia University Press New York

tales of moonlight and rain

Translations from the Asian Classics

Image has been suppressed

tales

of

moon l igh t

a n d

ra i n

Ueda Akinari

A Study and Translation by anthony h. chambers

C O L U M B I A U N I V E R S I T Y P R E S S N E W Y O R K

Image has been suppressed

Columbia University Press

Publishers Since 1893

New York Chichester, West Sussex

Copyright © 2007 Columbia University Press

All rights reserved

Library of Congress Cataloging-in-Publication Data

Ueda, Akinari, 1734–1809.

[Ugetsu monogatari. English]

Tales of moonlight and rain : a study and translation by Anthony H. Chambers.

p. cm. — (Translations from the Asian classics)

Includes bibliographical references.

ISBN 0-231-13912-8 (cloth : alk. paper)

ISBN 0-231-51124-8 (electronic)

I. Chambers, Anthony H. (Anthony Hood)

II. Title. III. Series.

PL794.8.U3413 2006

895.6'33—dc22 2006015127

Columbia University Press books are printed on permanent and durable acid-free paper.

Printed in the United States of America

c 10 9 8 7 6 5 4 3 2 1

Frontispiece: Tosa Hidenobu, portrait of Ueda Akinari (1786). (Tenri Central Library, Nara)

Acknowledgments vii

I n t ro du c t i o n 1

The Early Modern Period in Japan 2

About the Author 3

Bunjin, National Learning, and Yomihon 8

About Tales of Moonlight and Rain 11

About the Translation 34

Ta l e s o f M o o n li g h t a n d R a i n 45

pre fac e 47

B o o k O n e

Shiramine 51

The Chrysanthemum Vow 75

B o o k two

The Reed-Choked House 91

The Carp of My Dreams 110

contents

B o o k t h re e

The Owl of the Three Jewels 121

The Kibitsu Cauldron 139

B o o k f ou r

A Serpent’s Lust 155

B o o k five

The Blue Hood 186

On Poverty and Wealth 202

Bibliography 221

vi contents

Haruo Shirane provided the initial spark by asking me to translate three stories from Ueda Akinari’s Tales of Moon- light and Rain for his Early Modern Japanese Literature: An Anthology, 1600–1900, and then suggesting that I translate the whole collection. Deborah Losse, Lawrence E. Marceau, and Donald Richie deserve special thanks for their encour- agement and suggestions. The College of Liberal Arts and Sciences at Arizona State University provided time for the work by giving me a year’s sabbatical. Even with time, the study and translation would have been impossible without the pathbreaking work of earlier scholars and the compilers of the marvelous reference works we all depend on. Thanks go also to my incomparable circle of friends and colleagues, who sustain me emotionally and intellectually.

Michael Ashby read the first draft and made countless perceptive comments. I am also indebted to Jennifer Crewe, Anne McCoy, Irene Pavitt, and the rest of the staff at Colum- bia University Press. The anonymous readers recruited by the press offered encouragement, pointed out errors, and provided valuable advice. Any problems that remain are, of course, my own responsibility.

ACKNOWLEDGMENTS

The translation is dedicated to all my teachers, especially, Ch’en Shou-yi, who introduced me to the study of East Asia; Makoto Ueda, who introduced me to Akinari; Robert H. Brower, who tutored me in Japanese court poetry; and Edward G. Seidensticker, my principal mentor over the years.

viii acknowledgments

tales of moonlight and rain

Tales of Moonlight and Rain (Ugetsu monogatari), nine sto- ries by Ueda Akinari (1734–1809) published in Osaka and Kyoto in 1776, is the most celebrated example in Japan of the literature of the strange and marvelous. It is far more, however, than an engrossing collection of ghost stories. Japanese scholars regard it, along with Genji monogatari (The Tale of Genji, early eleventh century) and the stories of Ihara Saikaku (late seventeenth century), as among the finest works of fiction in the canon of traditional Japanese literature. The reasons for this esteem have to do primar- ily with Akinari’s elegant prose—a model of literary Japa- nese enriched by Chinese borrowings—and with his subtle exploration of the psychology of men and women at the extremes of experience, where they come into contact with the strange and anomalous: ghosts, fiends, dreams, and other manifestations of the world beyond logic and com- mon sense.

Tales of Moonlight and Rain exerted a powerful influ- ence in the twentieth century. Many novelists—including Izumi Kyoka (1873–1939), Tanizaki Jun’ichiro (1886–1965), Akutagawa Ryunosuke (1892–1927), Ishikawa Jun (1899– 1987), Enchi Fumiko (1905–1986), and Mishima Yukio

INTRODUCTION

2 introduction

(1925–1970)—were avid readers of the collection. Two of the tales inspired Mizoguchi Kenji’s cinematic masterpiece Ugetsu monogatari (1953; known to Western viewers as Ugetsu), which is widely regarded as “one of the greatest of all films.”1 Deeply rooted in its eighteenth-century cultural context, Tales of Moonlight and Rain is nonetheless a work of timeless significance and fascination.

The Early Modern Period in Japan

In 1603 Japan began to settle into a long era of relative calm and prosperity after a century of disastrous civil war (War- ring States period [Sengoku jidai], 1467–1568) and nearly forty years of gradual pacification and unification (Azuchi– Momoyama period) under the successive warlord-unifiers Oda Nobunaga (1534–1582), Toyotomi Hideyoshi (1536– 1598), and Tokugawa Ieyasu (1542–1616). The Tokugawa shogunate—the military regime established by Ieyasu— governed Japan for 265 years, an era that is commonly referred to as the Edo period, after the site of the shogun’s capital, or the Tokugawa period. The emperor and the court continued to hold ultimate, though symbolic, authority in Kyoto during these years, but real power was wielded by the Tokugawa bureaucracy until it collapsed in 1868 and the Meiji emperor moved from Kyoto to Edo, which was then renamed Tokyo.

Cultural historians refer to the years from 1603 to 1868 as the early modern period and have divided it into three parts on the basis of cultural and political developments: early (1603–1715), middle (1716–1800), and late (1801–1868).2 The first blossoming of early modern literature came toward the end of the seventeenth century, particularly with the work of three major figures: the fiction writer Ihara Saikaku (1642– 1693), the poet Matsuo Basho (1644–1694), and the drama- tist Chikamatsu Monzaemon (1653–1724).

introduction 3

The period we are most concerned with here, the eigh- teenth century, can be regarded as the time when Tokugawa culture reached its high point.3 The stability of the country under Tokugawa rule (among other factors) made possible a flourishing of artistic activity by and for commoners, who previously had enjoyed only limited access to high culture.4 The man now recognized as the outstanding Japanese author of fiction in the eighteenth century was such a commoner, Ueda Akinari. By the time he began writing, a good educa- tion was no longer the monopoly of the court aristocracy, the samurai class, and the clergy: literacy rates were comparable to those in Europe,5 and education had spread to large num- bers of affluent residents of the great cities of Kyoto, Osaka, and Edo.6 As a commoner, Akinari wrote primarily for an audience of other well-educated urban residents.

About the Author

Ueda Senjiro was born in 1734 in Osaka, then the commercial center of Japan.7 Akinari, the name by which he is known, is a pen name that he began to use in the early 1770s. His mother, Matsuo Osaki, was the granddaughter of a peasant from Yam- ato Province who had gone to Osaka to become a merchant; the identity of his father is not known. In his fourth year, he was adopted by a prosperous merchant named Ueda Mosuke. Surviving a severe bout of smallpox that left several of his fin- gers malformed, the young Akinari had a comfortable child- hood and received a good education, possibly at the Kaitokudo, one of the most prominent of the new schools chartered by the government to provide “an appropriately practical Confu- cian education” to the children of the merchant and artisan classes.8 The curriculum would have included the Confucian canon—the Four Books (Lun yü [Analects] of Confucius, Da xue [The Great Learning], Zhong yong [The Doctrine of the Mean], and Mengzi [Mencius]) and the Five Classics (I jing [The Book

4 introduction

of Changes], Shu jing [The Book of Documents], Shi jing [The Book of Songs], Li ji [The Book of Rites], and Chun qiu [Spring and Autumn Annals])—and Japanese classics, especially waka (thirty-one-syllable court poems), Ise monogatari (Tales of Ise, ca. 947), and The Tale of Genji.

Akinari’s earliest surviving literary efforts are haikai verses included in several collections published in 1753 and 1755. Although composing haikai (playful, humorous) poetry, an outgrowth of renga (linked verse), began as an amusing pas- time, it had evolved into a serious pursuit by the eighteenth century. Akinari continued to write haikai throughout his life—even if he did not take it as seriously as did some of his contemporaries9—and the pursuit brought him into contact with important literary figures in Osaka and Kyoto, includ- ing the painter and haikai poet Yosa Buson (1716–1783), and with proponents of kokugaku (National Learning or Nativist Study), which emphasized the philological study of ancient Japanese literature.10

Akinari never made his living as a writer, however. He married Ueyama Tama in 1760; they enjoyed a happy mar- riage but had no children. Akinari’s adoptive father died in 1761, leaving to Akinari the family business and the respon- sibility for supporting himself, his new bride, and his adop- tive mother, to whom he was devoted. They lost their busi- ness and all their belongings to a fire in 1771, after which Akinari turned to the study of medicine, probably under Tsuga Teisho (ca. 1718–ca. 1794), one of many intellectuals of the time who combined scholarship, writing, and medicine. Akinari worked as a physician in Osaka until 1787, when he retired from medicine and occupied himself with scholar- ship, teaching, and writing. How he supported his family during these years is unclear; he may have lived on accumu- lated savings, and he may have earned some money from teaching Japanese classics.

Along with his friend Buson and his sometime mentor Takebe Ayatari (1719–1774), Akinari was a classic example of

introduction 5

the eighteenth-century bunjin—a nonconformist, indepen- dent artist, typically a painter and writer, who, though not a member of the aristocracy, devoted himself or herself to high culture, stood aloof from commercial or political profit, and felt disdain for the “vulgarity” of contemporary society.11 What the bunjin of the mid-Edo period shared was “avoiding the ‘vulgar’ (zoku) and placing themselves on heights beyond the reach of the ‘vulgar.’”12 The bunjin ideal was inspired in part by the Chinese wen-jen (written with the same charac- ters as bunjin, signifying a person of letters) of earlier times, and one aspect of the eighteenth-century bunjin’s avoidance of vulgarity involved the study of Chinese culture, including vernacular Chinese fiction. This was true of Akinari.

Akinari’s first works of fiction, however, owe little to Chi- nese models and much to the ukiyo zoshi (books of the floating world) tradition of the late seventeenth and early eighteenth centuries, with its typically lighthearted, satirical treatment of the foibles of ordinary people. Akinari produced two col- lections of stories in this genre, Shodo kikimimi sekenzaru (A Worldly Monkey Who Hears About Everything, 1766) and Seken tekake katagi (Characters of Worldly Mistresses, 1767), which turned out to be the last significant ukiyo zoshi.13 Aki- nari quickly turned his attention to other interests.

One of these was National Learning. Akinari had begun a serious study of the Japanese classics, especially waka, before 1760. A few years later, he studied with Ayatari and then with Kato Umaki (1721–1777), a disciple of the great nativist scholar Kamo no Mabuchi (1697–1769). This led him to abandon the ukiyo zoshi tradition in favor of writing fiction that is far richer and more serious, as well as treatises on such classics as Tales of Ise; the Man’yoshu (Collection of Myriad Leaves, ca. 759), the oldest anthology of Japanese poetry; and the Kokin- shu (Collection of Ancient and Modern Poems, 905), the first imperially commissioned anthology of waka. His studies also embroiled him in a famous scholarly debate, which contin- ued over a number of years and on various subjects, with the

6 introduction

most noted of the National Learning scholars, Motoori Nori- naga (1730–1801)—a confrontation that Blake Morgan Young has characterized as “a clash between the rustic’s blind faith and the urbanite’s critical scepticism,” Norinaga, who lived in Ise, being the rustic.14 Their disagreements ranged from phonology to mythology. Norinaga’s arguments depended on his absolutely literal reading of ancient Japanese compendia of myth and history, while Akinari insisted on a more inter- pretative, empirical approach.15

Akinari’s studies of ancient Japanese literature merged with bunjin ideals, especially the avoidance of vulgarity and the fascination with Chinese fiction, to shape the solemn beauty of his masterpiece, Tales of Moonlight and Rain. The nine stories in this collection frequently allude to, quote from, and borrow words and phrases from Japanese classics and Chinese fiction and rise above zoku—even though most of the characters in the stories are commoners—to achieve the aesthetic ideal of ga (elegance, refinement), which had been associated with Kyoto court culture.16 No one doubts Akinari’s authorship of Moonlight and Rain, but he signed the work with a pen name and never acknowledged that he had written it. Although the collection is the principal basis for his fame, he probably would have preferred to be remem- bered for his waka, his studies in National Learning, and his expertise in a form of tea ceremony.

As a scholar, Akinari distinguished himself through edit- ing and publishing works by Kamo no Mabuchi and his cir- cle. In 1773 he wrote Ya kana sho (or Yasaisho), a commentary on the particles ya and kana, but for some reason he would not allow it to be published until 1787, when it appeared with a preface by Buson. Kaseiden, Akinari’s biographi- cal study of the great Man’yoshu poet Hitomaro (late sev- enth–early eighth centuries), apparently was written in 1781. The astonishing Reigotsu (ca. 1793) was “a comprehensive work in six sections, one each on the names of Shinto dei- ties, the names of Japan’s provinces, noted products of the

introduction 7

various regions, poetry, terminology, and systems of kana orthography,” but only the kana section survives.17 In 1794 he published Man’yoshu kaisetsu, a short study of the ancient anthology, and in 1800 he began a comprehensive commen- tary on the Man’yoshu, which, however, he left unfinished. Kinsa (1804) and Kinsa jogen (1804) bring together favorite poems from the Man’yoshu, with Akinari’s commentaries on them.

Akinari also compiled several miscellanies. Two are col- lections of humorous and satirical stories: Kakizome kigen kai (New Year’s Calligraphy and a Sea of Changing Feelings, 1787) and Kuse monogatari (Tales of Eccentricity, 1791; published 1822), whose title parodies Ise monogatari. Tsuzurabumi (Basket of Writings, 1805–1806), a collection of his prose and poetry, represents the final stage of Akinari’s serious liter- ary work, as he saw it; after it was published, he threw all his manuscripts down a well. Tandai shoshinroku (A Record of Daring and Prudence) was completed in 1808.18

Akinari wrote waka and haikai verse throughout his adult life and was one of the most distinguished waka poets of his time. His personal collection of waka, Aki no kumo (Autumn Clouds), was completed in 1807. He also distinguished him- self as an expert in senchado (the Way of sencha), a form of tea ceremony that employs tea leaves instead of the powdered tea of the better-known chanoyu ceremony. Seifu sagen (Triv- ial Words on Pure Elegance, 1794), his treatise on senchado, is a classic in the field. Pottery implements that Akinari made for the ceremony survive.19

Akinari did not abandon fiction after Moonlight and Rain. In 1808 and 1809, he gathered ten of his stories and essays under the title Harusame monogatari (Tales of the Spring Rain). The collection is uneven, partly because Akinari died before he could polish it to his satisfaction, and per- haps because he wrote more for his own enjoyment than for publication. The pieces in Spring Rain are less tightly structured than the stories in Moonlight and Rain, and the

8 introduction

element of the marvelous and strange is relatively unimport- ant. The language is plainer, and there is much less reliance on Chinese sources. Perhaps even more than the tales in Moonlight and Rain, the stories and essays in Spring Rain attest to Akinari’s studies in National Learning, particularly in the emphasis he placed on naoki kokoro (true heartedness, sincerity, guilelessness), which he apparently held to be the essential nature of the Japanese people. The stories in Spring Rain represent Akinari’s most important fiction aside from Moonlight and Rain.20

In 1793 Akinari and his wife moved from Osaka to Kyoto, where they lived in poverty near the temple Chion-in, on the east side of the capital. His wife died in 1797. Within a few months, Akinari, whose vision had been failing for some time, went blind, but then regained partial vision in one eye. He continued his writing and scholarship as he moved here and there in Kyoto, depending on friends for support, until his death in 1809 on the twenty-seventh day of the Sixth Month (August 8, in the Western calendar), in his seventy- sixth year. His grave is at the Buddhist temple Saifukuji, near the Nanzenji monastery.

Bunjin, National Learning, and Yomihon

The result of Akinari’s synthesis, in Moonlight and Rain, of a bunjin orientation with the National Learning was a new genre: the yomihon (books for reading).

The distinction between ga and zoku arose from ancient Chinese poetics and was embraced by Japanese artists of the Tokugawa period. As the painter Gion Nankai (1677–1751) said, “ga is neatness, propriety, elegance; zoku is vulgarity.”21 This analysis was applied to all the arts, including painting and literature. From the Japanese point of view, elegant liter- ary genres encompassed Chinese poetry and prose, includ- ing fiction; Japanese court poetry and linked verse; classi-

introduction 9

cal monogatari, such as Ise and Genji; and no dramas.22 The traditional ga–zoku aesthetic was modified, however, by Japanese artists in the seventeenth and eighteenth centuries from a system that distinguished clearly between the courtly and the common, into a quest for elegance in realms that had traditionally been considered vulgar. Thus Basho urged his students to raise their minds “to an enlightened state, [and then] return to zoku,” by which he meant “practicing ga while remaining in the ordinary zoku world of haikai.”23 In short, “Basho raised haikai poetry, traditionally a zoku form, to the world of ga, thereby confounding the traditional dis- tinctions of ga and zoku.”24

The objective of finding elegance in the vulgar dovetails with one of the goals of the scholars of National Learning, the “articulation of links between the mythological past, the recorded history of the aristocratic few, and the daily lives of common folk.”25 This agenda is related, of course, to the rising wealth and influence of the urban classes—primar- ily merchants and artisans—in the early modern period and their desire to participate in the high culture associ- ated with the court aristocracy. The consequent blurring of the distinction between ga and zoku can be seen clearly in Moonlight and Rain. As a bunjin, Akinari rejected the com- mon, and all the elegant genres are reflected in Moonlight and Rain. At the same time, the peasants (zoku) in “The Reed-Choked House,” for example, are remarkably well versed in waka (ga), and the inclusion of haikai (zoku) in the same context as waka and Chinese poetry (ga) in “The Owl of the Three Jewels” implicitly raises haikai to the same level of elegance. In Moonlight and Rain, then, peasants and haikai participate in the aristocratic tradition as Akinari lifts them—and eighteenth-century Japanese fiction—from the vulgar realm of ukiyo zoshi to the elegant sphere of court poetry and monogatari.26

The National Learning agenda is reflected in Moonlight and Rain in at least two other ways. First, the philological

10 introduction

study of ancient Japanese texts, one of the principal activi- ties of National Learning scholars, influenced Akinari’s choice of vocabulary and phrasing so often that a reader is hard put to count the examples.27 Indeed, the abundance of archaic words and expressions from, and allusions to, the Kojiki (Record of Ancient Matters, 712), the Man’yoshu, Tales of Ise, the Kokinshu, The Tale of Genji, and other clas- sics is the main reason that Moonlight and Rain is difficult to read. The classical lexicon also serves to associate Moonlight and Rain with court literature. Second, in opposition to the Confucian emphasis on rationalism, scholars of National Learning insisted that many things lie beyond the ability of human beings to understand, analyze, and explain—a belief that was based on an unquestioning acceptance of Japanese mythology.28 While Akinari rejected Norinaga’s uncritical embrace of ancient mythology, he did share the National Learning scholars’ propensity to “celebrate the mysterious wonders of life,”29 which takes an especially vivid form in Moonlight and Rain.

In synthesizing the bunjin aesthetic and National Learn- ing, Akinari produced a masterpiece in a new genre—the yomihon, a more serious form of literature than its prede- cessor, the ukiyo zoshi. The term yomihon comes from the genre’s characteristically heavy emphasis on the written text, as opposed to oral narratives and booklets in which illustra- tions play a central role. The language of yomihon, including Moonlight and Rain, is elegant, somewhat archaic, and often full of allusions to Chinese and Japanese antecedents. In short, the emphasis is on serious reading. The first yomihon writers were Tsuga Teisho, who probably instructed Akinari in medicine, and Takebe Ayatari, one of Akinari’s mentors in National Learning. Teisho’s Hanabusa soshi (A Garland of Heroes, 1749) is considered the first yomihon. Its prose style is characterized by wakan konko (a blend of Japanese and Chinese) and gazoku setchu (a blend of elegant and vulgar). Like Moonlight and Rain, A Garland of Heroes consists of nine

12 introduction

Why the preface bears the date “Meiwa 5, late spring” (the Third Month of 1768) has been the subject of considerable research and discussion, since the preface and the stories were first published eight years later. There are good reasons to think that a preface that Akinari had drafted in 1768, as part of the “Saigyo Stories” project, was followed by eight years of studying, writing, and revising before the tales in Moonlight and Rain reached their present form.33 Another possibility is that Akinari used the date of 1768 so that his work would appear to be contemporaneous with A Tale of the Western Hills, the preface of which is dated “Meiwa 5, spring, Second Month.”34

Moonlight and Rain belongs, of course, to a different genre—yomihon—from A Worldly Monkey and Worldly Mis- tresses and, presumably, their planned sequels. Neverthe- less, the titles of the unpublished works contain tantalizing suggestions of connections with Moonlight and Rain. First, both titles—“Tidings from a Cargo Ship in Various Prov- inces” and “Saigyo Stories: Poetic Sites Bundled in a Dyed Cloth”—anticipate the prominence of travel in Moonlight and Rain (in all but the last of the nine tales) and the location of all the stories in the provinces (as opposed to the great cities). Second, “Saigyo Stories: Poetic Sites Bundled in a Dyed Cloth” anticipates Moonlight and Rain in two additional ways: Saigyo, the beloved poet-monk of the twelfth century, appears in the first Moonlight and Rain story, “Shiramine”; and “poetic sites” (utamakura), place-names used frequently in poetry and listed in handbooks of poetic composition, are mentioned in almost all the tales, with special prominence in “Shiramine,” “The Carp of My Dreams,” “The Owl of the Three Jewels,” and “A Serpent’s Lust.” In short, “various provinces,” “Saigyo,” and “poetic sites” in the titles of the unpublished ukiyo zoshi are important elements in Moon- light and Rain. There can be little doubt that Moonlight and Rain grew from the germs of “Tidings from a Cargo Ship in Various Provinces” and “Saigyo Stories.” The resulting

introduction 13

yomihon is a work of far greater psychological depth, narra- tive sophistication, and historical and philological awareness than A Worldly Monkey and Worldly Mistresses, and it incor- porates two new elements: the adaptation of Chinese stories and the strange or anomalous.

t itle

The title Ugetsu monogatari (literally, “rain-moon tales”) comes from the phrase “misty moon after the rains” in the preface. It alludes to the no play Ugetsu, in which Saigyo appears, as he does in “Shiramine,” and in which rain and the moon are central images.35 Commentators have also pointed to a passage in “Mudan deng ji” (Peony Lantern), a story in Qu You’s Jiandeng xinhua (New Tales After Trim- ming the Lamp, 1378), one of Akinari’s principal sources for “The Kibitsu Cauldron,” which suggests that mysterious beings appear on cloudy, rainy nights and in mornings with a lingering moon. In any case, educated East Asian readers would probably guess immediately that a book containing the term “rain-moon” in its title would deal with the strange and marvelous.

sources

Much has been written about Akinari’s use of Chinese and Japanese sources in Moonlight and Rain—more than sixty Chinese sources, according to Noriko T. Reider, and more than a hundred Japanese.36 (For the titles of important sources, see the introductions to the tales.)

Akinari used his sources in several ways. For some tales— “The Chrysanthemum Vow,” “The Carp of My Dreams,” “The Kibitsu Cauldron,” and “A Serpent’s Lust”—he bor- rowed the story line of a Chinese work, always with signifi-

14 introduction

cant modifications that ease the transition to a Japanese set- ting. Further, many scenes and situations in the tales echo those in Chinese or Japanese sources. Examples include the description of Katsushiro’s house when he returns from the capital in “The Reed-Choked House,” which echoes the “Yomogiu” (The Wormwood Patch) chapter of The Tale of Genji; the depiction of the temple at Yoshino in “A Serpent’s Lust,” which echoes the “Wakamurasaki” (Lavender) chap- ter of Genji; and the arrival of Kaian at the village in “The Blue Hood,” which echoes chapter 5 of Shuihu zhuan (Water Margin, fourteenth century), attributed to Shi Nai’an and Luo Guanzhong. Akinari also borrowed words and phrases from his Chinese sources and from the Japanese classics he stud- ied, especially the Kojiki, the Man’yoshu, Tales of Ise, and The Tale of Genji. Finally, he seems to have structured individual stories along the lines of Chinese tales and in imitation of the structure of no plays, and the organization of the collec- tion as a whole seems to be influenced by the no.

Far from trying to hide his indebtedness to Chinese and Japanese precedents, Akinari undoubtedly hoped and expected that his readers would derive pleasure from recognizing his sources and appreciate his ingenuity in adapting them. The borrowings, allusions, and echoes that fill Moonlight and Rain also add richness and complexity to the tales. As with the references to earlier texts in The Tale of Genji and the use of honkadori (allusive variation) in Japanese court poetry, the reader’s awareness of other texts interacting with Akinari’s adds resonance and depth to the reading experience.37 The borrowings also draw the reader into the text and involve him or her in the creative process, as they reward, flatter, and delight the reader who is erudite enough to recognize them.38 Finally, the liberal use of Chinese and courtly Japanese sources lifts Moonlight and Rain, by association, into the elegant realm of Water Margin and Genji, the two works that Akinari mentions at the beginning of his preface, and, in the same way, lifts

introduction 15

Akinari himself into the lofty company of his Chinese and Japanese predecessors.

Even when Akinari’s borrowings from Chinese sources are most direct—in “The Chrysanthemum Vow,” “The Carp of My Dreams,” “The Kibitsu Cauldron,” and “A Serpent’s Lust”—he ingeniously adapted the stories to Japanese set- tings and enriched them with a psychological complexity that is absent in their Chinese counterparts. Again and again, the reader is struck by the wonderful aptness of the time and place in which Akinari placed his tales. By mak- ing Akana Soemon a samurai in “The Chrysanthemum Vow,” for example, he introduced the themes of samurai loyalty and honor, whereas the character who corresponds to Soemon in the Chinese story is a merchant who sim- ply forgets the date of his appointment.39 In “The Carp of My Dreams,” Akinari introduced the crucial theme of Bud- dhist compassion by placing a Buddhist monk at the cen- ter of the story and invoked a long tradition of descriptive Japanese literature and art by setting the story at Miidera. In “The Kibitsu Cauldron,” he introduced Shinto elements— prophecy and the role of spirits—by connecting the charac- ters to the Kibitsu Shrine. In “A Serpent’s Lust,” he imbeds the Chinese-inspired story line in the context of Japanese legends about storied places: Kumano, Yoshino, and Dojoji. In these carefully chosen settings, Akinari’s characters reveal distinctive personalities, unlike the characters in his Chinese sources. As Robert Ford Campany has pointed out, the authors of Chinese anomaly accounts were not concerned with “the ‘inner’ nature (xing of intellectual and emotional disposition, nor the structure of the self’s ascent toward perfection through self-cultivation, but pre- cisely humankind’s taxonomic place among other kinds of beings, the nature of its relationships to other kinds.”40 In Moonlight and Rain, by contrast, it is precisely the charac- ters’ inner natures—saga, the reading in Japanese of the character read xing in Chinese—that concerned Akinari, as

Image has been suppressed

16 introduction

Uzuki Hiroshi emphasizes in his commentaries.41 Donald Keene makes the same point: “The very fact that one can describe Katsushiro’s character places him in an altogether different category from Chao or Seiroku [who correspond to Katsushiro in the Chinese and Japanese antecedents to “The Reed-Choked House”], neither of whom displays any distinctive traits.”42

narrating the strange

Moonlight and Rain has been called a collection of “ghost stories,”43 “gothic tales,”44 and “tales of the supernatural.”45 In Japanese, they are called kaidan; indeed, the edition of 1776 includes the subtitle Kinko [present and past] kaidan. As Reider says, “Kaidan are tales of the strange and mys- terious, supernatural stories often depicting the horrific and gruesome,” and the word kaidan means “narrating the strange.”46 No one would argue with “strange and mysteri- ous,” but “supernatural” is probably an inappropriate word, since what is considered to be supernatural in one culture is regarded as merely strange—but natural—in another.47 Belief in revenants, spirit possessions, and other phenom- ena that we might call “supernatural” was widespread in eighteenth-century Japan and was apparently shared by Akinari.48 If the term “supernatural” is inappropriate, so is “fantastic,” as defined by Tzvetan Todorov, because “the basis of the fantastic is . . . the ambiguity as to whether the weird event is supernatural or not,”49 and such ambiguity is absent in Akinari’s world. “Strange” and “anomalous,” words that have been used in the study of Chinese stories, are more useful when discussing Moonlight and Rain.50

Strange beings abound in Japanese art, folklore, and liter- ature. They include kami (Shinto deities); spirits, deities, and divine beings from other traditions, such as Buddhist and Chinese lore; spirits of humans, living or dead, that can pos-

introduction 17

sess other people; revenants; oni (demons and fiends); tengu (goblins); trickster animals, such kitsune (fox) and tanuki (raccoon-dog); and other animals, such as serpents, that have strange powers. All of these, except kami and trickster ani- mals, figure prominently in Moonlight and Rain. The venge- ful ghost of the former emperor Sutoku returns to earth as king of the tengu in “Shiramine.” In “The Chrysanthemum Vow” and “The Reed-Choked House,” a faithful revenant ful- fills a promise. In “The Owl of the Three Jewels,” vengeful ghosts return as asura (J. ashura or shura), violent human beings who, in Buddhist lore, are reborn as violent demons. “The Kibitsu Cauldron” features the possessing spirit of a jealous woman—first when she is alive and then after she has died. In “A Serpent’s Lust,” a jealous serpent in the form of a woman seeks revenge on her husband. A monk turns into an oni and then miraculously stays alive for a year while meditating, in “The Blue Hood.” “On Poverty and Wealth” features a little man who introduces himself as the spirit of gold. “The Carp of My Dreams” differs from the other sto- ries in that no anomalous being plays a major role; instead, the story deals with the anomaly of a man who crosses the boundaries between human and animal, and between the waking world and the world of dreams.

Tengu and oni, which have no exact Western equivalents, require some explanation. Tengu are goblins said to live deep in the mountains. In Japanese art, they often resemble birds but sometimes take human form, with wings and a beak or long nose. They were “regarded as harbingers of war” because of their “insatiable desire to be destructive and to wreak havoc upon people’s lives.”51 It is fitting, then, that the malicious spirit of the former emperor Sutoku becomes a tengu to exact his revenge. The Minamoto leader Yoshi- tsune, who had defeated Sutoku’s enemies, was said to have learned martial arts from a tengu. Tengu were apparently brought under control by the Tokugawa government, which issued commands to them and expected them to obey.52 Oni

18 introduction

are usually depicted as grotesque, humanlike beings with horns, fangs, and claws, and clad in a tiger-skin loincloth or, sometimes, a monk’s robes, as in the Otsu prints that depict an oni-nembutsu (oni dressed as an itinerant monk).53 Some oni serve as soldiers in the Buddhist hell, but oni can also be benevolent, as was Ryogen (912–985), chief abbot of the Tendai sect, who is said to have become an oni after his death in order to protect the Enryakuji temple complex on Mount Hiei.54 Finally, the word oni is often applied to a cruel or frightening person. The mad abbot in “The Blue Hood” becomes an oni in this last sense of the word.

The categories suggested by Campany in his study of Chinese anomaly accounts help clarify the nature of the anomalies in Moonlight and Rain.55 As Campany says, “Most (but not all) anomalies . . . occur at or across bound- aries”56—for example, between humans and animals or the living and the dead. Two of Akinari’s stories involve “anom- aly by transformation,”57 in which a being is metamor- phosed across a boundary that normally separates humans from animals and the apparent from the real: “The Carp of My Dreams,” in which a man seems to become a fish,58 and “A Serpent’s Lust,” in which a woman turns out to be a serpent.59 A category not posited by Campany—the bound- ary between mineral and animal—also figures in “The Carp of My Dreams,” when paintings turn into real fish. Most of the stories in Moonlight and Rain involve “anom- aly by contact,” specifically, “contact with the realm of the dead”:60 “Shiramine,” “The Chrysanthemum Vow,” “The Reed-Choked House,” “The Owl of the Three Jewels,” and “The Kibitsu Cauldron,” which also features the vengeful, possessing spirit of a living person, a type not mentioned by Campany. “The Carp of My Dreams” belongs to the category of “contact through dreams.”61 In the last story, “On Poverty and Wealth,” contact occurs between a human being and a spirit.62 Finally, two stories involve “sexual con- tact with non-human (or non-living human) beings”:63 “The

introduction 11

stories adapted from Chinese sources to Japanese settings and grouped into five books, as does its sequel, Shigeshige yawa (1766). It seems likely that Akinari modeled his col- lection on Teisho’s.30 Ayatari’s Nishiyama monogatari (A Tale of the Western Hills) appeared in 1768, eight years before the publication of Moonlight and Rain. In contrast to Teisho’s and Akinari’s yomihon, A Tale of the Western Hills consists of ten chapters grouped into three books and concerns a con- temporary scandal in the capital. The prose style of Moon- light and Rain combines that of A Garland of Heroes with the elegant, neoclassical prose of A Tale of the Western Hills.31 Tales of Moonlight and Rain is unquestionably the finest of the early yomihon.

About Tales of Moonlight and Rain

composit ion and publicat ion

An advertisement at the end of Akinari’s story collection A Worldly Monkey Who Hears About Everything announces the forthcoming publication of two more works by the same author: Characters of Worldly Mistresses and “Shokoku kaisen dayori” (Tidings from a Cargo Ship in Various Provinces).32 Worldly Mistresses, published the following year, repeats the advertisement for “Tidings from a Cargo Ship,” adding “Se- kenzaru kohen” (A Worldly Monkey, Part Two) to the title, and announces a forthcoming work to be called “Saigyo hanashi utamakura somefuroshiki” (Saigyo Stories: Poetic Sites Bundled in a Dyed Cloth). The context and the titles suggest that both “Tidings from a Cargo Ship” and “Saigyo Stories” were to have been ukiyo zoshi like A Worldly Monkey and Worldly Mistresses, but neither “Tidings from a Cargo Ship” nor “Saigyo Stories” was published. In their place came Tales of Moonlight and Rain in 1776, with a preface dated 1768, the year after the publication of Worldly Mistresses.

introduction 19

Reed-Choked House,” in which “marital relations estab- lished during life continue after death,”64 and “A Serpent’s Lust.” “The Blue Hood” generally follows the pattern that Campany has described for Buddhist tales (for a detailed explanation, see the introduction to “The Blue Hood”).65 In most of Akinari’s stories, then, the anomalies correspond to categories that had been used for centuries by Chinese writers. This is not only because Akinari adapted Chinese stories, but also because Chinese lore had been trickling into Japan, along with Chinese Buddhist and secular texts, for more than a thousand years and naturally influenced the methods of Japanese storytellers.

sett ings

The chronological and geographical settings of each story in Moonlight and Rain are shown in the table. The names of provinces are accompanied by the names, in parentheses, of the corresponding modern prefectures.

Title Date Location

1 “Shiramine” 1168 Sanuki (Kagawa)

2 “The Chrysanthemum Vow” 1486 Harima (Hyogo)

3 “The Reed-Choked House” 1455–1461 Shimosa, Omi

(Chiba, Shiga)

4 “The Carp of My Dreams” 923–931 Omi (Shiga)

5 “The Owl of the Three Jewels” ca. 1616 Mount Koya, in Kii

(Wakayama)

6 “The Kibitsu Cauldron” ca. 1500 Kibi, Harima

(Okayama, Hyogo)

7 “A Serpent’s Lust” 794–1185? Kii, Yamato

(Wakayama, Nara)

8 “The Blue Hood” 1471–1472 Shimotsuke (Tochigi)

9 “On Poverty and Wealth” ca. 1595 Aizu (Fukushima)

20 introduction

Three things are striking about the settings. First, they are specific. Second, all the stories except “The Owl of the Three Jewels” take place before the establishment of the Tokugawa shogunate in 1603. Third, none of the stories is set in any of the three great cities of Akinari’s time—Kyoto (the capi- tal), Osaka (the center of commerce), and Edo (the seat of the shogunate)—where most of his readers lived. In other words, the stories are specifically located at a chronological, geographical, and, therefore, political distance from Akinari’s world. Various reasons can be imagined for this distancing, and it affects the stories in various ways.

None of the stories is set “in a certain province,” but always in a real place and, with the exception of “A Serpent’s Lust,” at a more or less precisely specified time, not “once upon a time” or its Japanese equivalent, mukashi mukashi. This specificity has the effect of grounding the strange beings and events in the real world, thus lending plausibility to the stories. Speci- ficity also makes the geographical and chronological distanc- ing more effective than vagueness would have done.

Improvements in highways and other means of trans- portation—not to mention prosperity and relative freedom from bandits and warlords—made domestic travel in Japan fairly safe and increasingly popular during the Edo period. By contrast, the dangers of travel in the fifteenth century are portrayed in “The Reed-Choked House.” Akinari is known to have traveled widely in Japan, possibly as far as the Kanto region (now greater Tokyo).66 Journeys and far-off destina- tions thus were subjects of intrinsic interest to the author and his intended audience. The opening words of the first story—“Allowed by the guards to pass through the Osaka Barrier, he found it hard to look away from the mountain’s autumn leaves, but he traveled on to Narumi Shore”—put the reader on notice that the narrative is moving away from the capital and toward the provinces. More than a love of travel is involved, however. From the perspective of the cap-

introduction 21

ital—and of Osaka and Edo—the provinces were more or less exotic places and definitely less civilized than the great cities. This conventional attitude has a long history in Japan. In The Tale of Genji, even Uji, just a few miles outside the capital, is depicted (with some hyperbole) as a wild, danger- ous place. Nor is this attitude unique to Japan. As Campany has written, “From the point of view of a center of urban culture, the ‘distance’ between the center and its periphery is seldom a matter of mere geographical space, or of the calendrical time required for the journey out and back. The peripheral is, from a centrist perspective, the anomalous— the external other.”67 Thus when a story moves away from the capital, the reader is prepared to encounter strange and wonderful things; mysterious happenings are more believ- able when they occur far from home; and a provincial setting facilitates the Japanese reader’s “suspension of disbelief,” in Coleridge’s phrase.

Nevertheless, anomalies, even when they occur in dis- tant provinces, represent disorder, and it would not do to suggest that the Tokugawa regime permitted disorder any- where in Japan—especially in such extreme forms as canni- balism, necrophilia, and sexual relations between humans and nonhumans.68 This is why the chronological settings of the stories, before Tokugawa rule began, is at least as significant as the geographical settings. As a collection of pre-Tokugawa anomalies in the provinces, Moonlight and Rain indirectly draws attention to the orderliness of the Tokugawa era and reinforces the normality of the center, the big city. “The Owl of the Three Jewels,” the only story that is set during Tokugawa times, is the exception that proves the rule: it opens with praise for the Tokugawa regime and portrays one of the predecessors of the Tokugawa as having been so bloodthirsty that he was reborn as an asura. The old days, according to these stories, were not nearly as good as the present.

22 introduction

structure of the stories

Much of Campany’s analysis of the structure of Chinese anomalous accounts is applicable to the Moonlight and Rain stories.69 He identifies ten typical structural elements, the last three of which are not present in all the stories he stud- ies. I use “The Owl of the Three Jewels” to illustrate these elements in Moonlight and Rain.

1. Chronological and geographical settings. The opening paragraph hints that “The Owl of the Three Jewels” takes place during the Edo period; this time frame is made explicit later: “more than eight hundred years” after the founding of Mount Koya. The second paragraph identifies the place: “In a village called Oka, in Ise.”

2. Specifier. The story provides “some indication . . . of the specific situation or string of events . . . , so that the focus is narrowed to a particular occasion when an event happened. This focusing is usually accomplished in part by introducing a protagonist.”70 In “The Owl of the Three Jewels,” a “man of the Hayashi clan transferred his affairs to his heir, shaved his head . . . , [and] changed his name to Muzen.”

3. Process. “Some sort of familiar process or type of inter- action is set underway—some human activity with a predict- able sequence.”71 Muzen “looked forward to traveling here and there in his old age” and sets out on a journey to Mount Koya with his son.

4. Hints. The story provides clues that “something anoma- lous is about to occur.”72 The most obvious in “The Owl of the Three Jewels” are Muzen’s decision to spend the night in front of Kobo Daishi’s mausoleum and the riveting cry of the owl.

5. Limen. Campany’s list of liminal markers includes several that figure in “The Owl of the Three Jewels”: travel through mountains, sunset, and a gateway.73 In addition, the extraordinary historical and spiritual associations of Mount

introduction 23

Koya make it, ipso facto, a liminal place, and the mausoleum itself represents liminality.

6. Pivot. “Something distinctly odd now happens and the reader—usually joined at this point . . . by the protago- nist—becomes unmistakably aware of the strangeness of the situation.”74 Muzen, “taking out his travel-inkstone, . . . wrote down the verse by lantern light, then strained his ears in hopes of hearing the voice of the bird again, when, to his surprise, he heard instead the stern voice of a forerunner, coming from the direction of the distant temples and gradu- ally drawing closer.”

7. Climax. “The full force of the anomaly hits home to both protagonist and reader.”75 The climax of “The Owl of the Three Jewels” comes when Joha identifies Hidetsugu and the other ghosts: “It was so horrible that the hair on Muzen’s head would have stood on end, had there been any hair, and he felt as though his innards and his spirit alike were flying away into space.” The climax continues to the point where the ghosts reveal their true form as asuras and threaten Muzen and his son.

8. Response. “Some tales . . . continue by reporting the protagonist’s response” to the climax.76 In “The Owl of the Three Jewels,” Muzen produces the verse required of him and then faints.

9. Outcome. “In tales in which there is some response by the protagonist, comment is usually made on how things worked out.”77 In “The Owl of the Three Jewels,” Muzen and his son wake up, return to the capital, and seek medical treatment.

10. Impact. “Very occasionally—especially in Buddhist miracle tales and in stories concerning the origins of cults and temples—further comment is made on reactions by persons other than the protagonist, or on some later situ- ation relevant to the tale. In every case these comments concern the lasting impact made by the narrated event on a person or group of people or on the landscape.”78 “The Owl of the Three Jewels,” of course, concerns the origin

24 introduction

of Mount Koya and the legends surrounding Kobo Daishi. The story concludes: “One day as he was passing the Sanjo Bridge, Muzen thought of the Brutality Mound [containing the remains of Hidetsugu and his family] and felt his gaze being drawn toward the temple. ‘It was horrible, even in broad daylight,’ he recounted to people in the capital. The story has been recorded here just as he told it.” These sen- tences describe the lasting impact of the events at Mount Koya both on Muzen and on the landscape (in the form of the Brutality Mound). The impact on other people is sug- gested in the survival of the story as it has been passed down from Muzen.

The structure of “The Owl of the Three Jewels” thus resembles that of Chinese anomalous accounts. A similar analysis could be made of the other stories in Moonlight and Rain, all of which follow this pattern. It is characteristic of Akinari’s tales that the last three elements—response, out- come, and impact—which are often absent in Chinese sto- ries, not only are present but are developed at length (except in “The Kibitsu Cauldron,” “A Serpent’s Lust,” and “The Blue Hood,” where they are present but abbreviated). The third through seventh elements are repeated several times in “A Serpent’s Lust,” which thus might be regarded as sev- eral stories in one. The response section of “On Poverty and Wealth” is exceptionally long, consisting of Sanai’s conversa- tion with the spirit of gold.

Another way to describe the Moonlight and Rain stories is in terms of the structure of no plays, as many commentators have noted.79 Indeed, Akinari invited comparisons by using the title of a no play—Ugetsu—in the title of the collection. Four elements of no structure can be found in some or all of the stories.

1. Michiyuki. Originating in the no theater and found in other performance and literary genres as well, a michiyuki

introduction 25

(going on the road) is a literary convention in which the route, sights, and impressions of a journey are evoked with a litany of familiar place-names, often modified by makura- kotoba (pillow-words) or other epithets. In addition to pro- viding a display of beautiful rhetoric and enriching the audience’s experience by prompting the memory and imagi- nation to envision famous scenes and their associations, a michiyuki guides the reader into a world apart, where unex- pected, wondrous beings and events are likely to be encoun- tered. Three of the stories in Moonlight and Rain include con- spicuous michiyuki: Saigyo’s poetic journey at the beginning of “Shiramine,” Muzen’s journey with his son in “The Owl of the Three Jewels,” and Kaian’s journey in “The Blue Hood.” A fourth story, “The Carp of My Dreams,” includes a par- ticularly beautiful passage—Kogi’s tour of Lake Biwa—that resembles a michiyuki, although it does not serve the same structural function.

2. Shite and waki. A typical no play has two important characters: the shite (central figure) and the waki ([man at] the side). The shite is preceded on the stage by the waki, who is frequently a traveler or an itinerant monk. The waki arrives at a famous site, where he meets a local person—the shite—whom he questions about the history of the area. As the waki draws out the shite’s story, it turns out that the shite is actually the ghost of a historical figure who is still clinging to this world because of obsessive anger, desire for revenge, or love. Often the shite asks the waki to pray for his or her release from obsession so that he or she might be reborn in Amida Buddha’s Western Paradise. Characters in at least four of the stories in Moonlight and Rain resemble the waki and shite roles, in greater or lesser degrees. Saigyo, in “Shiramine,” corresponds to the waki as his michiyuki takes him to Shiramine. There he encounters the ghost of Sutoku, who corresponds to the shite, and they engage in a mondo (dialogue), a common element in no plays.80 As in a no play, Sutoku’s true form is revealed near the end of the

26 introduction

story. In “The Chrysanthemum Vow,” Samon resembles the waki insofar as he elicits a story from Soemon, who corre- sponds to the shite. Samon ceases to behave like a waki at the end of the story, however. In “The Owl of the Three Jewels,” Muzen corresponds to the waki (and his son to a wakizure [waki companion]) as he travels to Mount Koya. Hidetsugu’s ghost, of course, corresponds to the shite (and his follow- ers to shitezure), and the fact that Hidetsugu turns out to be an asura encourages the reader to recall no plays, such as Yashima, in which the shite is an asura. Finally, the itinerant monk Kaian functions as the waki in “The Blue Hood,” and the mad abbot can be thought of as the shite. The Moonlight and Rain stories are not no plays, of course, and it would be a mistake to press the waki–shite analogy too far, but the resemblance is unmistakable.

3. Dreams. In some no plays of the mugen (dream mys- tery) type, it is conceivable that the waki dozes off at the end of the first part of the play and dreams of his subsequent encounter with a ghost. This possibility exists in “Shira- mine” and “The Owl of the Three Jewels.” In “Shiramine,” Saigyo “began to doze” just before Sutoku makes his appear- ance, and the strange events in “The Owl of the Three Jew- els” take place after Muzen and his son have laid out their bedding and as they are trying to go to sleep. The central event of “The Carp of My Dreams” occurs in a dream, but parallels with mugen no are less obvious in this story than in the other two.

4. Jo-ha-kyu. The fourth structural element of no plays that is shared by the Moonlight and Rain stories is the jo-ha- kyu (introduction-development-climax) rhythm—common not only in no, but in all Japanese traditional performing arts—in which the performance begins slowly, the pace gradually quickens, and a swift, dramatic climax is reached at the end. The development, typically the longest section, also consists of three parts, resulting in five sections alto- gether. This structure is fairly clear in, for example, “The

introduction 27

Blue Hood.” In the introduction (jo), Kaian reaches Tonda and meets his host. In the first part of the development (ha), the host explains Kaian’s odd reception and the problem that has been troubling the village. Kaian responds in the second part of the development, goes to the mountain tem- ple, and meets the abbot. In the third part of the develop- ment, Kaian confronts the abbot’s madness. Finally, in the climax (kyu), Kaian returns to the temple a year later and meets the abbot again, resolving the crisis faced by both the village and the abbot. The same analysis could be applied fruitfully to “Shiramine,” “The Owl of the Three Jewels,” “The Kibitsu Cauldron,” and “A Serpent’s Lust.”81

The themes or motifs of several of the stories recall spe- cific no plays as well. “The Owl of the Three Jewels” resem- bles a shuramono (asura play), such as Yashima, in which the angry ghost of a warrior returns to the scene of his defeat. “The Reed-Choked House” employs the traditional theme of a woman who waits loyally and patiently for her husband or lover—a motif that is common in Japanese court poetry and in no plays such as Matsukaze (Wind in the Pines) and Izutsu (The Well Curb). “A Serpent’s Lust,” which explicitly alludes to the Dojoji legend, naturally recalls the no play Dojoji and its serpent-woman.82

structure of the collect ion

Contemporary publishers of short-story collections com- monly put the best—or the most appealing, exciting, or evocative—story first, in order to hook readers and keep them reading. If the publisher of Moonlight and Rain had followed this practice, he might have chosen to begin with “The Reed-Choked House,” “The Kibitsu Cauldron,” or “The Blue Hood,” rather than with the stately, poetic “Shiramine,” which opens the collection as we have it. The organization

28 introduction

of Moonlight and Rain is so unlike that with which Western readers are accustomed that Kengi Hamada, in his transla- tion, apparently found it necessary to rearrange the stories to conform more closely to Western tastes, beginning his ver- sion with “The Reed-Choked House.”83 In fact, the structure of the collection—the grouping and ordering of the stories— is complex and certainly not random.

The nine stories are arranged in five books:

Book One

“Shiramine”

“The Chrysanthemum Vow”

Book Two

“The Reed-Choked House”

“The Carp of My Dreams”

Book Three

“The Owl of the Three Jewels”

“The Kibitsu Cauldron”

Book Four

“A Serpent’s Lust”

Book Five

“The Blue Hood”

“On Poverty and Wealth”

This organization recalls two parallels: first, Teisho’s A Gar- land of Heroes and Shigeshige yawa, which likewise consist of nine stories in five books; and, second, the five-part jo-ha-kyu rhythm of no. This structure applies not only to individual plays, but also to the arrangement of the five kinds of plays into a five-play no program.84

To some extent, Moonlight and Rain can be compared with a program of no plays.85 The five types of plays and their sequence in a full program are shown in the table.

It is tempting to try to apply this program of five plays to the five books of Moonlight and Rain—to say that Book One corresponds to the first category of no plays, for example—

introduction 29

but the analogy breaks down almost immediately, since “The Chrysanthemum Vow” has little in common with a wakino or kamino, nor are “The Reed-Choked House” and “The Carp of My Dreams” at all analogous to shuramono. The five books of Moonlight and Rain do, however, parallel the five-part rhythmic structure of the no program, with a gradual heightening of pace and excitement. The stories in Book One unfold at an appropriately dignified pace, with elevated language and high-ranking characters: an eminent poet-monk and the ghost of a deified former emperor in “Shiramine” and a scholar and a samurai in “The Chrysan- themum Vow.” The excitement picks up in Book Two and continues to grow, as the status of the characters declines, through Book Four. In Book Five, the pace and excite- ment reach a climax in “The Blue Hood,” and the collec- tion ends with an auspicious prophecy of good times under the Tokugawa government, in “On Poverty and Wealth.” In addition, there are unmistakable parallels between the five types of no plays and the central characters of some of the stories and the order in which they appear. The ghost of the former emperor Sutoku, in “Shiramine,” can be thought of as a deity, insofar as all emperors were believed to be divine descendants of Amaterasu, the sun goddess, and this one was actually enshrined at Shiramine. The second story, “The Chrysanthemum Vow,” involves the ghost of a samurai, and “The Reed Choked House” focuses on a beautiful woman. With “The Carp of My Dreams” the collection departs from

Number Name Typical central character (shite)

1 wakino or kamino A deity

2 shuramono The ghost of a warrior

3 katsuramono A beautiful young woman

4 kyojomono or genzaimono A mad person or a modern person

5 kirino or ki(chiku)mono A demon or another strange creature

30 introduction

any correspondence with the subject matter of a no pro- gram, but “The Owl of the Three Jewels” resembles a shura- mono and draws the collection back toward the no pattern. “The Kibitsu Cauldron” can be seen as a combination of a katsuramono, kyojomono, and kirino, since Isora starts out as a beautiful young woman and ends up as a possessing spirit. Finally, each of the remaining stories focuses on an anomalous creature: a serpent-woman, a necrophiliac and cannibalistic monk, and the spirit of gold. The sequencing of the nine stories does, then, approximate a no program in its gradually accelerating pace and in the nature of the cen- tral characters. While the presence of the jo-ha-kyu rhythm might be explained as the automatic choice of a Japanese artist of the time, the downward progression of characters suggests a deliberate imitation of the no form.

There are other ways of looking at the organization of Moonlight and Rain.86 The most persuasive is the link- ing structure described by the prominent Akinari scholar Takada Mamoru.87

Stories 1–2. The clash of wills between a ghost and a liv- ing man in “Shiramine” is followed by the meeting of minds between a ghost and a living man in “The Chrysanthemum Vow.” (The first two stories are also linked by an interest in Confucianism and by the theme of loyalty—Saigyo’s to Sutoku, and Soemon’s and Samon’s to each other. The intimacy and loyalty that bond Saigyo to his former master, Sutoku, in “Shiramine” anticipate an even closer relation- ship between two men in “The Chrysanthemum Vow.”)

Stories 2–3. The fraternal loyalty of a ghost in “The Chry- santhemum Vow” is echoed by the marital fidelity of a ghost in “The Reed-Choked House,” in which the loyal ghost is the woman who waits, not the man who returns. (There is a link by contrast, as well, in that the man who vows to return by a certain date keeps his promise in “The Chrysan- themum Vow,” but not in “The Reed-Choked House.” The

introduction 31

parallel between the two stories is even closer if we think of Samon and Soemon as lovers: the steadfast love between two men in “The Chrysanthemum Vow” is followed by the undependability of a husband in “The Reed-Choked House.” Upper-class men of exemplary character are replaced by a pusillanimous farmer and his devoted wife. Miyagi, the ideal wife, anticipates another idealized wife, with a twist, in “The Kibitsu Cauldron,” and a mockery of the ideal wife plays a central role in “A Serpent’s Lust.”)

Stories 3–4. The image of water links “The Reed-Choked House” to “The Carp of My Dreams”: the former ends with the legend of a girl who threw herself into the water, and the latter concerns an artist who dreams that he swims about like a fish. (“The Carp of My Dreams” is a lighthearted, peaceful story with a happy ending, a cheerful interlude after the solemnity of “Shiramine” and “The Chrysanthe- mum Vow” and the pathos of “The Reed-Choked House” and before the darkness of “The Owl of the Three Jewels.” Like Saigyo, in “Shiramine,” Kogi was an eminent monk, and “The Carp of My Dreams” makes many references to Buddhist teachings. The water motif is picked up again in “A Serpent’s Lust.”)

Stories 4–5. Kogi, in “The Carp of My Dreams,” comes back from the strange world of a watery dream to tell his story; similarly, Muzen, in “The Owl of the Three Jewels,” barely returns alive from the strange world of Mount Koya and a brush with the asura realm to tell others about his experience. Daytime turns to night; fish is replaced by owl. (The stories are also linked by the prominent role that Buddhism plays in each. If Muzen dreams of his encounter with Hidetsugu, the connection becomes even closer. The implicit contrast in “The Owl of the Three Jewels” between the revered Kobo Daishi and the murderous Hidetsugu—both of them famous historical figures—echoes the contrast between Saigyo and Sutoku in “Shiramine” and anticipates that between Kaian and the mad monk in “The Blue Hood.”)

32 introduction

Stories 5–6. The story of a cruel man who fell to the asura realm, in “The Owl of the Three Jewels,” is followed, in “The Kibitsu Cauldron,” by that of a betrayed woman whose jeal- ousy turns her into an angry possessing spirit, and this time the man does not survive. (The Shinto context of “The Kib- itsu Cauldron,” as well as the amoral spirit of Isora, form a sharp contrast with the Buddhism of “Shiramine,” “The Carp of My Dreams,” and “The Owl of the Three Jewels.”)

Stories 6–7. The lascivious husband of “The Kibitsu Cauldron” is followed, in “A Serpent’s Lust,” by a lascivi- ous and jealous serpent-woman who, like the jealous wife of “The Kibitsu Cauldron,” tries to kill her husband. (The reference to serpent-women in the first paragraph of “The Kibitsu Cauldron” anticipates “A Serpent’s Lust.” Both Sho- taro, in “The Kibitsu Cauldron,” and Toyoo, in “A Serpent’s Lust,” are led by a servant girl to meet a mysterious lady at a house that turns out to be something other than what it first appears to be. In both stories, one or more wise old men tries to help: in “The Kibutsu Cauldron,” he is a yin–yang master, while in “A Serpent’s Lust,” the first is a Shinto priest; the second, an overrated Buddhist monk; and the third, a Buddhist sage. Pervasive water imagery in “A Serpent’s Lust” echoes that in “The Reed-Choked House” and “The Carp of My Dreams.”)

Stories 7–8. The destructive power of lust is an issue in “The Blue Hood,” as in “A Serpent’s Lust,” but the focus shifts from a serpent-woman to a previously upright monk who becomes a fiend, and the weak husband is replaced by a Zen “priest of great virtue.” (Buddhism returns as a central concern, as in “Shiramine,” “The Carp of My Dreams,” and “The Owl of the Three Jewels.”)

Stories 8–9. “The Blue Hood” is linked to “On Poverty and Wealth” by the important role of Chinese verse—the lines with which Kaian leads the mad abbot to enlightenment, and the prophetic lines bestowed on Sanai by the spirit of gold.

introduction 33

Stories 9–1. The last story in Moonlight and Rain, “On Pov- erty and Wealth,” is linked to the first story, “Shiramine,” since both involve a philosophical dialogue between a human and a nonhuman, and the gold spirit’s prediction of a happy future for the nation echoes, and contrasts with, Sutoku’s grim predictions of war. (Another parallel between the first and last stories is that “Shiramine” alludes to the Hogen, Heiji, and Gempei conflicts, while “On Poverty and Wealth” reviews the struggles of the Warring States and Azuchi– Momoyama periods before predicting peace and prosperity under the Tokugawa. Tributes to the Tokugawa are included in the “Preface” and “The Owl of the Three Jewels” as well.)

At least two other patterns in the structure of Moonlight and Rain are noteworthy. First, anomalous beings and events grow more dangerous as the collection progresses and then less threatening again. They present no danger to Saigyo, Samon, Katsushiro, and Kogi, in the first four stories. In the fifth, however, Muzen is nearly killed, and Shotaro dies in the sixth. The danger recedes slightly in the seventh story, in which Tomiko is killed but Toyoo escapes, and the eighth, in which Kaian is threatened but survives because of his great virtue. Danger is not an element in the ninth story. The other notable pattern is the contrast between a steadfast character and an undependable, erratic, or vicious one. The motif is established in “Shiramine,” with Saigyo’s unshakable loyalty and Sutoku’s thirst for revenge, and is repeated in all the stories except “The Carp of My Dreams” and “On Poverty and Wealth.” Further, the loyalty that Samon and Soemon show for each other in “The Chrysanthemum Vow” con- trasts with the weakness of the central male characters in “The Reed-Choked House,” “The Kibitsu Cauldron,” and “A Serpent’s Lust,” and the theme is driven home with the warning in “The Chrysanthemum Vow”: “Bond not with a shallow man.”

34 introduction

About the Translation

Tales of Moonlight and Rain has already attracted an impres- sive company of translators and adapters, and yet it is a tru- ism that English translations of the work have been inad- equate. Takada understated the problem: “Readers of Ugetsu Monogatari in translation may not be able fully to grasp the classic beauty of the . . . original’s literary style, its elegant phraseology, and its precise mode of expression, replete with concatenations yet without redundancy.”88 To reflect adequately in translation the style and tone of the original text is a tall order. Akinari was a great master of Japanese, but few of us who translate from Japanese are great masters of English. His prose is terse, elliptical, sinewy, highly lit- erary, allusive, scholarly, dignified, elegant, and sometimes obscure—never slack or insipid. His is a neoclassical, self- conscious, quirky style, with many usages borrowed from archaic Japanese texts and Chinese sources, resulting in a rich, dense text that is meant to be read slowly and savored. I have tried to let the text speak for itself as directly as possible, rather than embroidering it with interpretations and expla- nations. Some of the existing translations strike me as wordy and insufficiently dignified, because they have gone too far in accommodating the Western reader and so fail to convey the tone, pace, and elegance of the original. Leon M. Zolbrod wanted his translation “to read as if the original were written in common English.”89 In this, he echoed Dryden: “I have endeavoured to make Virgil speak such English as he would himself have spoken, if he had been born in England, and in this present age.”90 Commonsensical as it sounds, this is an impossible goal.91 In any case, Akinari certainly did not write in common Japanese.

Translators of Moonlight and Rain who attempted to “[leave] the reader alone as much as possible and [move] the writer toward the reader”92—in other words, to “naturalize” Akinari’s prose into modern English—have vitiated the text.

introduction 35

In my translation, I have tried to leave Akinari alone as much as possible without doing violence to my mother tongue. I have looked for ways to convey in English the distinctive qualities of Akinari’s prose more successfully than some of my predecessors have done, even though I am aware of my limitations as an English stylist and of the ultimate impossibility of a close approximation. Akinari’s original (as opposed to the modern, edited texts on which we base our translations) presents each tale as a steady narrative flow, uninterrupted by paragraphs or quotation marks and only occasionally guided by punctuation. One effect of this pre- sentation (which was common at the time) is to blur the dis- tinction between narrative and dialogue: early editions give the impression that both narrative and dialogue are told in the voice of the narrator. I considered dispensing with quo- tations marks and other modern techniques for setting off dialogue, but finally decided that this would make the trans- lation too odd and remote for the tastes of most readers. For the same reason, I have also used English tenses, despite the advice of those who emphasize the inherent differences between Japanese and English narrative.93 Probably these compromises will satisfy no one, but I hope that my transla- tion will bring readers of English a little closer to the tone, texture, and excitement of Akinari’s masterpiece.

Some readers may think that I have accepted Vladimir Nabokov’s advice to provide “copious footnotes, footnotes reaching up like skyscrapers.”94 In the case of Moonlight and Rain, I believe that extensive notes are desirable to explain exotic references and to demonstrate the rich intertextuality of the stories. Even so, the notes are far from exhaustive. I have emphasized sources that readers are most likely to be familiar with, such as The Tale of Genji, and secondary mate- rial in English. I hope that the notes will be sufficient for the scholarly reader and not too tedious for the casual. Informa- tion that is immediately useful for understanding the text is provided in footnotes; longer notes, of interest primarily to

36 introduction

students and scholars, appear at the end of each story. Truly comprehensive notes can be found in Uzuki Hiroshi’s indis- pensable Ugetsu monogatari hyoshaku. Many of my notes paraphrase Uzuki’s.